Datasets:

---

bigcode

/

jupyter-parsed

like 4

Follow

BigCode 1.4k

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import torch\nfrom torch import optim\nimport torch.nn as nn\nimport torch.nn.functional as F\nimport torch.autograd as autograd\nfrom torch.autograd import Variable\n\nfrom sklearn.preprocessing import OneHotEncoder\nimport os, math, glob, argparse\nfrom utils.torch_utils import *\nfrom utils.utils import *\nfrom mpradragonn_predictor_pytorch import *\nimport matplotlib.pyplot as plt\nimport utils.language_helpers\n#plt.switch_backend('agg')\nimport numpy as np\nfrom models import *\n\nfrom wgan_gp_mpradragonn_analyzer_quantile_cutoff import *\n\nuse_cuda = torch.cuda.is_available()\ndevice = torch.device('cuda:0' if use_cuda else 'cpu')\n\nfrom torch.distributions import Normal as torch_normal\n\nclass IdentityEncoder :\n \n def __init__(self, seq_len, channel_map) :\n self.seq_len = seq_len\n self.n_channels = len(channel_map)\n self.encode_map = channel_map\n self.decode_map = {\n nt: ix for ix, nt in self.encode_map.items()\n }\n \n def encode(self, seq) :\n encoding = np.zeros((self.seq_len, self.n_channels))\n \n for i in range(len(seq)) :\n if seq[i] in self.encode_map :\n channel_ix = self.encode_map[seq[i]]\n encoding[i, channel_ix] = 1.\n\n return encoding\n \n def encode_inplace(self, seq, encoding) :\n for i in range(len(seq)) :\n if seq[i] in self.encode_map :\n channel_ix = self.encode_map[seq[i]]\n encoding[i, channel_ix] = 1.\n \n def encode_inplace_sparse(self, seq, encoding_mat, row_index) :\n raise NotImplementError()\n \n def decode(self, encoding) :\n seq = ''\n \n for pos in range(0, encoding.shape[0]) :\n argmax_nt = np.argmax(encoding[pos, :])\n max_nt = np.max(encoding[pos, :])\n seq += self.decode_map[argmax_nt]\n\n return seq\n \n def decode_sparse(self, encoding_mat, row_index) :\n raise NotImplementError()\n\n", "/home/ubuntu/anaconda3/envs/pytorch_p36_fbgan/lib/python3.6/site-packages/h5py/__init__.py:34: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n from ._conv import register_converters as _register_converters\n" ], [ "\nclass ActivationMaximizer(nn.Module) :\n \n def __init__(self, generator_dir, batch_size=1, seq_len=145, latent_size=128, sequence_template=None):\n super(ActivationMaximizer, self).__init__()\n self.generator = Generator_lang(4, seq_len, batch_size, 512)\n self.predictor = DragoNNClassifier(batch_size=batch_size).cnn\n \n self.load_generator(generator_dir)\n \n self.use_cuda = torch.cuda.is_available()\n \n self.x_mask = None\n self.x_template = None\n if sequence_template is not None :\n onehot_mask = np.zeros((seq_len, 4))\n onehot_template = np.zeros((seq_len, 4))\n\n for j in range(len(sequence_template)) :\n if sequence_template[j] == 'N' :\n onehot_mask[j, :] = 1.\n elif sequence_template[j] == 'A' :\n onehot_template[j, 0] = 1.\n elif sequence_template[j] == 'C' :\n onehot_template[j, 1] = 1.\n elif sequence_template[j] == 'G' :\n onehot_template[j, 2] = 1.\n elif sequence_template[j] == 'T' :\n onehot_template[j, 3] = 1.\n \n self.x_mask = Variable(torch.FloatTensor(onehot_mask).unsqueeze(0))\n self.x_template = Variable(torch.FloatTensor(onehot_template).unsqueeze(0))\n if self.use_cuda :\n self.x_mask = self.x_mask.to(device)\n self.x_template = self.x_template.to(device)\n \n self.predictor.eval()\n \n if self.use_cuda :\n self.generator.cuda()\n self.predictor.cuda()\n self.cuda()\n\n def load_generator(self, directory, iteration=None) :\n list_generator = glob.glob(directory + \"G*.pth\")\n generator_file = max(list_generator, key=os.path.getctime)\n self.generator.load_state_dict(torch.load(generator_file))\n \n def forward(self, z) :\n x = self.generator.forward(z)\n \n if self.x_mask is not None :\n x = x * self.x_mask + self.x_template\n \n return self.predictor.forward(x.unsqueeze(2).transpose(1, 3))\n \n def get_pattern(self, z) :\n x = self.generator.forward(z)\n \n if self.x_mask is not None :\n x = x * self.x_mask + self.x_template\n \n return x\n ", "_____no_output_____" ], [ "#Sequence length\nseq_len = 145\nbatch_size = 64\n\n#Sequence decoder\nacgt_encoder = IdentityEncoder(seq_len, {'A':0, 'C':1, 'G':2, 'T':3})\n\n#Sequence template\nsequence_template = 'N' * 145\n\n#Activation maximization model (pytorch)\nact_maximizer = ActivationMaximizer(batch_size=batch_size, seq_len=seq_len, generator_dir='./checkpoint/' + 'mpradragonn_sample' + '/', sequence_template=sequence_template)\n", "[*] Checkpoint 10 found!\n" ], [ "\n#Function for optimizing n sequences for a target predictor\ndef optimize_sequences(act_maximizer, n_seqs, batch_size=1, latent_size=128, n_iters=100, eps1=0., eps2=0.1, noise_std=1e-6, use_adam=True, run_name='default', store_intermediate_n_seqs=None, store_every_iter=100) :\n\n z = Variable(torch.randn(batch_size, latent_size, device=\"cuda\"), requires_grad=True)\n\n norm_var = torch_normal(0, 1)\n\n optimizer = None\n if use_adam :\n optimizer = optim.Adam([z], lr=eps2)\n else :\n optimizer = optim.SGD([z], lr=1)\n\n z.register_hook(lambda grad, batch_size=batch_size, latent_size=latent_size, noise_std=noise_std: grad + noise_std * torch.randn(batch_size, latent_size, device=\"cuda\"))\n\n seqs = []\n fitness_histo = []\n \n n_batches = n_seqs // batch_size\n \n for batch_i in range(n_batches) :\n \n if batch_i % 4 == 0 :\n print(\"Optimizing sequence batch \" + str(batch_i))\n \n #Re-initialize latent GAN seed\n z.data = torch.randn(batch_size, latent_size, device=\"cuda\")\n \n fitness_scores_batch = [act_maximizer(z)[:, 0].data.cpu().numpy().reshape(-1, 1)]\n\n for curr_iter in range(n_iters) :\n\n fitness_score = act_maximizer(z)[:, 0]\n \n fitness_loss = -torch.sum(fitness_score)\n z_prior = -torch.sum(norm_var.log_prob(z))\n\n loss = None\n if use_adam :\n loss = fitness_loss\n else :\n loss = eps1 * z_prior + eps2 * fitness_loss\n\n optimizer.zero_grad()\n loss.backward()\n optimizer.step()\n\n fitness_scores_batch.append(fitness_score.data.cpu().numpy().reshape(-1, 1))\n \n if store_intermediate_n_seqs is not None and batch_i * batch_size < store_intermediate_n_seqs and curr_iter % store_every_iter == 0 :\n onehot_batch = act_maximizer.get_pattern(z).data.cpu().numpy()\n seq_batch = [\n acgt_encoder.decode(onehot_batch[k]) for k in range(onehot_batch.shape[0])\n ]\n with open(run_name + \"_curr_iter_\" + str(curr_iter) + \".txt\", \"a+\") as f :\n for i in range(len(seq_batch)) :\n seq = seq_batch[i]\n\n f.write(seq + \"\\n\")\n \n onehot_batch = act_maximizer.get_pattern(z).data.cpu().numpy()\n seq_batch = [\n acgt_encoder.decode(onehot_batch[k]) for k in range(onehot_batch.shape[0])\n ]\n\n seqs.extend(seq_batch)\n fitness_histo.append(np.concatenate(fitness_scores_batch, axis=1))\n \n fitness_histo = np.concatenate(fitness_histo, axis=0)\n \n return seqs, fitness_histo\n", "_____no_output_____" ], [ "\nn_seqs = 4096#960\nn_iters = 1000\n\nrun_name = 'killoran_mpradragonn_' + str(n_seqs) + \"_sequences\" + \"_\" + str(n_iters) + \"_iters_sample_wgan\"\n\nseqs, fitness_scores = optimize_sequences(\n act_maximizer,\n n_seqs,\n batch_size=64,\n latent_size=128,\n n_iters=n_iters,\n eps1=0.,\n eps2=0.1,\n noise_std=1e-6,\n use_adam=True,\n run_name=\"samples/killoran_mpradragonn/\" + run_name,\n store_intermediate_n_seqs=None,#960,\n store_every_iter=100\n)\n", "Optimizing sequence batch 0\nOptimizing sequence batch 4\nOptimizing sequence batch 8\nOptimizing sequence batch 12\nOptimizing sequence batch 16\nOptimizing sequence batch 20\nOptimizing sequence batch 24\nOptimizing sequence batch 28\nOptimizing sequence batch 32\nOptimizing sequence batch 36\nOptimizing sequence batch 40\nOptimizing sequence batch 44\nOptimizing sequence batch 48\nOptimizing sequence batch 52\nOptimizing sequence batch 56\nOptimizing sequence batch 60\n" ], [ "#Plot fitness statistics of optimization runs\n\n#Plot k trajectories\nplot_n_traj = 100\n\nf = plt.figure(figsize=(8, 6))\n\nfor i in range(min(plot_n_traj, n_seqs)) :\n plt.plot(fitness_scores[i, :], linewidth=2, alpha=0.75)\n\nplt.xlabel(\"Training iteration\", fontsize=14)\nplt.ylabel(\"Fitness score\", fontsize=14)\n\nplt.xticks(fontsize=14)\nplt.yticks(fontsize=14)\n\nplt.xlim(0, n_iters)\nplt.ylim(-3, 3)\n\nplt.tight_layout()\nplt.show()\n\n#Plot mean trajectory\n\nf = plt.figure(figsize=(8, 6))\n\nplt.plot(np.mean(fitness_scores, axis=0), linewidth=2, alpha=0.75)\n\nplt.xlabel(\"Training iteration\", fontsize=14)\nplt.ylabel(\"Fitness score\", fontsize=14)\n\nplt.xticks(fontsize=14)\nplt.yticks(fontsize=14)\n\nplt.xlim(0, n_iters)\nplt.ylim(-3, 3)\n\nplt.tight_layout()\nplt.show()\n", "_____no_output_____" ], [ "#Save sequences to file\n\nwith open(run_name + \".txt\", \"wt\") as f :\n for i in range(len(seqs)) :\n seq = seqs[i]\n\n f.write(seq + \"\\n\")\n", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# Plotting and Functions", "_____no_output_____" ], [ "This notebook will work trough how to plot data and how to define functions. Throughout the lecture we will take a few moments to plot different functions and see how they depend on their parameters", "_____no_output_____" ], [ "## Plotting in Python: Matplot ", "_____no_output_____" ] ], [ [ "import matplotlib.pyplot as plt\nimport numpy as np\nimport scipy as sp", "_____no_output_____" ] ], [ [ "Pyplot is a powerful plotting library that can be used to make publication quaility plots. It is also useful for quikly plotting the results of a calcualtion. \n\nThis is a quick demonstration of its use\n\nNote: when you call al library `import matplotlib.pyplot as plt` the way that use it is to do the following `plt.function()` where `function()` is whatever you are trying to call from the library", "_____no_output_____" ] ], [ [ "# Define x and y values for some function\nx = [i for i in range(20)]\ny1 = [i**2 for i in x]\ny2 = [i**3 for i in x]", "_____no_output_____" ] ], [ [ "The methods used above to make the lists is considered very *pythonic*. It works the same as a loop, but outputs all the results into a list. The left-hand most argument is what the list elements will be and the right hand side is the the way the loop will work.", "_____no_output_____" ], [ "When you use pyplot to make a plot, you can add more than one data set to the figure until you render the plot. Once you render the plot it resets", "_____no_output_____" ] ], [ [ "plt.plot(x,y1)\nplt.plot(x,y2)\nplt.xlabel('X', fontsize=24)\nplt.ylabel('Y', fontsize=24)\nplt.legend(['Quadratic', 'Cubic'], loc=0)\nplt.show()", "_____no_output_____" ] ], [ [ "We can call also use numpy fucntions to make our plots. Numpy is a very powerful math library", "_____no_output_____" ] ], [ [ "# linspace will make a list of values from initial to final with however many increments you want\n# this example goes from 0-2.5 with 20 increments\nx=numpy.linspace(0,1.0,20)\nprint(x)", "_____no_output_____" ], [ "exp_func=np.exp(-2*np.pi*x)\nprint(exp_func)", "_____no_output_____" ], [ "plt.plot(x,exp_func, color=\"black\")\nplt.xlabel('x', fontsize=24)\nplt.ylabel(\"y(x)\", fontsize=24)\nplt.show()", "_____no_output_____" ] ], [ [ "All aspects of the plot can be changed. The best way to figure out what you want to do is to go to the Matplotlib gallery and choose an image that looks like what you are trying to do.\n\nhttps://matplotlib.org/gallery/index.html", "_____no_output_____" ], [ "### Example: Scatter plot with histograms", "_____no_output_____" ] ], [ [ "import numpy as np\n\n#Fixing random state for reproducibility\nnp.random.seed(19680801)\n\n# the random data\nx = np.random.randn(1000)\ny = np.random.randn(1000)\n\n# definitions for the axes\nleft, width = 0.1, 0.65\nbottom, height = 0.1, 0.65\nspacing = 0.005\n\n\nrect_scatter = [left, bottom, width, height]\nrect_histx = [left, bottom + height + spacing, width, 0.2]\nrect_histy = [left + width + spacing, bottom, 0.2, height]\n\n# start with a rectangular Figure\nplt.figure(figsize=(8, 8))\n\nax_scatter = plt.axes(rect_scatter)\nax_scatter.tick_params(direction='in', top=True, right=True)\nax_histx = plt.axes(rect_histx)\nax_histx.tick_params(direction='in', labelbottom=False)\nax_histy = plt.axes(rect_histy)\nax_histy.tick_params(direction='in', labelleft=False)\n\n# the scatter plot:\nax_scatter.scatter(x, y)\n\n# now determine nice limits by hand:\nbinwidth = 0.25\nlim = np.ceil(np.abs([x, y]).max() / binwidth) * binwidth\nax_scatter.set_xlim((-lim, lim))\nax_scatter.set_ylim((-lim, lim))\n\nbins = np.arange(-lim, lim + binwidth, binwidth)\nax_histx.hist(x, bins=bins)\nax_histy.hist(y, bins=bins, orientation='horizontal')\n\nax_histx.set_xlim(ax_scatter.get_xlim())\nax_histy.set_ylim(ax_scatter.get_ylim())\n\nplt.show()", "_____no_output_____" ] ], [ [ "I don't have to be an expert in making that kind of plot. I just have to understand and guess enough to figure out. I also google things I don't know\n\nhttps://www.google.com/search?client=firefox-b-1-d&q=pyplot+histogram+change+color\n\nhttps://stackoverflow.com/questions/42172440/python-matplotlib-histogram-color?rq=1\n\nhttps://matplotlib.org/examples/color/named_colors.html\n\nThen I can make small changes to have the plot look how I want it to look\n\nNotice below I changed \n\n`ax_scatter.scatter(x, y, color=\"purple\")`, \n\n`ax_histx.hist(x, bins=bins, color = \"skyblue\")`, \n\n`ax_histy.hist(y, bins=bins, orientation='horizontal', color=\"salmon\")`", "_____no_output_____" ] ], [ [ "#Fixing random state for reproducibility\nnp.random.seed(19680801)\n\n# the random data\nx = np.random.randn(1000)\ny = np.random.randn(1000)\n\n# definitions for the axes\nleft, width = 0.1, 0.65\nbottom, height = 0.1, 0.65\nspacing = 0.005\n\n\nrect_scatter = [left, bottom, width, height]\nrect_histx = [left, bottom + height + spacing, width, 0.2]\nrect_histy = [left + width + spacing, bottom, 0.2, height]\n\n# start with a rectangular Figure\nplt.figure(figsize=(8, 8))\n\nax_scatter = plt.axes(rect_scatter)\nax_scatter.tick_params(direction='in', top=True, right=True)\nax_histx = plt.axes(rect_histx)\nax_histx.tick_params(direction='in', labelbottom=False)\nax_histy = plt.axes(rect_histy)\nax_histy.tick_params(direction='in', labelleft=False)\n\n# the scatter plot:\nax_scatter.scatter(x, y, color=\"purple\")\n\n# now determine nice limits by hand:\nbinwidth = 0.25\nlim = np.ceil(np.abs([x, y]).max() / binwidth) * binwidth\nax_scatter.set_xlim((-lim, lim))\nax_scatter.set_ylim((-lim, lim))\n\nbins = np.arange(-lim, lim + binwidth, binwidth)\nax_histx.hist(x, bins=bins, color = \"skyblue\")\nax_histy.hist(y, bins=bins, orientation='horizontal', color=\"salmon\")\n\nax_histx.set_xlim(ax_scatter.get_xlim())\nax_histy.set_ylim(ax_scatter.get_ylim())\n\n\n\nplt.show()", "_____no_output_____" ] ], [ [ "Notice how I changed the colors on the plot based off of what I found on the stack exchange. The way to solve issues in the course and computational work is to google them.", "_____no_output_____" ], [ "## Plotting Exersice 1", "_____no_output_____" ], [ "Find a plot from the gallery that you like. Then make some sort of noticable change to it.", "_____no_output_____" ] ], [ [ "import numpy as np\nimport matplotlib.pyplot as plt\nfrom matplotlib.lines import Line2D\n\n\npoints = np.ones(5) # Draw 5 points for each line\nmarker_style = dict(color='tab:blue', linestyle=':', marker='o',\n markersize=15, markerfacecoloralt='tab:red')\n\nfig, ax = plt.subplots()\n\n# Plot all fill styles.\nfor y, fill_style in enumerate(Line2D.fillStyles):\n ax.text(-0.5, y, repr(fill_style),\n horizontalalignment='center', verticalalignment='center')\n ax.plot(y * points, fillstyle=fill_style, **marker_style)\n\nax.set_axis_off()\nax.set_title('fill style')\n\nplt.show()", "_____no_output_____" ], [ "import numpy as np\nimport matplotlib.pyplot as plt\nfrom matplotlib.lines import Line2D\n\n\npoints = np.ones(5) # Draw 5 points for each line\nmarker_style = dict(color='tab:green', linestyle=':', marker='o',\n markersize=15, markerfacecoloralt='tab:purple')\n\nfig, ax = plt.subplots()\n\n# Plot all fill styles.\nfor y, fill_style in enumerate(Line2D.fillStyles):\n ax.text(-0.5, y, repr(fill_style),\n horizontalalignment='center', verticalalignment='center')\n ax.plot(y * points, fillstyle=fill_style, **marker_style)\n\nax.set_axis_off()\nax.set_title('fill style')\n\nplt.show()", "_____no_output_____" ] ], [ [ "## Plotting Exersice 2", "_____no_output_____" ], [ "Plot a the following functions on the same plot from $ -2\\pi $ to $2\\pi$\n\n$$ \\sin(2\\pi x+\\pi)$$\n$$ \\cos(2\\pi x+\\pi)$$\n$$\\sin(2\\pi x+\\pi)+\\cos(2\\pi x+\\pi)$$", "_____no_output_____" ], [ "This might be useful:\nhttps://docs.scipy.org/doc/numpy/reference/generated/numpy.sin.html\nhttps://docs.scipy.org/doc/numpy/reference/generated/numpy.cos.html#numpy.cos", "_____no_output_____" ] ], [ [ "np.sin(np.pi/2.)\n", "_____no_output_____" ], [ "np.sin(np.array((0., 30., 45., 60., 90.)) * np.pi / 180. )\n", "_____no_output_____" ], [ "import matplotlib.pylab as plt\nx = np.linspace(-np.pi, np.pi, 201)\nplt.plot(x, np.sin(x))\nplt.xlabel('Angle [rad]')\nplt.ylabel('sin(x)')\nplt.axis('tight')\nplt.show()", "_____no_output_____" ], [ "import matplotlib.pylab as plt\nx = np.linspace(-2*np.pi, 2*np.pi, 201)\nplt.plot(x, np.sin(x))\nplt.xlabel('Angle [rad]')\nplt.ylabel('sin(x)')\nplt.axis('tight')\nplt.show()", "_____no_output_____" ], [ "import matplotlib.pylab as plt\nx = np.linspace(-2*np.pi, 2*np.pi, 201)\nplt.plot(x, np.sin(2*np.pi*x+np.pi))\nplt.plot(x, np.cos(2*np.pi*x+np.pi))\nplt.xlabel('Angle [rad]')\nplt.ylabel('sin(x)')\nplt.axis('tight')\nplt.show()", "_____no_output_____" ], [ "import matplotlib.pylab as plt\nx = np.linspace(-2*np.pi, 2*np.pi, 201)\nplt.plot(x, np.cos(2*np.pi*x+np.pi))\nplt.xlabel('Angle [rad]')\nplt.ylabel('cos(x)')\nplt.axis('tight')\nplt.show()", "_____no_output_____" ], [ "import matplotlib.pylab as plt\nx = np.linspace(-2*np.pi, 2*np.pi, 201)\nplt.plot(x, np.sin(2*np.pi*x+np.pi), color=\"blue\")\nplt.plot(x, np.cos(2*np.pi*x+np.pi), color=\"red\")\nplt.plot(x, np.sin(2*np.pi*x+np.pi)+np.cos(2*np.pi*x+np.pi), color=\"green\")\n\nplt.xlabel('x')\nplt.ylabel('y(x)')\nplt.axis('tight')\n\nplt.show()", "_____no_output_____" ], [ "import matplotlib.pylab as plt\nx = np.linspace(-2*np.pi, 2*np.pi, 201)\nplt.plot(x, np.sin(2*np.pi*x+np.pi), color=\"black\")\nplt.plot(x, np.cos(2*np.pi*x+np.pi), color=\"red\")\nplt.plot(x, np.sin(2*np.pi*x+np.pi)+np.cos(2*np.pi*x+np.pi), color=\"gray\")\n\nplt.xlabel('x')\nplt.ylabel('y(x)')\nplt.axis('tight')\n\nplt.show()", "_____no_output_____" ] ], [ [ "# Lecture plots", "_____no_output_____" ], [ "Periodically during lecture we will take a pause to plot some of the interesting functions that we use in class.", "_____no_output_____" ], [ "## Classical wavefunctions\n\nThe following plot shows the the spacial component of the standard wavefunction with a wavelength of $\\lambda=\\text{1.45 m}$ and a relative amplitude of $A=1$ when the time, $t=0$ and the phase $\\phi=1.0$.", "_____no_output_____" ] ], [ [ "x=np.linspace(0,3.0,100)\nsinx=np.sin(2*np.pi*x+0+1)\nplt.plot(x,sinx, color=\"black\")\nplt.xlabel('x', fontsize=24)\nplt.ylabel(\"y(x)\", fontsize=24)\nplt.show()", "_____no_output_____" ] ], [ [ "Make a new figure where you plot the same wave function at three time points in the future. Assume the frequency is $\\nu=.1 \\text{ ms / s} $ Use a different color for each plot", "_____no_output_____" ], [ "## Orthogonality", "_____no_output_____" ], [ "Graphically show that the the following two functions are orthogonal on the interval $-3\\pi$ to $3\\pi$\n$$ \\sin(x) \\text{ and } \\cos(3x)$$\n\nPlot both functions together, then plot the product of both functions and explain why it is orthogonal", "_____no_output_____" ] ], [ [ "import matplotlib.pylab as plt\nx = np.linspace(-3*np.pi, 3*np.pi, 201)\nplt.plot(x, np.sin(x))\nplt.xlabel('Angle [rad]')\nplt.ylabel('sin(x)')\nplt.axis('tight')\nplt.show()", "_____no_output_____" ], [ "import matplotlib.pylab as plt\nx = np.linspace(-3*np.pi, 3*np.pi, 201)\nplt.plot(x, np.sin(x))\nplt.plot(x, np.cos(3*x))\nplt.xlabel('Angle [rad]')\nplt.ylabel('sin(x)')\nplt.axis('tight')\nplt.show()", "_____no_output_____" ], [ "import matplotlib.pylab as plt\nx = np.linspace(-3*np.pi, 3*np.pi, 201)\nprod=np.sin(x)*np.cos(3*x)\nplt.plot(x, np.sin(x))\nplt.plot(x, np.cos(3*x))\nplt.xlabel('Angle [rad]')\nplt.ylabel('sin(x)')\nplt.axis('tight')\nplt.show()", "_____no_output_____" ], [ "import matplotlib.pylab as plt\nx = np.linspace(-3*np.pi, 3*np.pi, 201)\nprod=np.sin(x)*np.cos(3*x)\nplt.plot(x, prod, color=\"blue\")\nplt.xlabel('Angle [rad]')\nplt.ylabel('sin(x)')\nplt.axis('tight')\nplt.show()", "_____no_output_____" ], [ "prod=np.sin(x)*np.cos(3*x)\n\n\n", "_____no_output_____" ], [ "prod=np.sin(x)*np.cos(3*x)\nx = np.linspace(-3*np.pi, 3*np.pi, 201)\nexp_func=prod\nnp.trapz(exp_func,x)", "_____no_output_____" ] ], [ [ "Use the numpy trapezoid rule integrator to show the the two functions are orthogonal\n`np.trapz(y,x)`\n\nhttps://docs.scipy.org/doc/numpy/reference/generated/numpy.trapz.html", "_____no_output_____" ] ], [ [ "# Example code\nx=numpy.linspace(0,1.0,20)\nexp_func=np.exp(-2*np.pi*x)\nnp.trapz(exp_func,x)", "_____no_output_____" ], [ "# Your code here", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import os\nimport sys\nsys.path.append('../')\n\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport pandas as pd\nfrom pprint import pprint\nfrom scipy.optimize import curve_fit\n\nimport src.io as sio\nimport src.preprocessing as spp\nimport src.fitting as sft", "_____no_output_____" ], [ "AFM_FOLDER = sio.get_folderpath(\"20200818_Akiyama_AFM\")\nAFM_FOLDER1 = sio.get_folderpath(\"20200721_Akiyama_AFM\")\nAFM_FOLDER2 = sio.get_folderpath(\"20200824_Akiyama_AFM\")\nAFM_FOLDER3 = sio.get_folderpath(\"20200826_TFSC_Preamp_AFM/11613_Tip_5/Akiyama_Tip_Stage\")\nAFM_FOLDER4 = sio.get_folderpath(\"20200826_TFSC_Preamp_AFM/11613_Tip_5/Custom_Tip_Stage\")\nAFM_FOLDER5 = sio.get_folderpath(\"20200828_Tip_Approach1\")\nAFM_FOLDER6 = sio.get_folderpath(\"20200901_Tip_Approach_2/Actual_tip_approach\")", "_____no_output_____" ] ], [ [ "# Approach", "_____no_output_____" ] ], [ [ "params, data = sio.read_dat(AFM_FOLDER6 + \"HistoryData001.dat\")\namplitude = data[\"Amplitude (m)\"].values\nfig, ax = plt.subplots()\nax.plot(amplitude*1e9)\nax.set_ylabel(\"Amplitude (nm)\")\nax.set_xlabel(\"Time (a.u.)\")\n#plt.savefig(\"snap.jpg\", dpi=600)", "_____no_output_____" ] ], [ [ "## 20200721_Akiyama_AFM", "_____no_output_____" ] ], [ [ "params, data = sio.read_dat(AFM_FOLDER1 + \"frq-sweep002.dat\")\nfreq_shift = data[\"Frequency Shift (Hz)\"].values\namplitude = data[\"Amplitude (m)\"].values\nphase = data[\"Phase (deg)\"].values\namp_freq_sweep = sft.fit_fano(freq_shift, amplitude, linear_offset=True)\nphase_freq_sweep = sft.fit_fano(freq_shift, phase)", "_____no_output_____" ], [ "%matplotlib inline\n\nfig, (ax1, ax2) = plt.subplots(nrows=2, sharex=True)\nax1.plot(freq_shift, amplitude*1e12)\n#ax1.plot(freq_shift, amp_freq_sweep.best_fit)\nax1.set_ylabel(\"Amplitude (pm)\")\n\nax2.plot(freq_shift, phase)\n#ax2.plot(freq_shift, phase_freq_sweep.best_fit)\nax2.set_ylabel(data.columns[3])\nax2.set_xlabel(data.columns[0])\n\n#plt.savefig(\"second.jpg\", dpi=600)", "_____no_output_____" ] ], [ [ "Quality factor can be calculated as $ Q = \\frac{f_R}{\\Delta f} $", "_____no_output_____" ] ], [ [ "print(f'Q-factor= {params[\"f_res (Hz)\"] / amp_freq_sweep.params[\"fwhm\"].value}')", "_____no_output_____" ] ], [ [ "## 20200818_Akiyama_AFM", "_____no_output_____" ] ], [ [ "params, data = sio.read_dat(AFM_FOLDER + \"frq-sweep001.dat\")\n#pprint(params, sort_dicts=False)\nfreq_shift = data[\"Frequency Shift (Hz)\"]\namplitude = data[\"Amplitude (m)\"]\nphase = data[\"Phase (deg)\"]\nfano = sft.fit_fano(freq_shift, amplitude)\nlorentzian = sft.fit_fano(freq_shift, phase)\nparams", "_____no_output_____" ] ], [ [ "## Equations for calculating Q factor\n\n$$ Q = \\frac{f_R}{\\Delta f} $$\n\n$$ Q = \\frac{A(\\omega_0)}{A_{in}} $$", "_____no_output_____" ] ], [ [ "f_res = 44379.7064\nsigma = 62.2841355\nprint(f_res/sigma)\n\nA_drive = 50e-3\nA_res = 28.3e-6 * 1 / 500e-6\nprint(A_res/A_drive)\n\n# Calibration\nA_drive = 50e-3\nosc_amp = 50e-9\n\nprint(osc_amp/A_drive)", "712.5362830154398\n1.132\n1e-06\n" ] ], [ [ "## Plot frequency sweep curves", "_____no_output_____" ] ], [ [ "fig, (ax1, ax2) = plt.subplots(nrows=2, sharex=True)\nax1.plot(freq_shift, amplitude)\nax1.plot(freq_shift, fano.best_fit)\nax1.set_ylabel(data.columns[2])\n\nax2.plot(freq_shift, phase)\nax2.plot(freq_shift, lorentzian.best_fit)\nax2.set_ylabel(data.columns[3])\nax2.set_xlabel(data.columns[1])", "_____no_output_____" ] ], [ [ "## Extract fit values", "_____no_output_____" ] ], [ [ "print(\"{} = {:.1f} +- {:.1f}\".format(fano.params[\"sigma\"].name, fano.params[\"sigma\"].value, fano.params[\"sigma\"].stderr))\nprint(\"{} = {:.2e} +- {:.0e}\".format(fano.params[\"amplitude\"].name, fano.params[\"amplitude\"].value, fano.params[\"amplitude\"].stderr))", "_____no_output_____" ] ], [ [ "# 20200824_Akiyama_AFM\n\n## Automatically read files from disk\n\nReads all files stored in **AFM_FOLDER2 = \"20200824_Akiyama_AFM/\"** and plots the amplitude and phase data.\n\nOptionally, the data can be fit to Fano resonances by setting the variable\n```python\nfit = True\n```\nThe Q-factor is calculated as:\n\n$$ Q = \\frac{f_R}{\\Delta f} = \\frac{f_R}{2 \\sigma} $$\n\nErrors are calculated as (this also gives an estimate of the SNR):\n$$ \\frac{\\Delta Q}{Q} = \\sqrt{ \\left( \\frac{\\Delta (\\Delta f)}{\\Delta f} \\right)^2 + \\left( \\frac{\\Delta (\\sigma)}{\\sigma} \\right)^2 } $$\n\nAnother estimate of the SNR, is the Chi square or weighted sum of squared deviations (lower is better):\n$$ \\chi^2 = \\sum_{i} {\\frac{(O_i - C_i)^2}{\\sigma_i^2}} $$", "_____no_output_____" ] ], [ [ "%matplotlib inline\n\nfit = False # Setting to True will take slightly longer due to the fitting protocols\n\nfiles = []\nfor file in os.listdir(AFM_FOLDER2):\n if file.endswith(\".dat\"):\n files.append(file) \n\nfig, ax = plt.subplots(nrows=len(files), ncols=2)\n\nfor idx, file in enumerate(files):\n params, data = sio.read_dat(AFM_FOLDER2 + file)\n freq_shift = data[\"Frequency Shift (Hz)\"]\n amplitude = data[\"Amplitude (m)\"]\n phase = data[\"Phase (deg)\"]\n\n ax[idx, 0].plot(freq_shift, amplitude)\n ax[idx, 0].set_ylabel(data.columns[2])\n ax[idx, 0].set_title(file)\n\n ax[idx, 1].plot(freq_shift, phase)\n ax[idx, 1].set_ylabel(data.columns[3])\n ax[idx, 1].set_title(file)\n \n if fit:\n fano1 = sft.fit_fano(freq_shift, amplitude)\n q_factor = (params[\"Center Frequency (Hz)\"] + fano1.params[\"center\"].value) / (2 * fano1.params[\"sigma\"].value)\n q_factor_err = q_factor * np.sqrt((fano1.params[\"center\"].stderr/fano1.params[\"center\"].value)**2 + (fano1.params[\"sigma\"].stderr/fano1.params[\"sigma\"].value)**2)\n ax[idx, 0].plot(freq_shift, fano1.best_fit, label=\"Q={:.0f}$\\pm{:.0f}$\".format(q_factor, q_factor_err))\n ax[idx, 0].legend()\n fano2 = sft.fit_fano(freq_shift, phase, linear_offset=True)\n ax[idx, 1].plot(freq_shift, fano2.best_fit)\n print(\"chi-square ({}) = {:.2e}\".format(file, fano1.chisqr))\n\nfig.tight_layout()\nfig.text(0.5, 0.02, data.columns[1], ha='center', va='center')", "_____no_output_____" ] ], [ [ "## 20200826_TFSC_Preamp_AFM \n### 11613_Tip_5", "_____no_output_____" ] ], [ [ "%matplotlib inline\n\nfit = False # Setting to True will take slightly longer due to the fitting protocols\n\nfiles = []\nfor file in os.listdir(AFM_FOLDER4):\n if file.endswith(\".dat\"):\n files.append(file) \n\nfig, ax = plt.subplots(nrows=len(files), ncols=2)\n\nfor idx, file in enumerate(files):\n params, data = sio.read_dat(AFM_FOLDER4 + file)\n freq_shift = data[\"Frequency Shift (Hz)\"]\n amplitude = data[\"Amplitude (m)\"]\n phase = data[\"Phase (deg)\"]\n\n ax[idx, 0].plot(freq_shift, amplitude)\n ax[idx, 0].set_ylabel(data.columns[2])\n ax[idx, 0].set_title(file)\n\n ax[idx, 1].plot(freq_shift, phase)\n ax[idx, 1].set_ylabel(data.columns[3])\n ax[idx, 1].set_title(file)\n \n if fit:\n fano1 = sft.fit_fano(freq_shift, amplitude)\n q_factor = (params[\"Center Frequency (Hz)\"] + fano1.params[\"center\"].value) / (fano1.params[\"sigma\"].value)\n q_factor_err = q_factor * np.sqrt((fano1.params[\"center\"].stderr/fano1.params[\"center\"].value)**2 + (fano1.params[\"sigma\"].stderr/fano1.params[\"sigma\"].value)**2)\n ax[idx, 0].plot(freq_shift, fano1.best_fit, label=\"Q={:.0f}$\\pm{:.0f}$\".format(q_factor, q_factor_err))\n ax[idx, 0].legend()\n fano2 = sft.fit_fano(freq_shift, phase, linear_offset=True)\n ax[idx, 1].plot(freq_shift, fano2.best_fit)\n print(\"chi-square ({}) = {:.2e}\".format(file, fano1.chisqr))\n\nfig.tight_layout()\nfig.text(0.5, 0.02, data.columns[1], ha='center', va='center')", "_____no_output_____" ], [ "omega_0 = 1\nomega = np.linspace(0, 2, 1000)\nQ = 1\n\nratio = omega / omega_0\n\nphi = np.arctan(-ratio / (Q * (1 - ratio**2)))\n\nfid, ax = plt.subplots()\nax.plot(ratio, phi)", "_____no_output_____" ] ], [ [ "# Calibration from Thermal Noise density\n\nFrom Atomic Force Microscopy, Second Edition by Bert Voigtländer\n\nSection 11.6.5 Experimental Determination of the Sensitivity and Spring Constant in AFM Without Tip-Sample Contact\n\nEq. 11.28 and 11.26", "_____no_output_____" ] ], [ [ "%matplotlib widget\nfile = \"SignalAnalyzer_Spectrum001\"\nparams, data = sio.read_dat(AFM_FOLDER4 + file)\n\ncalibration_params = sft.find_afm_calibration_parameters(data, frequency_range=[40000, 48000], Q=1000, f_0_guess=44000)\nfig, ax = plt.subplots()\nax.plot(calibration_params[\"Frequency (Hz)\"], calibration_params[\"PSD squared (V**2/Hz)\"])\nax.plot(calibration_params[\"Frequency (Hz)\"], calibration_params[\"PSD squared fit (V**2/Hz)\"])\nprint(\"Calibration (m/V) =\", calibration_params[\"Calibration (m/V)\"])", "_____no_output_____" ], [ "%matplotlib inline\n\nfit = False # Setting to True will take slightly longer due to the fitting protocols\n\nfiles = []\nfor file in os.listdir(\"../../Data/\" + AFM_FOLDER4):\n if file.endswith(\".dat\"):\n files.append(file) \n\n\nfiles = [\"frq-sweep002.dat\"]\n \nfig, ax = plt.subplots(nrows=len(files), ncols=2)\n\nfor idx, file in enumerate(files):\n params, data = sio.read_dat(AFM_FOLDER4 + file)\n freq_shift = data[\"Frequency Shift (Hz)\"]\n amplitude = data[\"Amplitude (m)\"]\n phase = data[\"Phase (deg)\"]\n \n if len(files) == 1:\n ax[0].plot(freq_shift, amplitude)\n ax[0].set_ylabel(data.columns[2])\n\n ax[1].plot(freq_shift, phase)\n ax[1].set_ylabel(data.columns[3])\n else:\n ax[idx, 0].plot(freq_shift, amplitude)\n ax[idx, 0].set_ylabel(data.columns[2])\n ax[idx, 0].set_title(file)\n\n ax[idx, 1].plot(freq_shift, phase)\n ax[idx, 1].set_ylabel(data.columns[3])\n ax[idx, 1].set_title(file)\n\n if fit:\n fano1 = sft.fit_fano(freq_shift, amplitude)\n #q_factor = (params[\"Center Frequency (Hz)\"] + fano1.params[\"center\"].value) / (fano1.params[\"sigma\"].value)\n #q_factor_err = q_factor * np.sqrt((fano1.params[\"center\"].stderr/fano1.params[\"center\"].value)**2 + (fano1.params[\"sigma\"].stderr/fano1.params[\"sigma\"].value)**2)\n ax[idx, 0].plot(freq_shift, fano1.best_fit)\n ax[idx, 0].legend()\n fano2 = sft.fit_fano(freq_shift, phase, linear_offset=True)\n ax[idx, 1].plot(freq_shift, fano2.best_fit)\n print(\"chi-square ({}) = {:.2e}\".format(file, fano1.chisqr))\n\nfig.tight_layout()\nfig.text(0.5, 0.02, data.columns[1], ha='center', va='center')", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import sklearn\nimport pickle\nfrom PIL import Image, ImageOps\nimport numpy as np\nimport matplotlib.pyplot as plt", "_____no_output_____" ], [ "#cargar de un pickle\n\nwith open('/home/jupyter/Pickles/Descriptores/train.pickle', 'rb') as train:\n train_list = pickle.load(train)\n\n\nwith open('/home/jupyter/Pickles/Descriptores/validation.pickle', 'rb') as validation:\n validation_list = pickle.load(validation)", "_____no_output_____" ], [ "with open('/home/jupyter/Pickles/Imagenes/images_train.pickle', 'rb') as trains:\n images_train_list = pickle.load(trains)\n\n\nwith open('/home/jupyter/Pickles/Imagenes/images_validation.pickle', 'rb') as validations:\n images_validation_list = pickle.load(validations)", "_____no_output_____" ], [ "#cerca con todas las imagenes\nv_val = np.reshape(validation_list[70], (1,4096))\nv_train = np.reshape(train_list, (1194,4096))", "_____no_output_____" ], [ "v_train = sklearn.preprocessing.normalize(train_list, norm='l2', axis=1, copy=True, return_norm=False)\nv_val = sklearn.preprocessing.normalize(v_val, norm='l2', axis=1, copy=True, return_norm=False)", "_____no_output_____" ], [ "train_list_t = v_train.transpose()", "_____no_output_____" ], [ "res= np.matmul(v_val, train_list_t)", "_____no_output_____" ], [ "ranks = np.argsort(res, axis=1)[:,::-1]\nx_train_img = []\nx_val_img = []", "_____no_output_____" ], [ "x_val_img.append(np.array(images_validation_list[70]))", "_____no_output_____" ], [ "j = 0\nfor j in range (1194):\n x_train_img.append(np.array(images_train_list[j]))", "_____no_output_____" ], [ "h,w = (224, 224)\nnew_image= Image.new('RGB', (h*5,w*1))\n\n# Visualize ranks of the 10 queries\noffset = 10 # it will show results from query #'offset' to #offset+10\nfor q in range(1):\n ranks_q = ranks[q*(offset+1),:]\n for i in range(5):\n new_image.paste(Image.fromarray(x_train_img[ranks_q[i]]), (h*(1+i),w*q))\n # visualize query\n ima_q = Image.fromarray(x_val_img[0])\n ima_q = ImageOps.expand(ima_q, border=15, fill='orange')\n\n new_image.paste(ima_q, (0,w*q))", "_____no_output_____" ], [ "plt.imshow(new_image)\nplt.axis('off')\nplt.show()", "_____no_output_____" ], [ "new_image.save('no_funciona.jpg')", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "<a href=\"https://colab.research.google.com/github/yarusx/cat-vs-dogo/blob/main/cat_vs_dog_0_0_2.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>", "_____no_output_____" ] ], [ [ "import matplotlib.pyplot as plt\nimport numpy as np\nimport os\nimport tensorflow as tf\n\nfrom tensorflow.keras.preprocessing import image_dataset_from_directory", "_____no_output_____" ], [ "_URL = 'https://storage.googleapis.com/mledu-datasets/cats_and_dogs_filtered.zip'\npath_to_zip = tf.keras.utils.get_file('cats_and_dogs.zip', origin=_URL, extract=True)\nPATH = os.path.join(os.path.dirname(path_to_zip), 'cats_and_dogs_filtered')\n\ntrain_dir = os.path.join(PATH, 'train')\nvalidation_dir = os.path.join(PATH, 'validation')\n\nBATCH_SIZE = 32\nIMG_SIZE = (160, 160)\n\ntrain_dataset = image_dataset_from_directory(train_dir,\n shuffle=True,\n batch_size=BATCH_SIZE,\n image_size=IMG_SIZE)", "Downloading data from https://storage.googleapis.com/mledu-datasets/cats_and_dogs_filtered.zip\n68608000/68606236 [==============================] - 1s 0us/step\nFound 2000 files belonging to 2 classes.\n" ], [ "validation_dataset = image_dataset_from_directory(validation_dir,\n shuffle=True,\n batch_size=BATCH_SIZE,\n image_size=IMG_SIZE)", "Found 1000 files belonging to 2 classes.\n" ], [ "class_names = train_dataset.class_names\n\n# plt.figure(figsize=(10, 10))\n# for images, labels in train_dataset.take(1):\n# for i in range(9):\n# ax = plt.subplot(3, 3, i + 1)\n# plt.imshow(images[i].numpy().astype(\"uint8\"))\n# plt.title(class_names[labels[i]])\n# plt.axis(\"off\")", "_____no_output_____" ], [ "val_batches = tf.data.experimental.cardinality(validation_dataset)\ntest_dataset = validation_dataset.take(val_batches // 5)\nvalidation_dataset = validation_dataset.skip(val_batches // 5)", "_____no_output_____" ], [ "AUTOTUNE = tf.data.experimental.AUTOTUNE\n\ntrain_dataset = train_dataset.prefetch(buffer_size=AUTOTUNE)\nvalidation_dataset = validation_dataset.prefetch(buffer_size=AUTOTUNE)\ntest_dataset = test_dataset.prefetch(buffer_size=AUTOTUNE)", "_____no_output_____" ], [ "data_augmentation = tf.keras.Sequential([\n tf.keras.layers.experimental.preprocessing.RandomFlip('horizontal'),\n tf.keras.layers.experimental.preprocessing.RandomRotation(0.2),\n])", "_____no_output_____" ], [ "# for image, _ in train_dataset.take(1):\n# plt.figure(figsize=(10, 10))\n# first_image = image[0]\n# for i in range(9):\n# ax = plt.subplot(3, 3, i + 1)\n# augmented_image = data_augmentation(tf.expand_dims(first_image, 0))\n# plt.imshow(augmented_image[0] / 255)\n# plt.axis('off')", "_____no_output_____" ], [ "rescale = tf.keras.layers.experimental.preprocessing.Rescaling(1./127.5, offset= -1)", "_____no_output_____" ], [ "# Create the base model from the pre-trained model MobileNet V2\nIMG_SHAPE = IMG_SIZE + (3,)\nbase_model = tf.keras.applications.MobileNetV2(input_shape=IMG_SHAPE,\n include_top=False,\n weights='imagenet')", "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/mobilenet_v2/mobilenet_v2_weights_tf_dim_ordering_tf_kernels_1.0_160_no_top.h5\n9412608/9406464 [==============================] - 0s 0us/step\n" ], [ "image_batch, label_batch = next(iter(train_dataset))\nfeature_batch = base_model(image_batch)\nprint(feature_batch.shape)", "(32, 5, 5, 1280)\n" ], [ "base_model.trainable = False", "_____no_output_____" ], [ "global_average_layer = tf.keras.layers.GlobalAveragePooling2D()\nfeature_batch_average = global_average_layer(feature_batch)\nprint(feature_batch_average.shape)", "(32, 1280)\n" ], [ "prediction_layer = tf.keras.layers.Dense(1)\nprediction_batch = prediction_layer(feature_batch_average)\nprint(prediction_batch.shape)", "(32, 1)\n" ], [ "inputs = tf.keras.Input(shape=(160, 160, 3))\nx = data_augmentation(inputs)\nx = rescale(x)\nx = base_model(x, training=False)\nx = global_average_layer(x)\nx = tf.keras.layers.Dropout(0.4)(x)\noutputs = prediction_layer(x)\nmodel = tf.keras.Model(inputs, outputs)", "_____no_output_____" ], [ "base_learning_rate = 0.0001\nmodel.compile(optimizer=tf.keras.optimizers.Adam(lr=base_learning_rate),\n loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),\n metrics=['accuracy'])", "_____no_output_____" ], [ "loss0, accuracy0 = model.evaluate(validation_dataset)\nprint(\"initial loss: {:.2f}\".format(loss0))\nprint(\"initial accuracy: {:.2f}\".format(accuracy0))", "26/26 [==============================] - 1s 45ms/step - loss: 0.6649 - accuracy: 0.5408\ninitial loss: 0.66\ninitial accuracy: 0.54\n" ], [ "initial_epochs = 1\nhistory = model.fit(train_dataset,\n epochs=initial_epochs,\n validation_data=validation_dataset)\nval_acc = history.history['val_accuracy']", "63/63 [==============================] - 5s 77ms/step - loss: 0.6902 - accuracy: 0.6105 - val_loss: 0.4842 - val_accuracy: 0.6856\n" ], [ "while np.mean(val_acc)*100 < 98.5:\n initial_epochs = 3\n history = model.fit(train_dataset,\n epochs=initial_epochs,\n validation_data=validation_dataset)\n \n val_acc = history.history['val_accuracy']\n \ntry:\n !mkdir -p saved_model\nexcept:\n pass\n\nmodel.save('saved_model/dvc/')\n\n!zip -r dvc.zip saved_model/dvc/\n\nfrom google.colab import files\nfiles.download(\"dvc.zip\")", "Epoch 1/3\n63/63 [==============================] - 5s 75ms/step - loss: 0.1765 - accuracy: 0.9230 - val_loss: 0.0850 - val_accuracy: 0.9752\nEpoch 2/3\n63/63 [==============================] - 5s 74ms/step - loss: 0.1732 - accuracy: 0.9235 - val_loss: 0.0831 - val_accuracy: 0.9765\nEpoch 3/3\n63/63 [==============================] - 5s 73ms/step - loss: 0.1832 - accuracy: 0.9210 - val_loss: 0.0858 - val_accuracy: 0.9715\nEpoch 1/3\n63/63 [==============================] - 5s 77ms/step - loss: 0.1874 - accuracy: 0.9150 - val_loss: 0.0790 - val_accuracy: 0.9790\nEpoch 2/3\n63/63 [==============================] - 5s 75ms/step - loss: 0.1710 - accuracy: 0.9310 - val_loss: 0.0793 - val_accuracy: 0.9777\nEpoch 3/3\n63/63 [==============================] - 5s 73ms/step - loss: 0.1723 - accuracy: 0.9260 - val_loss: 0.0767 - val_accuracy: 0.9765\nEpoch 1/3\n63/63 [==============================] - 5s 73ms/step - loss: 0.1757 - accuracy: 0.9245 - val_loss: 0.0798 - val_accuracy: 0.9765\nEpoch 2/3\n63/63 [==============================] - 5s 73ms/step - loss: 0.1678 - accuracy: 0.9280 - val_loss: 0.0710 - val_accuracy: 0.9790\nEpoch 3/3\n63/63 [==============================] - 5s 72ms/step - loss: 0.1597 - accuracy: 0.9325 - val_loss: 0.0757 - val_accuracy: 0.9777\nEpoch 1/3\n63/63 [==============================] - 5s 72ms/step - loss: 0.1569 - accuracy: 0.9320 - val_loss: 0.0688 - val_accuracy: 0.9814\nEpoch 2/3\n63/63 [==============================] - 5s 72ms/step - loss: 0.1576 - accuracy: 0.9350 - val_loss: 0.0723 - val_accuracy: 0.9802\nEpoch 3/3\n63/63 [==============================] - 5s 73ms/step - loss: 0.1667 - accuracy: 0.9315 - val_loss: 0.0694 - val_accuracy: 0.9802\nEpoch 1/3\n63/63 [==============================] - 5s 73ms/step - loss: 0.1567 - accuracy: 0.9305 - val_loss: 0.0655 - val_accuracy: 0.9827\nEpoch 2/3\n63/63 [==============================] - 5s 73ms/step - loss: 0.1526 - accuracy: 0.9330 - val_loss: 0.0641 - val_accuracy: 0.9839\nEpoch 3/3\n63/63 [==============================] - 5s 73ms/step - loss: 0.1564 - accuracy: 0.9335 - val_loss: 0.0676 - val_accuracy: 0.9802\nEpoch 1/3\n63/63 [==============================] - 5s 74ms/step - loss: 0.1537 - accuracy: 0.9300 - val_loss: 0.0669 - val_accuracy: 0.9802\nEpoch 2/3\n63/63 [==============================] - 5s 78ms/step - loss: 0.1556 - accuracy: 0.9285 - val_loss: 0.0694 - val_accuracy: 0.9802\nEpoch 3/3\n63/63 [==============================] - 5s 74ms/step - loss: 0.1600 - accuracy: 0.9300 - val_loss: 0.0691 - val_accuracy: 0.9802\nEpoch 1/3\n63/63 [==============================] - 5s 74ms/step - loss: 0.1506 - accuracy: 0.9335 - val_loss: 0.0622 - val_accuracy: 0.9827\nEpoch 2/3\n63/63 [==============================] - 5s 74ms/step - loss: 0.1511 - accuracy: 0.9305 - val_loss: 0.0622 - val_accuracy: 0.9827\nEpoch 3/3\n63/63 [==============================] - 5s 78ms/step - loss: 0.1444 - accuracy: 0.9420 - val_loss: 0.0586 - val_accuracy: 0.9827\nEpoch 1/3\n63/63 [==============================] - 5s 83ms/step - loss: 0.1524 - accuracy: 0.9385 - val_loss: 0.0641 - val_accuracy: 0.9814\nEpoch 2/3\n63/63 [==============================] - 5s 77ms/step - loss: 0.1547 - accuracy: 0.9335 - val_loss: 0.0584 - val_accuracy: 0.9839\nEpoch 3/3\n63/63 [==============================] - 5s 79ms/step - loss: 0.1479 - accuracy: 0.9405 - val_loss: 0.0653 - val_accuracy: 0.9802\nEpoch 1/3\n63/63 [==============================] - 5s 76ms/step - loss: 0.1475 - accuracy: 0.9415 - val_loss: 0.0600 - val_accuracy: 0.9814\nEpoch 2/3\n63/63 [==============================] - 5s 78ms/step - loss: 0.1344 - accuracy: 0.9510 - val_loss: 0.0563 - val_accuracy: 0.9864\nEpoch 3/3\n63/63 [==============================] - 5s 79ms/step - loss: 0.1363 - accuracy: 0.9450 - val_loss: 0.0627 - val_accuracy: 0.9814\nEpoch 1/3\n63/63 [==============================] - 5s 79ms/step - loss: 0.1341 - accuracy: 0.9450 - val_loss: 0.0570 - val_accuracy: 0.9839\nEpoch 2/3\n63/63 [==============================] - 5s 76ms/step - loss: 0.1428 - accuracy: 0.9345 - val_loss: 0.0603 - val_accuracy: 0.9802\nEpoch 3/3\n63/63 [==============================] - 5s 78ms/step - loss: 0.1259 - accuracy: 0.9470 - val_loss: 0.0565 - val_accuracy: 0.9827\nEpoch 1/3\n63/63 [==============================] - 5s 79ms/step - loss: 0.1578 - accuracy: 0.9295 - val_loss: 0.0596 - val_accuracy: 0.9814\nEpoch 2/3\n63/63 [==============================] - 5s 78ms/step - loss: 0.1392 - accuracy: 0.9370 - val_loss: 0.0549 - val_accuracy: 0.9827\nEpoch 3/3\n63/63 [==============================] - 5s 77ms/step - loss: 0.1464 - accuracy: 0.9365 - val_loss: 0.0579 - val_accuracy: 0.9839\nEpoch 1/3\n63/63 [==============================] - 5s 76ms/step - loss: 0.1479 - accuracy: 0.9385 - val_loss: 0.0597 - val_accuracy: 0.9814\nEpoch 2/3\n63/63 [==============================] - 5s 77ms/step - loss: 0.1339 - accuracy: 0.9395 - val_loss: 0.0529 - val_accuracy: 0.9839\nEpoch 3/3\n63/63 [==============================] - 5s 78ms/step - loss: 0.1478 - accuracy: 0.9400 - val_loss: 0.0601 - val_accuracy: 0.9790\nEpoch 1/3\n63/63 [==============================] - 5s 80ms/step - loss: 0.1355 - accuracy: 0.9450 - val_loss: 0.0556 - val_accuracy: 0.9814\nEpoch 2/3\n63/63 [==============================] - 5s 79ms/step - loss: 0.1335 - accuracy: 0.9425 - val_loss: 0.0576 - val_accuracy: 0.9814\nEpoch 3/3\n63/63 [==============================] - 5s 75ms/step - loss: 0.1315 - accuracy: 0.9435 - val_loss: 0.0591 - val_accuracy: 0.9827\nEpoch 1/3\n63/63 [==============================] - 5s 77ms/step - loss: 0.1465 - accuracy: 0.9415 - val_loss: 0.0561 - val_accuracy: 0.9839\nEpoch 2/3\n63/63 [==============================] - 5s 78ms/step - loss: 0.1251 - accuracy: 0.9520 - val_loss: 0.0584 - val_accuracy: 0.9790\nEpoch 3/3\n63/63 [==============================] - 5s 76ms/step - loss: 0.1364 - accuracy: 0.9460 - val_loss: 0.0501 - val_accuracy: 0.9851\nEpoch 1/3\n63/63 [==============================] - 5s 74ms/step - loss: 0.1365 - accuracy: 0.9360 - val_loss: 0.0514 - val_accuracy: 0.9839\nEpoch 2/3\n63/63 [==============================] - 5s 76ms/step - loss: 0.1302 - accuracy: 0.9435 - val_loss: 0.0538 - val_accuracy: 0.9839\nEpoch 3/3\n63/63 [==============================] - 5s 75ms/step - loss: 0.1406 - accuracy: 0.9385 - val_loss: 0.0551 - val_accuracy: 0.9839\nEpoch 1/3\n63/63 [==============================] - 5s 73ms/step - loss: 0.1400 - accuracy: 0.9360 - val_loss: 0.0551 - val_accuracy: 0.9827\nEpoch 2/3\n63/63 [==============================] - 5s 79ms/step - loss: 0.1442 - accuracy: 0.9380 - val_loss: 0.0517 - val_accuracy: 0.9839\nEpoch 3/3\n63/63 [==============================] - 5s 76ms/step - loss: 0.1339 - accuracy: 0.9465 - val_loss: 0.0550 - val_accuracy: 0.9839\nEpoch 1/3\n63/63 [==============================] - 5s 79ms/step - loss: 0.1499 - accuracy: 0.9345 - val_loss: 0.0568 - val_accuracy: 0.9814\nEpoch 2/3\n63/63 [==============================] - 5s 76ms/step - loss: 0.1401 - accuracy: 0.9435 - val_loss: 0.0527 - val_accuracy: 0.9839\nEpoch 3/3\n63/63 [==============================] - 5s 81ms/step - loss: 0.1285 - accuracy: 0.9410 - val_loss: 0.0575 - val_accuracy: 0.9827\nEpoch 1/3\n63/63 [==============================] - 5s 78ms/step - loss: 0.1457 - accuracy: 0.9355 - val_loss: 0.0564 - val_accuracy: 0.9839\nEpoch 2/3\n63/63 [==============================] - 5s 76ms/step - loss: 0.1210 - accuracy: 0.9520 - val_loss: 0.0491 - val_accuracy: 0.9876\nEpoch 3/3\n63/63 [==============================] - 5s 79ms/step - loss: 0.1300 - accuracy: 0.9450 - val_loss: 0.0513 - val_accuracy: 0.9851\nINFO:tensorflow:Assets written to: saved_model/dvc/assets\nupdating: saved_model/dvc/ (stored 0%)\nupdating: saved_model/dvc/assets/ (stored 0%)\nupdating: saved_model/dvc/variables/ (stored 0%)\nupdating: saved_model/dvc/variables/variables.index (deflated 75%)\nupdating: saved_model/dvc/variables/variables.data-00000-of-00001 (deflated 8%)\nupdating: saved_model/dvc/saved_model.pb (deflated 92%)\n" ], [ "from google.colab import drive\ndrive.mount('/content/drive')\n!unzip -q /content/drive/MyDrive/dvc.zip\ndvc = tf.keras.models.load_model('/content/saved_model/dvc')", "Mounted at /content/drive\nreplace saved_model/dvc/variables/variables.index? [y]es, [n]o, [A]ll, [N]one, [r]ename: A\n" ], [ "try:\n !mkdir -p saved_model\nexcept:\n pass\n\nmodel.save('saved_model/dvc/')\n\n!zip -r dvc.zip saved_model/dvc/\n\nfrom google.colab import files\nfiles.download(\"dvc.zip\")", "_____no_output_____" ], [ "from keras.preprocessing.image import load_img, img_to_array\n\n# load and prepare the image\ndef load_image(filename):\n\t# load the image\n\timg = load_img(filename, target_size=(160, 160))\n\t# convert to array\n\timg = img_to_array(img)\n\t# reshape into a single sample with 3 channels\n\timg = img.reshape(1, 160, 160, 3)\n\treturn img\n\nimg = load_image('/content/drive/MyDrive/dayana_1.JPG')", "_____no_output_____" ], [ "categories = [\"Cat\", \"Dog\"]", "_____no_output_____" ], [ "prediction = dvc_model.predict(img)\nprediction = tf.nn.sigmoid(prediction)\nprint(prediction)\nplt.figure()\nplt.imshow(img[0]/255)\nplt.title(categories[int(np.round_(prediction))])", "_____no_output_____" ], [ "loss0, accuracy0 = model.evaluate(validation_dataset)", "26/26 [==============================] - 1s 40ms/step - loss: 0.0848 - accuracy: 0.9765\n" ], [ "# #Retrieve a batch of images from the test set\n# image_batch, label_batch = test_dataset.as_numpy_iterator().next()\n# predictions = model.predict_on_batch(image_batch).flatten()\n\n# # Apply a sigmoid since our model returns logits\n# predictions = tf.nn.sigmoid(predictions)\n# predictions = tf.where(predictions < 0.5, 0, 1)\n\n# print('Predictions:\\n', predictions.numpy())\n# print('Labels:\\n', label_batch)\n\n# plt.figure(figsize=(10, 10))\n# for i in range(9):\n# ax = plt.subplot(3, 3, i + 1)\n# plt.imshow(image_batch[i].astype(\"uint8\"))\n# plt.title(class_names[predictions[i]])\n# plt.axis(\"off\")", "_____no_output_____" ] ] ]

[ "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# some_file.py\nimport sys\n\n# insert at 1, 0 is the script path (or '' in REPL)\nsys.path.insert(1, \"/Users/dhruvbalwada/work_root/sogos/\")", "_____no_output_____" ], [ "import os\nfrom numpy import *\nimport pandas as pd\nimport xarray as xr\nimport matplotlib as mpl\nimport matplotlib.pyplot as plt\nimport numpy as np", "_____no_output_____" ], [ "from xgcm import Grid\nfrom xgcm.autogenerate import generate_grid_ds", "_____no_output_____" ], [ "import sogos.download_product as dlp\nimport sogos.load_product as ldp\nimport sogos.time_tools as tt\nimport sogos.geo_tools as gt\nimport sogos.download_file as df", "_____no_output_____" ], [ "import gsw\nimport cmocean as cmocean", "_____no_output_____" ] ], [ [ "# Download Latest Data", "_____no_output_____" ] ], [ [ "data_dir = \"/Users/dhruvbalwada/work_root/sogos/data/raw/climatology/\"", "_____no_output_____" ] ], [ [ "FTP ADDRESS: ftp://kakapo.ucsd.edu/pub/gilson/argo_climatology/", "_____no_output_____" ], [ "Data prior to 2017 (till Dec 2016) is in a single file ", "_____no_output_____" ] ], [ [ "# download the big climatology files\nwget.download(\n \"ftp://kakapo.ucsd.edu/pub/gilson/argo_climatology/RG_ArgoClim_Salinity_2017.nc.gz\",\n data_dir,\n)\nwget.download(\n \"ftp://kakapo.ucsd.edu/pub/gilson/argo_climatology/RG_ArgoClim_Temperature_2017.nc.gz\",\n data_dir,\n)", "_____no_output_____" ], [ "from ftplib import FTP", "_____no_output_____" ], [ "ftp_address = \"ftp://kakapo.ucsd.edu/pub/gilson/argo_climatology/RG_ArgoClim_2019\"", "_____no_output_____" ], [ "url_root = \"/pub/gilson/argo_climatology/\"\nftp_root = \"kakapo.ucsd.edu\"", "_____no_output_____" ], [ "ftp = FTP(ftp_root)\nftp.login()\nftp.cwd(url_root)\ncontents = ftp.nlst(\"RG_ArgoClim_2017*\")", "_____no_output_____" ], [ "contents = ftp.nlst(\"RG_ArgoClim_201*\")", "_____no_output_____" ], [ "for i in contents:\n print(\"Downloading\" + i)\n wget.download(\"ftp://kakapo.ucsd.edu/pub/gilson/argo_climatology/\" + i, data_dir)", "DownloadingRG_ArgoClim_201701.nc.gz\nDownloadingRG_ArgoClim_201702.nc.gz\nDownloadingRG_ArgoClim_201703.nc.gz\nDownloadingRG_ArgoClim_201704.nc.gz\nDownloadingRG_ArgoClim_201705.nc.gz\nDownloadingRG_ArgoClim_201706.nc.gz\nDownloadingRG_ArgoClim_201707.nc.gz\nDownloadingRG_ArgoClim_201708.nc.gz\nDownloadingRG_ArgoClim_201709.nc.gz\nDownloadingRG_ArgoClim_201710.nc.gz\nDownloadingRG_ArgoClim_201711.nc.gz\nDownloadingRG_ArgoClim_201712.nc.gz\nDownloadingRG_ArgoClim_201801.nc.gz\nDownloadingRG_ArgoClim_201802.nc.gz\nDownloadingRG_ArgoClim_201803.nc.gz\nDownloadingRG_ArgoClim_201804.nc.gz\nDownloadingRG_ArgoClim_201805.nc.gz\nDownloadingRG_ArgoClim_201806.nc.gz\nDownloadingRG_ArgoClim_201807.nc.gz\nDownloadingRG_ArgoClim_201808.nc.gz\nDownloadingRG_ArgoClim_201809.nc.gz\nDownloadingRG_ArgoClim_201810.nc.gz\nDownloadingRG_ArgoClim_201811.nc.gz\nDownloadingRG_ArgoClim_201812.nc.gz\nDownloadingRG_ArgoClim_201901.nc.gz\nDownloadingRG_ArgoClim_201902.nc.gz\nDownloadingRG_ArgoClim_201903.nc.gz\nDownloadingRG_ArgoClim_201904.nc.gz\nDownloadingRG_ArgoClim_201905.nc.gz\nDownloadingRG_ArgoClim_201906.nc.gz\nDownloadingRG_ArgoClim_201907.nc.gz\nDownloadingRG_ArgoClim_201908.nc.gz\nDownloadingRG_ArgoClim_201909.nc.gz\n" ] ], [ [ "## Load some data ", "_____no_output_____" ] ], [ [ "Tclim = xr.open_dataset(data_dir + \"RG_ArgoClim_Temperature_2017.nc\", decode_times=False)\nSclim = xr.open_dataset(data_dir + \"RG_ArgoClim_Salinity_2017.nc\", decode_times=False)", "_____no_output_____" ], [ "Climextra = xr.open_mfdataset(data_dir+ 'RG_ArgoClim_201*', decode_times=False)", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/ipykernel_launcher.py:1: FutureWarning: In xarray version 0.15 the default behaviour of `open_mfdataset`\nwill change. To retain the existing behavior, pass\ncombine='nested'. To use future default behavior, pass\ncombine='by_coords'. See\nhttp://xarray.pydata.org/en/stable/combining.html#combining-multi\n\n \"\"\"Entry point for launching an IPython kernel.\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/xarray/backends/api.py:931: FutureWarning: The datasets supplied have global dimension coordinates. You may want\nto use the new `combine_by_coords` function (or the\n`combine='by_coords'` option to `open_mfdataset`) to order the datasets\nbefore concatenation. Alternatively, to continue concatenating based\non the order the datasets are supplied in future, please use the new\n`combine_nested` function (or the `combine='nested'` option to\nopen_mfdataset).\n from_openmfds=True,\n" ], [ "RG_clim = xr.merge([Tclim, Sclim, Climextra])", "_____no_output_____" ], [ "# Calendar type was missing, and giving errors in decoding time\nRG_clim.TIME.attrs['calendar'] = '360_day'\nRG_clim = xr.decode_cf(RG_clim)", "_____no_output_____" ], [ "## Add density and other things \nSA = xr.apply_ufunc(gsw.SA_from_SP, RG_clim.ARGO_SALINITY_MEAN+RG_clim.ARGO_SALINITY_ANOMALY, RG_clim.PRESSURE , \n RG_clim.LONGITUDE, RG_clim.LATITUDE,\n dask='parallelized', output_dtypes=[float,]).rename('SA')\nCT = xr.apply_ufunc(gsw.CT_from_t, SA, RG_clim.ARGO_TEMPERATURE_MEAN+RG_clim.ARGO_SALINITY_ANOMALY, RG_clim.PRESSURE, \n dask='parallelized', output_dtypes=[float,]).rename('CT')\nSIGMA0 = xr.apply_ufunc(gsw.sigma0, SA, CT, dask='parallelized', output_dtypes=[float,]).rename('SIGMA0')", "_____no_output_____" ], [ "RG_clim = xr.merge([RG_clim, SIGMA0])", "_____no_output_____" ], [ "T_region = RG_clim.ARGO_TEMPERATURE_ANOMALY.groupby('TIME.season').mean() + RG_clim.ARGO_TEMPERATURE_MEAN\nS_region = RG_clim.ARGO_SALINITY_ANOMALY.groupby('TIME.season').mean() + RG_clim.ARGO_SALINITY_MEAN\nrho_region = RG_clim.SIGMA0.groupby('TIME.season').mean()", "_____no_output_____" ], [ "plt.figure(figsize=(18,3))\nplt.subplot(141)\nT_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='DJF').plot.contourf(levels=11, vmin=-9, vmax=9);\n\nplt.subplot(142)\nT_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='MAM').plot.contourf(levels=11, vmin=-9, vmax=9);\n\nplt.subplot(143)\nT_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='JJA').plot.contourf(levels=11, vmin=-9, vmax=9);\n\nplt.subplot(144)\nT_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='SON').plot.contourf(levels=11, vmin=-9, vmax=9);", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n" ], [ "plt.figure(figsize=(18,3))\nplt.subplot(141)\nS_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='DJF').plot.contourf(levels=11, vmin=33.7, vmax=34.2)\n\nplt.subplot(142)\nS_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='MAM').plot.contourf(levels=11, vmin=33.7, vmax=34.2)\n\nplt.subplot(143)\nS_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='JJA').plot.contourf(levels=11, vmin=33.7, vmax=34.2)\n\nplt.subplot(144)\nS_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='SON').plot.contourf(levels=11, vmin=33.7, vmax=34.2)\nplt.tight_layoutout()", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n" ], [ "plt.figure(figsize=(18,3))\nplt.subplot(141)\nrho_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='DJF').plot.contourf(levels=11)\n\nplt.subplot(142)\nrho_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='MAM').plot.contourf(levels=11)\n\nplt.subplot(143)\nrho_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='JJA').plot.contourf(levels=11)\n\nplt.subplot(144)\nrho_region.sel(LATITUDE=slice(-65,-45), LONGITUDE=slice(20,60)).isel(PRESSURE=0).sel(season='SON').plot.contourf(levels=11)", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n" ] ], [ [ "### Some Climatological Mean Sections ", "_____no_output_____" ] ], [ [ "dens_section30 = RG_clim.SIGMA0.sel(LONGITUDE=30, method='nearest').sel(LATITUDE=slice(-70,-40)).load()\ndens_section40 = RG_clim.SIGMA0.sel(LONGITUDE=40, method='nearest').sel(LATITUDE=slice(-70,-40)).load()", "_____no_output_____" ], [ "glider = {\"start_month\": 4.99, \"end_month\":7.8, \"start_lat\": -51.5, \"end_lat\": -53, \"max_depth\": 1000}", "_____no_output_____" ], [ "plt.figure(figsize=(15,4))\n\nplt.subplot(121)\nRG_clim.ARGO_TEMPERATURE_MEAN.sel(LONGITUDE=30, method='nearest').sel(LATITUDE=slice(-70,-40)\n ).plot.contourf(vmin=-10, levels=24)\nRG_clim.ARGO_TEMPERATURE_MEAN.sel(LONGITUDE=30, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contour(linestyles='-.',levels=[1])\ndens_section.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C0',\n linestyles='dashed', linewidths=4)\nplt.plot([glider['start_lat'], glider['start_lat']], [4, glider['max_depth']], color='C2', alpha=0.5)\nplt.plot([glider['end_lat'], glider['end_lat']], [4, glider['max_depth']], color='C2', alpha=0.5)\nplt.gca().invert_yaxis()\n\nplt.subplot(122)\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=30, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contourf(vmin=33.77, vmax=35.21, levels=24, cmap=cmocean.cm.haline)\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=30, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contour(levels=[34.2], linestyles='dashdot')\ndens_section.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C1', \n linestyles='dashed', linewidths=4)\nplt.plot([glider['start_lat'], glider['start_lat']], [4, 1900], color='C3', alpha=0.5)\n\nplt.gca().invert_yaxis()\n\nplt.tight_layout()\n\n#plt.savefig('../figures/clim_TS_30E.png')", "_____no_output_____" ], [ "plt.figure(figsize=(15,4))\n\nplt.subplot(121)\nRG_clim.ARGO_TEMPERATURE_MEAN.sel(LONGITUDE=40, method='nearest').sel(LATITUDE=slice(-70,-40)\n ).plot.contourf(vmin=-10,levels=24)\nRG_clim.ARGO_TEMPERATURE_MEAN.sel(LONGITUDE=40, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contour(linestyles='-.',levels=[1])\ndens_section40.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section40.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C0',\n linestyles='dashed', linewidths=4)\nplt.gca().invert_yaxis()\n\nplt.subplot(122)\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=40, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contourf(vmin=33.77, vmax=35.21,levels=24, cmap=cmocean.cm.haline)\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=40, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contour(levels=[34.2], linestyles='dashdot')\ndens_section40.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section40.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C1', \n linestyles='dashed', linewidths=4)\n\nplt.gca().invert_yaxis()\n\nplt.tight_layout()\nplt.savefig('../figures/clim_TS_40E.png')", "_____no_output_____" ], [ "plt.figure(figsize=(12,4))\n\nplt.subplot(121)\nRG_clim.ARGO_TEMPERATURE_MEAN.sel(LONGITUDE=40, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contourf(levels=24)\nRG_clim.ARGO_TEMPERATURE_MEAN.sel(LONGITUDE=40, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contour(levels=[1])\nplt.gca().invert_yaxis()\n\nplt.subplot(122)\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=40, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contourf(levels=24)\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=40, method='nearest').sel(LATITUDE=slice(-70,-40)).plot.contour(levels=[34.2])\nplt.gca().invert_yaxis()\n\nplt.tight_layout()", "_____no_output_____" ], [ "# Not much seasonality below 200m\ndens_section.groupby('TIME.month').mean().mean('month').sel(PRESSURE=slice(0,1000)).plot.contourf(cmap='Blues')\ndens_section.groupby('TIME.month').mean().isel(month=0).sel(PRESSURE=slice(0,1000)).plot.contour(levels=[27.2])\ndens_section.groupby('TIME.month').mean().isel(month=3).sel(PRESSURE=slice(0,1000)).plot.contour(levels=[27.2])\ndens_section.groupby('TIME.month').mean().isel(month=6).sel(PRESSURE=slice(0,1000)).plot.contour(levels=[27.2])\ndens_section.groupby('TIME.month').mean().isel(month=9).sel(PRESSURE=slice(0,1000)).plot.contour(levels=[27.2])\n\nplt.gca().invert_yaxis()", "_____no_output_____" ] ], [ [ "## N2\n\n\\begin{equation}\nN^2 = db/dz\n\\end{equation}\n\n$b = -\\frac{g}{\\rho_0} \\rho'$\n\n$b = g(\\alpha \\triangle T - \\beta \\triangle S)$", "_____no_output_____" ] ], [ [ "RG_clim", "_____no_output_____" ], [ "RG_clim = generate_grid_ds(RG_clim, \n {'Z':'PRESSURE', 'X':'LONGITUDE', 'Y':'LATITUDE'})", "_____no_output_____" ], [ "grid = Grid(RG_clim, periodic='X')", "_____no_output_____" ], [ "g = 9.81\nrho0 = 1000", "_____no_output_____" ], [ "dens_clim_monthly = RG_clim.SIGMA0.groupby('TIME.month').mean()", "_____no_output_____" ], [ "dens_clim_monthly", "_____no_output_____" ], [ "N2_clim_monthly = grid.interp(-g/rho0* grid.diff(dens_clim_monthly, 'Z', boundary='extend') / -(grid.diff(RG_clim.PRESSURE, 'Z', boundary='extend')), 'Z', boundary='extend')", "_____no_output_____" ], [ "N2_clim_monthly_SO = N2_clim_monthly.sel(LATITUDE=slice(-70, -30)).load()", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/core.py:119: RuntimeWarning: divide by zero encountered in true_divide\n return func(*args2)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/core.py:119: RuntimeWarning: invalid value encountered in true_divide\n return func(*args2)\n" ], [ "N2_clim_monthly_SO = N2_clim_monthly_SO.rename('N2')", "_____no_output_____" ], [ "plt.figure(figsize=(18,12))\n\nplt.subplot(221)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=1).plot(vmin=-5e-5)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=1).plot.contour(levels=[2e-5])\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=30, method='nearest').sel(\n LATITUDE=slice(-70,-40)).plot.contour(levels=[34.35], linestyles='dashdot')\n\n\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C1', \n linestyles='dashed', linewidths=4)\nplt.title('January')\nplt.gca().invert_yaxis()\n\nplt.subplot(222)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=4).plot(vmin=-5e-5)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=4).plot.contour(levels=[2e-5])\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=30, method='nearest').sel(\n LATITUDE=slice(-70,-40)).plot.contour(levels=[34.35], linestyles='dashdot')\n\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C1', \n linestyles='dashed', linewidths=4)\nplt.title('April')\nplt.gca().invert_yaxis()\n\nplt.subplot(223)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=7).plot(vmin=-5e-5)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=7).plot.contour(levels=[2e-5])\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=30, method='nearest').sel(\n LATITUDE=slice(-70,-40)).plot.contour(levels=[34.35], linestyles='dashdot')\n\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C1', \n linestyles='dashed', linewidths=4)\nplt.title('July')\nplt.gca().invert_yaxis()\n\nplt.subplot(224)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=10).plot(vmin=-5e-5)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=10).plot.contour(levels=[2e-5])\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=30, method='nearest').sel(\n LATITUDE=slice(-70,-40)).plot.contour(levels=[34.35], linestyles='dashdot')\n\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C1', \n linestyles='dashed', linewidths=4)\nplt.title('October')\nplt.gca().invert_yaxis()\n\nplt.savefig('../figures/clim_N2_30E.png')", "_____no_output_____" ], [ "plt.figure(figsize=(8,3))\n\nplt.subplot(121)\n#plt.pcolormesh(N2_clim_monthly_SO.LATITUDE.sel(LATITUDE=slice(-70, -40)), \n# N2_clim_monthly_SO.LATITUDE.sel(=slice(-70, -40)), \nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=4).plot(vmin=-5e-5,\n rasterized=True,add_colorbar=False)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=4).plot.contour(levels=[2e-5])\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=30, method='nearest').sel(\n LATITUDE=slice(-70,-40)).plot.contour(levels=[34.35], linestyles='dashdot')\n\n\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C1', \n linestyles='dashed', linewidths=4)\nplt.title('April')\nplt.gca().invert_yaxis()\nplt.ylim([1500, 0])\nplt.xlabel('Latitude')\nplt.ylabel('Depth (m)')\n\nplt.subplot(122)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=9).plot(vmin=-5e-5,\n rasterized=True,add_colorbar=False)\nN2_clim_monthly_SO.sel(LATITUDE=slice(-70, -40)).sel(LONGITUDE=30, method='nearest').sel(month=9).plot.contour(levels=[2e-5])\nRG_clim.ARGO_SALINITY_MEAN.sel(LONGITUDE=30, method='nearest').sel(\n LATITUDE=slice(-70,-40)).plot.contour(levels=[34.35], linestyles='dashdot')\n\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(colors='k', levels=7)\ndens_section30.groupby('TIME.month').mean().mean('month').plot.contour(levels=[27.189, 27.752], colors='C1', \n linestyles='dashed', linewidths=4)\nplt.title('September')\nplt.gca().invert_yaxis()\nplt.ylim([1500, 0])\nplt.xlabel('Latitude')\nplt.ylabel('Depth (m)')\n\nplt.tight_layout()\nplt.savefig('N2_climatology.pdf')", "_____no_output_____" ], [ "N2.sel(LATITUDE=slice(-60, -40)).sel(LONGITUDE=30, method='nearest').isel(TIME=-1).plot()", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/core.py:119: RuntimeWarning: divide by zero encountered in true_divide\n return func(*args2)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/core.py:119: RuntimeWarning: invalid value encountered in true_divide\n return func(*args2)\n" ] ], [ [ "The apply ufunc way, not working yet. ", "_____no_output_____" ] ], [ [ "CT_clim = CT.groupby('TIME.month').mean()\nSA_clim = SA.groupby('TIME.month').mean()", "_____no_output_____" ], [ "CT_clim_region = CT_clim.sel(LATITUDE=slice(-65,-35), LONGITUDE=slice(20,50)).load()\nSA_clim_region = SA_clim.sel(LATITUDE=slice(-65,-35), LONGITUDE=slice(20,50)).load()", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n" ], [ "(N2, pmid) = xr.apply_ufunc(gsw.Nsquared, SA_clim_region, CT_clim_region, RG_clim.PRESSURE, \n dask='parallelized', \n input_core_dims=[['PRESSURE'],['PRESSURE'],['PRESSURE']],\n output_core_dims=[['PRESSURE'],['PRESSURE']], exclude_dims=set(['PRESSURE']))", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/gsw/stability.py:85: RuntimeWarning: divide by zero encountered in true_divide\n N2 = ((g_local**2) / (specvol_mid * db_to_pa * dp))\n" ] ], [ [ "### Gestrophic Velocities", "_____no_output_____" ] ], [ [ "psi = xr.apply_ufunc(gsw.geo_strf_dyn_height, SA, CT , RG_clim.PRESSURE, \n dask='parallelized', output_dtypes=[float,]).rename('psi')", "_____no_output_____" ], [ "psi", "_____no_output_____" ], [ "vels = xr.apply_ufunc(gsw.geostrophic_velocity, psi, psi.LONGITUDE, psi.LATITUDE, \n dask='parallelized', output_core_dims=[4,4], output_dtypes=[float,]).rename('vels')", "_____no_output_____" ], [ "vels", "_____no_output_____" ] ], [ [ "### Mixed Layer Depth\n\nEnded up going with Holte's climatology for MLD work ", "_____no_output_____" ] ], [ [ "delta_dens = RG_clim.SIGMA0 - RG_clim.SIGMA0.isel(PRESSURE=0)", "_____no_output_____" ], [ "import nc_time_axis", "_____no_output_____" ], [ "RG_clim.SIGMA0.sel(LONGITUDE=30, method='nearest').sel( LATITUDE=slice(-63,-45)).isel(TIME=-1).plot.contourf()\nplt.gca().invert_yaxis()", "_____no_output_____" ], [ "RG_clim.SIGMA0.sel(LONGITUDE=30, LATITUDE=-50, method='nearest').isel(TIME=-1).plot()\nRG_clim.SIGMA0.sel(LONGITUDE=30, LATITUDE=-50, method='nearest').isel(TIME=-7).plot()\nRG_clim.SIGMA0.sel(LONGITUDE=30, LATITUDE=-60, method='nearest').isel(TIME=-1).plot()\nRG_clim.SIGMA0.sel(LONGITUDE=30, LATITUDE=-60, method='nearest').isel(TIME=-7).plot()", "_____no_output_____" ], [ "delta_dens.sel(LONGITUDE=30, LATITUDE=-50, method='nearest').isel(TIME=-1).plot()", "_____no_output_____" ], [ "temp = delta_dens.where(delta_dens>0.03).sel(LONGITUDE=30, LATITUDE=-50, method='nearest').isel(TIME=-1).plot()", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/core.py:119: RuntimeWarning: invalid value encountered in greater\n return func(*args2)\n" ], [ "temp = delta_dens.where(delta_dens>0.03)", "_____no_output_____" ], [ "MLD = temp.PRESSURE.where(temp == temp.min('PRESSURE')).min('PRESSURE')", "_____no_output_____" ], [ "MLD_clim = temp.PRESSURE.where(temp == temp.min('PRESSURE')).min('PRESSURE').groupby('TIME.month').mean()", "_____no_output_____" ], [ "MLD_clim.load()", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/core.py:119: RuntimeWarning: invalid value encountered in greater\n return func(*args2)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/utils.py:29: RuntimeWarning: All-NaN slice encountered\n return func(*args, **kwargs)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/toolz/functoolz.py:488: RuntimeWarning: All-NaN slice encountered\n ret = f(ret)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: divide by zero encountered in true_divide\n x = np.divide(x1, x2, out)\n/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/dask/array/numpy_compat.py:40: RuntimeWarning: invalid value encountered in true_divide\n x = np.divide(x1, x2, out)\n" ], [ "MLD_clim.month", "_____no_output_____" ], [ "plt.figure(figsize=(12,4))\n\nplt.subplot(121)\nMLD_clim.sel(LATITUDE=slice(-75,-25), LONGITUDE=slice(20,90)).sel(month=1).plot(vmin=0, vmax=120)\n\nplt.subplot(122)\nMLD_clim.sel(LATITUDE=slice(-75,-25), LONGITUDE=slice(20,90)).sel(month=7).plot(vmin=0, vmax=120)", "_____no_output_____" ], [ "MLD_clim.max('month').sel(LATITUDE=slice(-75,-25), LONGITUDE=slice(20,380)).plot(vmin=0, vmax=150)", "_____no_output_____" ], [ "deltaH = MLD_clim.max('month') - MLD_clim.min('month')", "/Users/dhruvbalwada/code/miniconda/envs/sogos/lib/python3.7/site-packages/xarray/core/nputils.py:223: RuntimeWarning: All-NaN slice encountered\n result = getattr(npmodule, name)(values, axis=axis, **kwargs)\n" ], [ "deltaH.sel(LATITUDE=slice(-75,-25), LONGITUDE=slice(20,380)).plot(vmin=0, vmax=80)", "_____no_output_____" ], [ "MLD_clim.sel(LATITUDE=-45, LONGITUDE=35, method='nearest').plot(label='45S')\nMLD_clim.sel(LATITUDE=-50, LONGITUDE=35, method='nearest').plot(label='50S')\nMLD_clim.sel(LATITUDE=-55, LONGITUDE=35, method='nearest').plot(label='55S')\nMLD_clim.sel(LATITUDE=-60, LONGITUDE=35, method='nearest').plot(label='60S')\nplt.legend()", "_____no_output_____" ], [ "MLD_clim.sel(LATITUDE=-45, LONGITUDE=45, method='nearest').plot(label='45S')\nMLD_clim.sel(LATITUDE=-50, LONGITUDE=45, method='nearest').plot(label='50S')\nMLD_clim.sel(LATITUDE=-55, LONGITUDE=45, method='nearest').plot(label='55S')\nMLD_clim.sel(LATITUDE=-60, LONGITUDE=45, method='nearest').plot(label='60S')\nplt.legend()", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "## Mecanismo de manivela-biela corredera", "_____no_output_____" ] ], [ [ "import numpy as np\nimport matplotlib.pyplot as plt\nfrom numpy import pi,cos,sin,tan\n%matplotlib inline\n\ndef plotvector(p0,u,color=\"r\"):\n \"\"\"\n Gráfica un vector, dados su punto inicial \n y sus componentes rectangulares\n \"\"\"\n plt.plot([p0[0],p0[0]+u[0]],[p0[1],p0[1]+u[1]], color=color, ls=\"-\", marker='o')\n\nr2, r3 = 3, 8\nfor t2 in np.linspace(0,2*pi,10):\n color = np.array([t2,0,t2])/(2*pi)\n t3 = np.arcsin(-r2*sin(t2)/r3)\n r1 = r2*cos(t2) + r3*cos(t3)\n R2 = np.array([r2*cos(t2),r2*sin(t2)])\n R3 = np.array([r3*cos(t3),r3*sin(t3)])\n plotvector([0,0], R2, color)\n plotvector(R2, R3, color)\n plt.grid('on')\n plt.axis('equal');", "_____no_output_____" ] ] ]

[ "markdown", "code" ]

[ [ "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "#Add needed imports\nimport numpy as np\nimport pandas as pd\nfrom imblearn.over_sampling import SMOTE\nimport seaborn as sns\nfrom sklearn.preprocessing import OrdinalEncoder\nfrom sklearn.dummy import DummyClassifier\nfrom imblearn.over_sampling import SMOTENC\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.preprocessing import StandardScaler\nfrom sklearn.metrics import accuracy_score,confusion_matrix, precision_score, recall_score,f1_score\nfrom sklearn.tree import DecisionTreeClassifier \nfrom sklearn.linear_model import LogisticRegression\nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.neural_network import MLPClassifier\nfrom xgboost import XGBClassifier\nfrom sklearn.neural_network import MLPClassifier\nfrom sklearn.ensemble import GradientBoostingClassifier\nfrom sklearn.model_selection import RepeatedStratifiedKFold, GridSearchCV\nfrom sklearn import svm\nimport shap\nimport os\n\n#Read data\nproccessed_data_path =os.path.join(os.path.pardir,os.path.pardir,'data','processed')\ntrain_path = os.path.join(proccessed_data_path,'dataset10.csv')\ndf = pd.read_csv(train_path)\nlabels=df['Churn']\nx = df.drop(columns=['Churn','Unnamed: 0'],axis = 'columns')\ny=np.ravel(labels)\nx_train,x_test,y_train,y_test = train_test_split(x,y,test_size=0.3)\noversample = SMOTENC(categorical_features=[2])\noversample = SMOTE()\nx_train, y_train = oversample.fit_resample(x_train, y_train)\n\nsc = StandardScaler()\nx_train = sc.fit_transform(x_train)\nx_test = sc.transform(x_test)", "_____no_output_____" ], [ "svm_model = svm.SVC(random_state=0,gamma='auto')\nrf_model=RandomForestClassifier(random_state=0)\ndt_model=DecisionTreeClassifier(random_state=0,criterion='entropy',max_depth = 7,min_samples_leaf=30) \nlr_model= LogisticRegression(random_state=0, max_iter=300)\nmlp_model =MLPClassifier(random_state=0,activation='relu', solver='sgd',learning_rate='adaptive')\nxgb_model = XGBClassifier(random_state=0 ,learning_rate=0.05, max_depth=7,eval_metric='mlogloss',use_label_encoder =False)\ngmb_model= GradientBoostingClassifier(random_state=0,n_estimators=20,learning_rate=0.75,max_features=4,max_depth=5)\nmodel_params = {\n 'svm': {\n 'model': svm_model,\n 'params' : {\n 'C': [15,10],\n 'kernel': ['rbf','linear']\n } \n },\n 'rf': {\n 'model': rf_model,\n 'params' : {\n 'n_estimators': [1,5,10]\n }\n },\n 'dt': {\n 'model': dt_model,\n 'params' : {}\n },\n 'lr' : {\n 'model':lr_model,\n 'params': {\n 'C': [1,5,10]\n }\n },\n 'mlp' : {\n 'model':mlp_model,\n 'params': {}\n },\n 'xg_boost' : {\n 'model':xgb_model,\n 'params': {}\n },\n 'gbm' : {\n 'model':gmb_model,\n 'params': {}\n }\n}", "_____no_output_____" ], [ "scores = []\ncv = RepeatedStratifiedKFold(n_splits=10, n_repeats=3, random_state=100)\nfor model_name, mp in model_params.items():\n clf = GridSearchCV(mp['model'], mp['params'], cv=cv, return_train_score=False)\n clf.fit(x_train,y_train)\n conf_matrix =confusion_matrix(y_test,clf.predict(x_test))\n scores.append({\n 'model': model_name,\n 'best_score': clf.best_score_,\n 'best_params': clf.best_params_,\n 'precision':precision_score(y_test,clf.predict(x_test)),\n 'recall':recall_score(y_test,clf.predict(x_test)),\n 'f1_score':f1_score(y_test,clf.predict(x_test)),\n 'true positives':conf_matrix[0][0],\n 'true negatives':conf_matrix[1][1],\n 'false postives':conf_matrix[0][1],\n 'false negatives':conf_matrix[1][0]\n })\n \ndf = pd.DataFrame(scores,columns=['model','best_score','precision','recall','f1_score','true positives','true negatives','false postives','false negatives','best_params'])\nprint(df)", "Stochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\nStochastic Optimizer: Maximum iterations (200) reached and the optimization hasn't converged yet.\n" ], [ "import shap\nxgb_model = XGBClassifier(random_state=0 ,learning_rate=0.05, max_depth=7,eval_metric='mlogloss',use_label_encoder =False)\nxgb_model.fit(x_train,y_train)\nexplainer = shap.TreeExplainer(xgb_model)\nshap_values = explainer.shap_values(x)\nshap.summary_plot(shap_values, x)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code" ] ]

[ "BSD-3-Clause" ]

[ "BSD-3-Clause" ]

[ "BSD-3-Clause" ]

[ [ [ "<img width=\"10%\" alt=\"Naas\" src=\"https://landen.imgix.net/jtci2pxwjczr/assets/5ice39g4.png?w=160\"/>", "_____no_output_____" ], [ "# FTP - Get file\n<a href=\"https://app.naas.ai/user-redirect/naas/downloader?url=https://raw.githubusercontent.com/jupyter-naas/awesome-notebooks/master/FTP/FTP_Get_file.ipynb\" target=\"_parent\"><img src=\"https://naasai-public.s3.eu-west-3.amazonaws.com/open_in_naas.svg\"/></a>", "_____no_output_____" ], [ "**Tags:** #ftp #file #naas_drivers #operations #snippet #naas", "_____no_output_____" ], [ "**Author:** [Jeremy Ravenel](https://www.linkedin.com/in/ACoAAAJHE7sB5OxuKHuzguZ9L6lfDHqw--cdnJg/)", "_____no_output_____" ], [ "## Input", "_____no_output_____" ], [ "### Import library", "_____no_output_____" ] ], [ [ "from naas_drivers import ftp", "_____no_output_____" ] ], [ [ "### Variables", "_____no_output_____" ] ], [ [ "path = \"/path/to/file/in/ftp\"\nuser = \"my user\"\npasswd = \"my passwd\"", "_____no_output_____" ] ], [ [ "## Model", "_____no_output_____" ], [ "### Connect to ftp", "_____no_output_____" ] ], [ [ "ftp = ftp.connect(user, passwd)", "_____no_output_____" ] ], [ [ "## Output", "_____no_output_____" ], [ "### Get the path", "_____no_output_____" ] ], [ [ "ftp = ftp.connect(user, passwd)\nftp.get(path)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# import dependencies\nimport pandas as pd\nimport requests\nimport json\n", "_____no_output_____" ], [ "# read csv on covid-19 covid vulnerability index data and convert to dataframe\nccvi = pd.read_csv('../resources/ccvi.csv')\n\n# drop rows that contain any null values (there are 655 of them)\nccvi = ccvi.dropna(how='any')\n\n# display dataframe\nccvi\n", "_____no_output_____" ], [ "# get covid data for each race by state\ncovid = pd.read_csv('../resources/CRDT_Data.csv')\n\n# display dataframe\ncovid\n", "_____no_output_____" ], [ "# dictionary for convertying state names to corresponding numbers or abbreviations\nstates = {\n 'southcarolina': {'num': '45', 'abbr': 'SC'},\n 'southdakota': {'num': '46', 'abbr': 'SD'},\n 'tennessee': {'num': '47', 'abbr': 'TN'},\n 'texas': {'num': '48', 'abbr': 'TX'},\n 'vermont': {'num': '50', 'abbr': 'VT'},\n 'utah': {'num': '49', 'abbr': 'UT'},\n 'virginia': {'num': '51', 'abbr': 'VA'},\n 'washington': {'num': '53', 'abbr': 'WA'},\n 'westvirginia': {'num': '54', 'abbr': 'WV'},\n 'wisconsin': {'num': '55', 'abbr': 'WI'},\n 'wyoming': {'num': '56', 'abbr': 'WY'},\n 'puertorico': {'num': '72', 'abbr': 'PR'},\n 'alabama': {'num': '01', 'abbr': 'AL'},\n 'alaska': {'num': '02', 'abbr': 'AK'},\n 'arizona': {'num': '04', 'abbr': 'AZ'},\n 'arkansas': {'num': '05', 'abbr': 'AR'},\n 'california': {'num': '06', 'abbr': 'CA'},\n 'colorado': {'num': '08', 'abbr': 'CO'},\n 'delaware': {'num': '10', 'abbr': 'CT'},\n 'districtofcolumbia': {'num': '11', 'abbr': 'DE'},\n 'connecticut': {'num': '09', 'abbr': 'DC'},\n 'florida': {'num': '12', 'abbr': 'FL'},\n 'georgia': {'num': '13', 'abbr': 'GA'},\n 'idaho': {'num': '16', 'abbr': 'ID'},\n 'hawaii': {'num': '15', 'abbr': 'HI'},\n 'illinois': {'num': '17', 'abbr': 'IL'},\n 'indiana': {'num': '18', 'abbr': 'IN'},\n 'iowa': {'num': '19', 'abbr': 'IA'},\n 'kansas': {'num': '20', 'abbr': 'KS'},\n 'kentucky': {'num': '21', 'abbr': 'KS'},\n 'louisiana': {'num': '22', 'abbr': 'LA'},\n 'maine': {'num': '23', 'abbr': 'ME'},\n 'maryland': {'num': '24', 'abbr': 'MD'},\n 'massachusetts': {'num': '25', 'abbr': 'MA'},\n 'michigan': {'num': '26', 'abbr': 'MI'},\n 'minnesota': {'num': '27', 'abbr': 'MN'},\n 'mississippi': {'num': '28', 'abbr': 'MS'},\n 'missouri': {'num': '29', 'abbr': 'MO'},\n 'montana': {'num': '30', 'abbr': 'MT'},\n 'nebraska': {'num': '31', 'abbr': 'NE'},\n 'nevada': {'num': '32', 'abbr': 'NV'},\n 'newhampshire': {'num': '33', 'abbr': 'NH'},\n 'newjersey': {'num': '34', 'abbr': 'NJ'},\n 'newmexico': {'num': '35', 'abbr': 'NM'},\n 'newyork': {'num': '36', 'abbr': 'NY'},\n 'northcarolina': {'num': '37', 'abbr': 'NC'},\n 'northdakota': {'num': '38', 'abbr': 'ND'},\n 'oregon': {'num': '41', 'abbr': 'OR'},\n 'pennsylvania': {'num': '42', 'abbr': 'PA'},\n 'rhodeisland': {'num': '44', 'abbr': 'RI'}\n}\n\n# all statistical categories to to be queried \npops = 'B01003_001E,B02001_002E,B02001_003E,B02001_004E,B02001_005E,B02001_006E,B03001_003E'\n\n# create list of racial groups to iterate through\nraces = ['total','white','black','native','asian','pacific','hispanic']\n\n# dictionary with all data to be used from all states that made the data avaliable\nstateData = {}\n\n# all states without necessary data\nerror_states = []\n", "_____no_output_____" ], [ "# iterate through states\nfor state in states:\n \n try:\n\n # get census state number\n state_num = states[state]['num']\n\n # create url to request data from api\n url = f'https://api.census.gov/data/2019/acs/acs5?get=NAME,{pops}&for=tract:*&in=state:{state_num}'\n\n # set returned data to a variable\n response = requests.get(url).json()\n\n # create list to store dictionaries with data for each census tract\n tracts = []\n\n # create dictionaries with population data for each census tract \n # (with properly formatted fips code)\n for r in response:\n if r[0] != 'NAME':\n tracts.append({\n 'FIPS': int(f'{r[8]}{r[9]}{r[10]}'),\n 'total': int(r[1]),\n 'white': int(r[2]),\n 'black': int(r[3]),\n 'native': int(r[4]),\n 'asian': int(r[5]),\n 'pacific': int(r[6]),\n 'hispanic': int(r[7])\n })\n\n # create dataframe with census population data\n populations = pd.DataFrame(tracts)\n\n # merge population data and ccvi data on census tract fips code\n ccvi_and_pop = pd.merge(populations, ccvi, on='FIPS')\n\n # create dictionary to hold data for each racial demographic\n demogs = {\n 'total': {},\n 'white': {},\n 'black': {},\n 'native': {},\n 'asian': {},\n 'pacific': {},\n 'hispanic': {}\n }\n\n # iterate through list of races\n for race in races:\n\n # calculate total population for each race\n demogs[race]['population'] = int(ccvi_and_pop[race].sum())\n\n # calculate average ccvi for each race\n demogs[race]['ccvi'] = (ccvi_and_pop[race]*ccvi_and_pop['ccvi']).sum()/demogs[race]['population']\n\n # calculate population of each race as a percentage of total population\n demogs[race]['population_percent'] = (demogs[race]['population']/demogs['total']['population'])*100\n\n # get covid data for each race by state\n covid = pd.read_csv('../resources/CRDT_Data.csv')\n\n # filter to only include data for selected state\n covid = covid.loc[covid['State'] == states[state]['abbr'],:]\n\n # filter to only include data from 2020\n covid = covid.loc[covid['Date'] < 20210000,:]\n\n # create dataframe with only relevant columns for covid cases\n cases = covid[['Cases_Total','Cases_White','Cases_Black','Cases_AIAN','Cases_Asian','Cases_NHPI','Cases_Ethnicity_Hispanic']]\n\n # create dataframe with only relevant columns for covid deaths\n deaths = covid[['Deaths_Total','Deaths_White','Deaths_Black','Deaths_AIAN','Deaths_Asian','Deaths_NHPI','Deaths_Ethnicity_Hispanic']]\n\n # iterate through covid data for selected races and place data in a dictionary\n for i in range(0, len(cases.columns)):\n\n # total cases for each race\n demogs[races[i]]['cases'] = int(cases[cases.columns[i]].values[0])\n\n # number of cases for each race as a percentage of total cases\n demogs[races[i]]['percent_of_cases'] = (demogs[races[i]]['cases']/demogs['total']['cases'])*100\n\n # percent discrepancy between percent of total cases and percent of total population for by each race\n # (theoretically each race should account for the same percent of cases as their percent of the population)\n demogs[races[i]]['discrepancy_percent'] = (demogs[races[i]]['percent_of_cases']/demogs[races[i]]['population_percent'])*100\n\n # total deaths for each race\n demogs[races[i]]['deaths'] = int(deaths[deaths.columns[i]].values[0])\n\n # chance of an infection resulting in death for each race\n demogs[races[i]]['chance_of_death'] = (demogs[races[i]]['deaths']/demogs[races[i]]['cases'])*100\n\n # number of deaths for each race as a percentage of total deaths\n demogs[races[i]]['percent_of_deaths'] = (demogs[races[i]]['deaths']/demogs['total']['deaths'])*100\n\n # create dataframe without total population values\n demographics = pd.DataFrame(demogs).drop(columns=['total'])\n\n\n # create dictionary to hold calculated values to be used in max patch\n for_max = {}\n\n # iterate through statistical categories\n for row in list(demographics.index):\n\n # create a list that holds all values within the row of a statistical category\n values = demographics.loc[row].values\n\n # iterate through races\n for i in range(1, len(races)):\n\n # get population numbers\n if row == 'population':\n for_max[races[i]] = {}\n for_max[races[i]][row] = int(values[i-1])\n\n # calculate inverted ccvi values\n elif row == 'ccvi':\n for_max[races[i]]['inverted_ccvi'] = round(100-(values[i-1])*100, 2)\n\n # calculate chances for where next infection will occure\n elif row == 'discrepancy_percent':\n for_max[races[i]]['chance_of_infection'] = round((values[i-1]/values.sum())*100, 2)\n\n # get values for chance of infection resulting in death\n elif row == 'chance_of_death':\n for_max[races[i]][row] = round(values[i-1], 2)\n\n # create keys to hold number of cases and deaths generated by Max alogrithm\n for key in for_max:\n for_max[key]['generated_cases'] = 0\n for_max[key]['generated_deaths'] = 0\n\n stateData[state] = for_max\n \n except:\n \n error_states.append(state)\n\n", "<ipython-input-5-4a8ca7f482d8>:57: RuntimeWarning: invalid value encountered in double_scalars\n demogs[race]['ccvi'] = (ccvi_and_pop[race]*ccvi_and_pop['ccvi']).sum()/demogs[race]['population']\n" ], [ "# display avaliable states in alphabetical order\nstateData = sorted(stateData)\nfor state in stateData:\n print(state)", "alaska\narkansas\ncalifornia\ncolorado\ngeorgia\nillinois\niowa\nmaine\nminnesota\nmissouri\nnebraska\noregon\ntennessee\nutah\nwashington\nwyoming\n" ], [ "# display unavaliable states\nfor state in error_states:\n print(state)", "southcarolina\nsouthdakota\ntexas\nvermont\nvirginia\nwestvirginia\nwisconsin\npuertorico\nalabama\narizona\ndelaware\ndistrictofcolumbia\nconnecticut\nflorida\nidaho\nhawaii\nindiana\nkansas\nkentucky\nlouisiana\nmaryland\nmassachusetts\nmichigan\nmississippi\nmontana\nnevada\nnewhampshire\nnewjersey\nnewmexico\nnewyork\nnorthcarolina\nnorthdakota\npennsylvania\nrhodeisland\n" ], [ "stateData", "_____no_output_____" ], [ "with open(\"../resources/stateData.json\", \"w\") as outfile:\n json.dump(stateData, outfile)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "# User testing for for Scikit-Yellowbrick\n", "_____no_output_____" ], [ "### Using data that was recorded from sensors during Data Science Certificate Program at GW\nhttps://github.com/georgetown-analytics/classroom-occupancy \n\nData consist of temperature, humidity, CO2 levels, light, # of bluetooth devices, noise levels and count of people in the room.", "_____no_output_____" ] ], [ [ "import pandas as pd\n%matplotlib inline", "_____no_output_____" ], [ "dataset = pd.read_csv('dataset.csv')", "_____no_output_____" ], [ "dataset.head(5)", "_____no_output_____" ], [ "dataset.count_total.describe()", "_____no_output_____" ], [ "#add a new column to create a binary class for room occupancy \ncountmed = dataset.count_total.median()\ndataset['room_occupancy'] = dataset['count_total'].apply(lambda x: 'occupied' if x > 4 else 'empty')", "_____no_output_____" ], [ "# map room occupancy to a number\ndataset['room_occupancy_num'] = dataset.room_occupancy.map({'empty':0, 'occupied':1})", "_____no_output_____" ], [ "dataset.head(5)", "_____no_output_____" ], [ "dataset.room_occupancy.describe()", "_____no_output_____" ], [ "import os\nimport sys \n\n# Modify the path \nsys.path.append(\"..\")\n\nimport pandas as pd\nimport yellowbrick as yb \nimport matplotlib.pyplot as plt\n\nplt.rcParams['figure.figsize'] = (12, 8)", "_____no_output_____" ], [ "g = yb.anscombe()", "_____no_output_____" ] ], [ [ "## Feature Analysis", "_____no_output_____" ], [ "Feature analysis visualizers are designed to visualize instances in data space in order to detect features or targets that might impact downstream fitting. Because ML operates on high-dimensional data sets (usually at least 35), the visualizers focus on aggregation, optimization, and other techniques to give overviews of the data. It is our intent that the steering process will allow the data scientist to zoom and filter and explore the relationships between their instances and between dimensions.\n\nAt the moment we have three feature analysis visualizers implemented:\n\nRank2D: rank pairs of features to detect covariance\n\nRadViz: plot data points along axes ordered around a circle to detect separability\n\nParallel Coordinates: plot instances as lines along vertical axes to detect clusters\n\nFeature analysis visualizers implement the Transformer API from Scikit-Learn, meaning they can be used as intermediate transform steps in a Pipeline (particularly a VisualPipeline). They are instantiated in the same way, and then fit and transform are called on them, which draws the instances correctly. Finally show or show is called which displays the image.", "_____no_output_____" ] ], [ [ "from yellowbrick.features.rankd import Rank2D \nfrom yellowbrick.features.radviz import RadViz \nfrom yellowbrick.features.pcoords import ParallelCoordinates", "_____no_output_____" ] ], [ [ "### Rank2D\n\nRank1D and Rank2D evaluate single features or pairs of features using a variety of metrics that score the features on the scale [-1, 1] or [0, 1] allowing them to be ranked. A similar concept to SPLOMs, the scores are visualized on a lower-left triangle heatmap so that patterns between pairs of features can be easily discerned for downstream analysis.", "_____no_output_____" ] ], [ [ "# Load the classification data set\ndata = dataset\n\n# Specify the features of interest\nfeatures = ['temperature','humidity','co2','light','noise','bluetooth_devices']\n\n# Extract the numpy arrays from the data frame\nX = data[features].as_matrix()\ny = data['count_total'].as_matrix()", "_____no_output_____" ], [ "# Instantiate the visualizer with the Covariance ranking algorithm \nvisualizer = Rank2D(features=features, algorithm='covariance')\n\nvisualizer.fit(X, y) # Fit the data to the visualizer\nvisualizer.transform(X) # Transform the data\nvisualizer.show() # Draw/show/show the data", "_____no_output_____" ], [ "# Instantiate the visualizer with the Pearson ranking algorithm\nvisualizer = Rank2D(features=features, algorithm='pearson')\n\nvisualizer.fit(X, y) # Fit the data to the visualizer\nvisualizer.transform(X) # Transform the data\nvisualizer.show() # Draw/show/show the data", "_____no_output_____" ] ], [ [ "### RadViz\n\nRadViz is a multivariate data visualization algorithm that plots each feature dimension uniformely around the circumference of a circle then plots points on the interior of the circle such that the point normalizes its values on the axes from the center to each arc. This meachanism allows as many dimensions as will easily fit on a circle, greatly expanding the dimensionality of the visualization.\nData scientists use this method to dect separability between classes. E.g. is there an opportunity to learn from the feature set or is there just too much noise?", "_____no_output_____" ] ], [ [ "# Specify the features of interest and the classes of the target \nfeatures = ['temperature','humidity','co2','light','noise','bluetooth_devices']\nclasses = ['empty', 'occupied']\n\n# Extract the numpy arrays from the data frame \nX = data[features].as_matrix()\ny = data.room_occupancy_num.as_matrix()", "_____no_output_____" ], [ "# Instantiate the visualizer\nvisualizer = visualizer = RadViz(classes=classes, features=features)\n\nvisualizer.fit(X, y) # Fit the data to the visualizer\nvisualizer.transform(X) # Transform the data\nvisualizer.show() # Draw/show/show the data", "_____no_output_____" ] ], [ [ "For regression, the RadViz visualizer should use a color sequence to display the target information, as opposed to discrete colors.", "_____no_output_____" ], [ "## Parallel Coordinates\n\n### !!! On this step notebook crashes and has to be restarted", "_____no_output_____" ] ], [ [ "# Specify the features of interest and the classes of the target \n#features = ['temperature','humidity','co2','light','noise','bluetooth_devices']\n#classes = ['empty', 'occupied']\n\n# Extract the numpy arrays from the data frame \n#X = data[features].as_matrix()\n#y = data.room_occupancy_num.as_matrix()", "_____no_output_____" ], [ "# Instantiate the visualizer\n#visualizer = visualizer = ParallelCoordinates(classes=classes, features=features)\n\n#visualizer.fit(X, y) # Fit the data to the visualizer\n#visualizer.transform(X) # Transform the data\n#visualizer.show() # Draw/show/show the data", "_____no_output_____" ] ], [ [ "## Regressor Evaluation", "_____no_output_____" ], [ "Regression models attempt to predict a target in a continuous space. Regressor score visualizers display the instances in model space to better understand how the model is making predictions. We currently have implemented two regressor evaluations:\n\nResiduals Plot: plot the difference between the expected and actual values\n\nPrediction Error: plot expected vs. the actual values in model space\n\nEstimator score visualizers wrap Scikit-Learn estimators and expose the Estimator API such that they have fit(), predict(), and score() methods that call the appropriate estimator methods under the hood. Score visualizers can wrap an estimator and be passed in as the final step in a Pipeline or VisualPipeline.", "_____no_output_____" ] ], [ [ "# Regression Evaluation Imports \n\nfrom sklearn.linear_model import Ridge, Lasso \nfrom sklearn.model_selection import train_test_split\n\nfrom yellowbrick.regressor import PredictionError, ResidualsPlot", "_____no_output_____" ] ], [ [ "### Residuals Plot\n\nA residual plot shows the residuals on the vertical axis and the independent variable on the horizontal axis. If the points are randomly dispersed around the horizontal axis, a linear regression model is appropriate for the data; otherwise, a non-linear model is more appropriate.", "_____no_output_____" ] ], [ [ "# Load the data\ndf = data\nfeature_names = ['temperature','humidity','co2','light','noise','bluetooth_devices']\ntarget_name = 'count_total'\n\n# Get the X and y data from the DataFrame \nX = df[feature_names].as_matrix()\ny = df[target_name].as_matrix() \n\n# Create the train and test data \nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)", "_____no_output_____" ], [ "# Instantiate the linear model and visualizer \nridge = Ridge()\nvisualizer = ResidualsPlot(ridge)\n\nvisualizer.fit(X_train, y_train) # Fit the training data to the visualizer\nvisualizer.score(X_test, y_test) # Evaluate the model on the test data \ng = visualizer.show() # Draw/show/show the data", "_____no_output_____" ] ], [ [ "### Prediction Error Plot\n\nPlots the actual targets from the dataset against the predicted values generated by our model. This allows us to see how much variance is in the model. Data scientists diagnose this plot by comparing against the 45 degree line, where the prediction exactly matches the model.", "_____no_output_____" ] ], [ [ "# Load the data\ndf = data\nfeature_names = ['temperature','humidity','co2','light','noise','bluetooth_devices']\ntarget_name = 'count_total'\n\n# Get the X and y data from the DataFrame \nX = df[feature_names].as_matrix()\ny = df[target_name].as_matrix() \n\n# Create the train and test data \nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)", "_____no_output_____" ], [ "# Instantiate the linear model and visualizer \nlasso = Lasso()\nvisualizer = PredictionError(lasso)\n\nvisualizer.fit(X_train, y_train) # Fit the training data to the visualizer\nvisualizer.score(X_test, y_test) # Evaluate the model on the test data \ng = visualizer.show() # Draw/show/show the data", "_____no_output_____" ] ], [ [ "## Classifier Evaluation\n\nClassification models attempt to predict a target in a discrete space, that is assign an instance of dependent variables one or more categories. Classification score visualizers display the differences between classes as well as a number of classifier-specific visual evaluations. We currently have implemented three classifier evaluations:\n\nClassificationReport: Presents the confusion matrix of the classifier as a heatmap\n\nROCAUC: Presents the graph of receiver operating characteristics along with area under the curve\n\nClassBalance: Displays the difference between the class balances and support\n\nEstimator score visualizers wrap Scikit-Learn estimators and expose the Estimator API such that they have fit(), predict(), and score() methods that call the appropriate estimator methods under the hood. Score visualizers can wrap an estimator and be passed in as the final step in a Pipeline or VisualPipeline.", "_____no_output_____" ] ], [ [ "# Classifier Evaluation Imports \n\nfrom sklearn.naive_bayes import GaussianNB\nfrom sklearn.linear_model import LogisticRegression \nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.model_selection import train_test_split\n\nfrom yellowbrick.classifier import ClassificationReport, ROCAUC, ClassBalance", "_____no_output_____" ] ], [ [ "### Classification report\n\nThe classification report visualizer displays the precision, recall, and F1 scores for the model. Integrates numerical scores as well color-coded heatmap in order for easy interpretation and detection.", "_____no_output_____" ] ], [ [ "# Load the classification data set\ndata = dataset\n\n# Specify the features of interest and the classes of the target \nfeatures = ['temperature','humidity','co2','light','noise','bluetooth_devices']\nclasses = ['empty', 'occupied']\n\n# Extract the numpy arrays from the data frame \nX = data[features].as_matrix()\ny = data.room_occupancy_num.as_matrix()\n\n# Create the train and test data \nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)", "_____no_output_____" ], [ "# Instantiate the classification model and visualizer \nbayes = GaussianNB()\nvisualizer = ClassificationReport(bayes, classes=classes)\n\nvisualizer.fit(X_train, y_train) # Fit the training data to the visualizer\nvisualizer.score(X_test, y_test) # Evaluate the model on the test data \ng = visualizer.show() # Draw/show/show the data", "_____no_output_____" ] ], [ [ "### ROCAUC\n\nPlot the ROC to visualize the tradeoff between the classifier's sensitivity and specificity.", "_____no_output_____" ] ], [ [ "# Instantiate the classification model and visualizer \nlogistic = LogisticRegression()\nvisualizer = ROCAUC(logistic)\n\nvisualizer.fit(X_train, y_train) # Fit the training data to the visualizer\nvisualizer.score(X_test, y_test) # Evaluate the model on the test data \ng = visualizer.show() # Draw/show/show the data", "_____no_output_____" ] ], [ [ "### ClassBalance\n\nClass balance chart that shows the support for each class in the fitted classification model.", "_____no_output_____" ] ], [ [ "# Instantiate the classification model and visualizer \nforest = RandomForestClassifier()\nvisualizer = ClassBalance(forest, classes=classes)\n\nvisualizer.fit(X_train, y_train) # Fit the training data to the visualizer\nvisualizer.score(X_test, y_test) # Evaluate the model on the test data \ng = visualizer.show() # Draw/show/show the data", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

[ "FSFAP" ]

[ "FSFAP" ]

[ "FSFAP" ]

[ [ [ "# 5. Putting it all together\n**Bring together all of the skills you acquired in the previous chapters to work on a real-life project. From connecting to a database and populating it, to reading and querying it.**", "_____no_output_____" ], [ "It's time to put all your effort so far to good use on a census case study.\n\n### Census case study\nThe case study is broken down into three parts. \n1. we are going to prepare SQLAlchemy and the database. \n2. we will load the data into the database. \n3. we solve a few data science type problems with our query knowledge.\n\n### Part 1: preparing SQLAlchemy and the database\nFor part 1 we are going to focus on preparing SQLAlchemy and the database. You might remember this example from Chapter 1. We import `create_engine` and `Metadata`, then create the engine and initialize the metadata.\n```python\nfrom sqlalchemy import create_engine, MetaData\nengine = create_engine('sqlite:///census_nyc.sqlite')\nmetadata = MetaData()\n```\n\n### Part 1: preparing SQLAlchemy and the database\nThen we will build the census table to hold our data. You might remember the employees table we built in Chapter 4. We begin by importing the `Table` and `Column` objects along with all the types we are going to use in our table. Next we define our Table using the Table object by giving it a name, the metadata object, and then each of the columns we want in our table. Finally we create the table in the database by using the create all method on the metadata with the engine.\n```python\nfrom sqlalchemy import Table, Column, String, Integer, Numeric, Boolean\n\nengine = create_engine('sqlite:///')\nmetadata = MetaData()\n\nemployees = Table('employees', metadata,\n Column('id', Integer()),\n Column('name', String(255)),\n Column('salary', Numeric()),\n Column('active', Boolean()))\nmetadata.create_all(engine)\n```", "_____no_output_____" ], [ "## Setup the engine and metadata\nIn this exercise, your job is to create an engine to the database that will be used in this chapter. Then, you need to initialize its metadata.\n\nRecall how you did this in Chapter 1 by leveraging `create_engine()` and `MetaData()`.", "_____no_output_____" ], [ "- Import `create_engine` and `MetaData` from `sqlalchemy`.\n- Create an `engine` to the chapter 5 database by using `'sqlite:///chapter5.sqlite'` as the connection string.\n- Create a MetaData object as `metadata`.", "_____no_output_____" ] ], [ [ "# Import create_engine, MetaData\nfrom sqlalchemy import create_engine, MetaData\n\n# Define an engine to connect to chapter5.sqlite: engine\nengine = create_engine('sqlite:///chapter5.sqlite')\n\n# Initialize MetaData: metadata\nmetadata = MetaData()", "_____no_output_____" ] ], [ [ "## Create the table to the database\nHaving setup the engine and initialized the metadata, you will now define the `census` table object and then create it in the database using the `metadata` and `engine` from the previous exercise. To create it in the database, you will have to use the `.create_all()` method on the `metadata` with `engine` as the argument.", "_____no_output_____" ], [ "- Import `Table`, `Column`, `String`, and `Integer` from `sqlalchemy`.\n- Define a `census` table with the following columns:\n - `'state'` - String - length of 30\n - `'sex'` - String - length of 1\n - `'age'` - Integer\n - `'pop2000'` - Integer\n - `'pop2008'` - Integer\n- Create the table in the database using the `metadata` and `engine`.", "_____no_output_____" ] ], [ [ "# Import Table, Column, String, and Integer\nfrom sqlalchemy import Table, Column, String, Integer\n\n# Build a census table: census\ncensus = Table('census', metadata,\n Column('state', String(30)),\n Column('sex', String(1)),\n Column('age', Integer),\n Column('pop2000', Integer),\n Column('pop2008', Integer))\n\n# Create the table in the database\nmetadata.create_all(engine)", "_____no_output_____" ] ], [ [ "---\n## Populating the database\nWith our table in place, we can now load the data into it. The US Census Agency gave us a CSV file full of data that we need to load into the table.\n\n### Part 2: populating the database\nWe'll start that by building a `values_list` like we did in chapter 4 with this exercise. \n```python\nvalues_list = []\nfor row in csv_reader:\n data = {'state': row[0], 'sex': row[1], 'age': row[2],\n 'pop2000': row[3], 'pop2008': row[4]}\n values_list.append(data)\n```\nWe begin by defining an empty list then looping over the rows of the CSV. Then we build a dictionary for each CSV row that has the data for that row matched up with the column we want to store it in. Then we append the dictionary to the values list.\n\n### Part 2: Populating the Database\nNow we can insert that `values_list` as we did in Chapter 4 like this example. We we start by importing the `insert` statement. Then we build an insert statement for our table, finally we use the execute method on our connection with the statement and values list to insert the data into the table.\n```python\nfrom sqlalchemy import insert\nstmt = insert(employees)\nresult_proxy = connection.execute(stmt, values_list)\nprint(result_proxy.rowcount)\n```\n```\n\n2\n```\nTo review how many rows were inserted, we use the `rowcount` method of the `ResultProxy`.", "_____no_output_____" ], [ "## Reading the data from the CSV\nLeverage the Python CSV module from the standard library and load the data into a list of dictionaries.", "_____no_output_____" ], [ "- Create an empty list called `values_list`.\n- Iterate over the rows of `csv_reader` with a for loop, creating a dictionary called `data` for each row and append it to `values_list`.\n - Within the for loop, `row` will be a list whose entries are `'state'`, `'sex'`, `'age'`, `'pop2000'` and `'pop2008'` (in that order).", "_____no_output_____" ] ], [ [ "import csv\n\ncsv_reader = csv.reader(open('census.csv'))\n\n# Create an empty list: values_list\nvalues_list = []\n\n# Iterate over the rows\nfor row in csv_reader:\n # Create a dictionary with the values\n data = {'state': row[0], 'sex': row[1], 'age': row[2], \n 'pop2000': row[3], 'pop2008': row[4]}\n # Append the dictionary to the values list\n values_list.append(data)", "_____no_output_____" ] ], [ [ "## Load data from a list into the Table\nUsing the multiple insert pattern, in this exercise, you will load the data from `values_list` into the table.", "_____no_output_____" ], [ "- Import `insert` from `sqlalchemy`.\n- Build an insert statement for the `census` table.\n- Execute the statement `stmt` along with `values_list`. You will need to pass them both as arguments to `connection.execute()`.\n- Print the `rowcount` attribute of `results`.", "_____no_output_____" ] ], [ [ "# Import insert\nfrom sqlalchemy import insert\n\n# Build insert statement: stmt\nstmt = insert(census)\n\n# Use values_list to insert data: results\nresults = connection.execute(stmt, values_list)\n\n# Print rowcount\nprint(results.rowcount)", "8772\n" ] ], [ [ "---\n## Querying the database\n### Part 3: answering data science questions with queries\nHere is an example of how we calculated an average in an exercise from Chapter 3. We began by importing the select statement. Next we built a select statement that creates a weighted average. We do this by summing the result of multiplying the age with the population and dividing that by the sum of the total population and labeling that average age. Next we grouped by the sex column to determine the average `age` for each `sex`. Finally, we executed the query and fetched all the results.\n```python\nfrom sqlalchemy import select\nstmt = select([census.columns.sex,\n (func.sum(census.columns.pop2008 *\n census.columns.age) /\n func.sum(census.columns.pop2008)\n ). label('avarage_age')])\nstmt = stmt.group_by(census.columns.sex)\nresutls = connection.execute(stmt).fetchall()\n```\n\n### Part 3: answering data science questions with queries\nWe learned how to calculate a percentage by using the case and cast clauses in Chapter 3. We begin by importing `case`, `cast`, and `Float`. Then we build a select statement that calculates the sum of the `pop2008` column in cases where the state is New York. Then we divided that by the sum of the total population which is cast to a Float so we would get Decimal values. Finally, we multiplied by 100 to get a percentage and labeled it `ny_percent`.\n```python\nfrom sqlalchemy import case, cast, Float\nstmt = select([\n (func.sum(\n case([\n (census.columns.state == 'New York',\n census.columns.pop2008)\n ], else_=0)) /\n cast(func.sum(census.columns.pop2008),\n Float) * 100). label('ny_percent')])\n```\n\nAlso from Chapter 3, we learned how calculate the difference between two columns grouped by another column. We start by building a `select` statement, that selects the column we want to determine the change by, which in this case is `age`. Then we calculate the difference between the population in 2008 and in 2000, and we label that `pop_change`. Remember to wrap the difference calculation in parentheses so you can label it. Next, we order by `pop_change` and finally we limit it to just 5 results.\n```python\nstmt = select([census.columns.age,\n (census.columns.pop2008 -\n census.columns.pop2000).label('pop_chage')\n ])\nstmt = stmt.order_by('pop_change')\nstmt = stmt.limit(5)\n```", "_____no_output_____" ], [ "## Determine the average age by population\nTo calculate a weighted average, we first find the total sum of weights multiplied by the values we're averaging, then divide by the sum of all the weights.\n\nFor example, if we wanted to find a weighted average of `data = [10, 30, 50]` weighted by `weights = [2,4,6]`, we would compute *(2*10 + 4*30 + 6*50) / (2+4+6)*, or `sum(weights * data) / sum(weights)`.\n\nIn this exercise, however, you will make use of **`func.sum()`** together with select to `select` the weighted average of a column from a table. You will still work with the `census` data, and you will compute the average of age weighted by state population in the year 2000, and then group this weighted average by sex.", "_____no_output_____" ], [ "- Import `select` and `func` from `sqlalchemy`.\n- Write a statement to `select` the average of age (`age`) weighted by population in **2000** (`pop2000`) from `census`.", "_____no_output_____" ] ], [ [ "# Import select and func\nfrom sqlalchemy import select, func\n\n# Select the average of age weighted by pop2000\nstmt = select([func.sum(census.columns.pop2000 *\n census.columns.age) /\n func.sum(census.columns.pop2000)])", "_____no_output_____" ] ], [ [ "- Modify the select statement to alias the new column with weighted average as `'average_age'` using `.label()`.", "_____no_output_____" ] ], [ [ "\n# Import select and func\nfrom sqlalchemy import select, func\n\n# Relabel the new column as average_age\nstmt = select([(func.sum(census.columns.pop2000 * \n census.columns.age) / \n func.sum(census.columns.pop2000)).label('average_age')\n\t\t\t ])", "_____no_output_____" ] ], [ [ "- Modify the select statement to select the `sex` column of `census` in addition to the weighted average, with the `sex` column coming first.\n- Group by the `sex` column of `census`.", "_____no_output_____" ] ], [ [ "# Import select and func\nfrom sqlalchemy import select, func\n\n# Add the sex column to the select statement\nstmt = select([census.columns.sex,\n (func.sum(census.columns.pop2000 * \n census.columns.age) / \n func.sum(census.columns.pop2000)).label('average_age'), \n\t\t\t ])\n\n# Group by sex\nstmt = stmt.group_by(census.columns.sex)", "_____no_output_____" ] ], [ [ "- Execute the statement on the `connection` and fetch all the results.\n- Loop over the results and print the values in the `sex` and `average_age` columns for each record in the results.", "_____no_output_____" ] ], [ [ "\n# Import select and func\nfrom sqlalchemy import select, func\n\n# Select sex and average age weighted by 2000 population\nstmt = select([census.columns.sex,\n (func.sum(census.columns.pop2000 * \n census.columns.age) / \n func.sum(census.columns.pop2000)).label('average_age')\n ])\n\n# Group by sex\nstmt = stmt.group_by(census.columns.sex)\n\n# Execute the query and fetch all the results\nconnection = engine.connect()\nresults = connection.execute(stmt).fetchall()\n\n# Print the sex and average age column for each result\nfor result in results:\n print(result.sex, result.average_age)", "F 37\nM 34\n" ] ], [ [ "## Determine the percentage of population by gender and state\nIn this exercise, you will write a query to determine the percentage of the population in 2000 that comprised of women. You will group this query by state.", "_____no_output_____" ], [ "- Import `case`, `cast` and `Float` from `sqlalchemy`.\n- Define a statement to select `state` and the percentage of women in 2000.\n - Inside `func.sum()`, use `case()` to select women (using the `sex` column) from `pop2000`. Remember to specify `else_=0` if the `sex` is not `'F'`.\n - To get the percentage, divide the number of women in the year 2000 by the overall population in 2000. Cast the divisor - `census.columns.pop2000` - to `Float` before multiplying by 100.\n- Group the query by `state`.\n- Execute the query and store it as `results`.\n- Print `state` and `percent_female` for each record.", "_____no_output_____" ] ], [ [ "\n# import case, cast and Float from sqlalchemy\nfrom sqlalchemy import case, cast, Float, desc\n\n# Build a query to calculate the percentage of women in 2000: stmt\nstmt = select([census.columns.state, \n (func.sum(\n case([\n (census.columns.sex == 'F', \n census.columns.pop2000)\n ], else_=0)) /\n cast(func.sum(census.columns.pop2000), \n Float) * 100).label('percent_female')\n])\n\n# Group By state\nstmt = stmt.group_by(census.columns.state)\n\nstmt = stmt.order_by(desc('percent_female'))\n\n# Execute the query and store the results: results\nresults = connection.execute(stmt).fetchall()\n\n# Print the percentage\nfor result in results:\n print(result.state, result.percent_female)", "District of Columbia 53.129626141738385\nRhode Island 52.07343391902215\nMaryland 51.93575549972231\nMississippi 51.92229481794672\nMassachusetts 51.843023571316785\nNew York 51.83453865150073\nAlabama 51.832407770179465\nLouisiana 51.75351596554121\nPennsylvania 51.74043473051053\nSouth Carolina 51.73072129765755\nConnecticut 51.66816507130644\nVirginia 51.657252447241795\nDelaware 51.61109733558627\nNew Jersey 51.51713956125773\nMaine 51.50570813418951\nNorth Carolina 51.482262322084594\nMissouri 51.46888602639692\nOhio 51.46550350015544\nTennessee 51.430689699449275\nWest Virginia 51.40042318092286\nFlorida 51.36488001165242\nKentucky 51.32687036927168\nArkansas 51.26992846221834\nHawaii 51.118011836915514\nGeorgia 51.11408350339436\nOklahoma 51.11362457075227\nIllinois 51.11224234802867\nNew Mexico 51.0471720798335\nVermont 51.018573209949466\nMichigan 50.97246518318712\nIndiana 50.95480313297678\nIowa 50.950398342534264\nNebraska 50.8584549336086\nNew Hampshire 50.858019844961746\nKansas 50.821864107754735\nWisconsin 50.61486452653393\nSouth Dakota 50.52583581373275\nWashington 50.518565087218334\nTexas 50.515721664207966\nNorth Dakota 50.50069363231332\nMinnesota 50.49332944301148\nOregon 50.4294670361772\nCalifornia 50.35233214901979\nMontana 50.32202690728538\nArizona 50.22361303057914\nIdaho 49.98972623903102\nUtah 49.97295275106927\nWyoming 49.94595542648306\nColorado 49.84767060299562\nNevada 49.36736361384359\nAlaska 49.301497893484594\n" ] ], [ [ "*Interestingly, the District of Columbia had the highest percentage of women in 2000, while Alaska had the highest percentage of males.*", "_____no_output_____" ], [ "## Determine the difference by state from the 2000 and 2008 censuses\nIn this final exercise, you will write a query to calculate the states that changed the most in population. You will limit your query to display only the top 10 states.", "_____no_output_____" ], [ "- Build a statement to:\n - Select `state`.\n - Calculate the difference in population between 2008 (`pop2008`) and 2000 (`pop2000`).\n- Group the query by `census.columns.state` using the `.group_by()` method on `stmt`.\n- Order by `'pop_change'` in descending order using the `.order_by()` method with the `desc()` function on `'pop_change'`.\n- ~Limit the query to the top `10` states using the `.limit()` method.~\n- Execute the query and store it as `results`.\n- Print the state and the population change for each result. ", "_____no_output_____" ] ], [ [ "# Build query to return state name and population difference from 2008 to 2000\nstmt = select([census.columns.state, \n (census.columns.pop2008-\n census.columns.pop2000).label('pop_change')\n])\n\n# Group by State\nstmt = stmt.group_by(census.columns.state)\n\n# Order by Population Change\nstmt = stmt.order_by(desc('pop_change'))\n\n# Limit to top 10\n##stmt = stmt.limit(10)\n\n# Use connection to execute the statement and fetch all results\nresults = connection.execute(stmt).fetchall()\n\n# Print the state and population change for each record\nfor result in results:\n print('{}:{}'.format(result.state, result.pop_change))", "Texas:40137\nCalifornia:35406\nFlorida:21954\nArizona:14377\nGeorgia:13357\nNorth Carolina:11574\nVirginia:6639\nColorado:6425\nUtah:5934\nIllinois:5412\nNevada:5367\nWashington:4666\nTennessee:4621\nMissouri:4547\nMinnesota:3763\nOklahoma:3677\nPennsylvania:3384\nSouth Carolina:3360\nWisconsin:2945\nOregon:2817\nMaryland:2551\nArkansas:2549\nIdaho:2500\nIndiana:2336\nNew Mexico:2095\nKentucky:2021\nNebraska:1924\nIowa:1915\nMississippi:1864\nNew York:1851\nNew Jersey:1773\nKansas:1772\nOhio:1585\nAlabama:1576\nHawaii:1454\nSouth Dakota:990\nMontana:960\nDelaware:858\nWyoming:830\nAlaska:740\nDistrict of Columbia:659\nNorth Dakota:585\nWest Virginia:537\nMaine:358\nRhode Island:197\nNew Hampshire:189\nVermont:7\nMassachusetts:-242\nLouisiana:-300\nConnecticut:-392\nMichigan:-2592\n" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "# Prepare environment", "_____no_output_____" ] ], [ [ "!pip install git+https://github.com/katarinagresova/ensembl_scraper.git@6d3bba8e6be7f5ead58a3bbaed6a4e8cd35e62fd", "Collecting git+https://github.com/katarinagresova/ensembl_scraper.git@6d3bba8e6be7f5ead58a3bbaed6a4e8cd35e62fd\n Cloning https://github.com/katarinagresova/ensembl_scraper.git (to revision 6d3bba8e6be7f5ead58a3bbaed6a4e8cd35e62fd) to /tmp/pip-req-build-fz97hoif\n Running command git clone --filter=blob:none -q https://github.com/katarinagresova/ensembl_scraper.git /tmp/pip-req-build-fz97hoif\n Running command git rev-parse -q --verify 'sha^6d3bba8e6be7f5ead58a3bbaed6a4e8cd35e62fd'\n Running command git fetch -q https://github.com/katarinagresova/ensembl_scraper.git 6d3bba8e6be7f5ead58a3bbaed6a4e8cd35e62fd\n Resolved https://github.com/katarinagresova/ensembl_scraper.git to commit 6d3bba8e6be7f5ead58a3bbaed6a4e8cd35e62fd\n Preparing metadata (setup.py) ... \u001b[?25ldone\n\u001b[?25hRequirement already satisfied: bio in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.3.3)\nRequirement already satisfied: biopython in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.79)\nRequirement already satisfied: certifi in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (2021.10.8)\nRequirement already satisfied: charset-normalizer in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (2.0.11)\nRequirement already satisfied: idna in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (3.3)\nRequirement already satisfied: joblib in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.1.0)\nRequirement already satisfied: numpy in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.22.2)\nRequirement already satisfied: pandas in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.4.0)\nRequirement already satisfied: plac in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.3.4)\nRequirement already satisfied: pyfiglet in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (0.8.post1)\nRequirement already satisfied: python-dateutil in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (2.8.2)\nRequirement already satisfied: pytz in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (2021.3)\nRequirement already satisfied: PyYAML>=5.4.1 in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (6.0)\nRequirement already satisfied: requests in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (2.27.1)\nRequirement already satisfied: scikit-learn in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.0.2)\nRequirement already satisfied: scipy in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.8.0)\nRequirement already satisfied: six in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.16.0)\nRequirement already satisfied: threadpoolctl in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (3.1.0)\nRequirement already satisfied: tqdm in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (4.62.3)\nRequirement already satisfied: twobitreader in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (3.1.7)\nRequirement already satisfied: urllib3 in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from scraper==0.0.1) (1.26.8)\nRequirement already satisfied: mygene in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from bio->scraper==0.0.1) (3.2.2)\nRequirement already satisfied: biothings-client>=0.2.6 in /home/katarina/git/genomic_benchmarks/venv/lib/python3.8/site-packages (from mygene->bio->scraper==0.0.1) (0.2.6)\n\u001b[33mWARNING: You are using pip version 21.3.1; however, version 22.0.3 is available.\nYou should consider upgrading via the '/home/katarina/git/genomic_benchmarks/venv/bin/python -m pip install --upgrade pip' command.\u001b[0m\n" ] ], [ [ "# Create config file", "_____no_output_____" ] ], [ [ "import yaml\n\nconfig = {\n \"root_dir\": \"../../datasets/\",\n \"organisms\": {\n \"homo_sapiens\": {\n \"regulatory_feature\"\n }\n }\n}\n\nuser_config = 'user_config.yaml'\nwith open(user_config, 'w') as handle:\n yaml.dump(config, handle)", "_____no_output_____" ] ], [ [ "# Prepare directories", "_____no_output_____" ] ], [ [ "from pathlib import Path\n\nBASE_FILE_PATH = Path(\"../../datasets/human_ensembl_regulatory/\")\n\n# copied from https://stackoverflow.com/a/57892171\ndef rm_tree(pth: Path):\n for child in pth.iterdir():\n if child.is_file():\n child.unlink()\n else:\n rm_tree(child)\n pth.rmdir()\n\nif BASE_FILE_PATH.exists():\n rm_tree(BASE_FILE_PATH)", "_____no_output_____" ] ], [ [ "# Run tool", "_____no_output_____" ] ], [ [ "!python -m scraper.ensembl_scraper -c user_config.yaml", "Processing organisms: 0%| | 0/1 [00:00<?, ?it/s]\nProcessing feature files: 0%| | 0/1 [00:00<?, ?it/s]\u001b[AINFO:root:download_file(): Going to download file from path ftp://ftp.ensembl.org/pub/release-100/mysql/regulation_mart_100/hsapiens_regulatory_feature__regulatory_feature__main.txt.gz\nINFO:root:download_file(): File downloaded to path ../../datasets//tmp//homo_sapiens_regulatory_feature.txt.gz.\nINFO:root:parse_feature_file(): Going to parse file ../../datasets//tmp//homo_sapiens_regulatory_feature.txt.gz\nINFO:root:parse_feature_file(): Done parsing file ../../datasets//tmp//homo_sapiens_regulatory_feature.txt.gz\n\n\nProcessing feature types: 0%| | 0/6 [00:00<?, ?it/s]\u001b[A\u001b[AINFO:root:find_sequences(): Going to find sequences based on genomic loci.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 downloaded to path ../../datasets//tmp/hg38.2bit.\nINFO:root:find_sequences(): Done finding sequences.\nINFO:root:remove_low_quality(): Going to preprocess sequences.\nINFO:root:remove_low_quality(): Original number of sequences: 141250\nINFO:root:remove_low_quality(): Number of sequences after contigs rejection: 141250\nINFO:root:remove_low_quality(): Number of sequences after outlier rejection: 123915\nINFO:root:remove_low_quality(): Number of sequences after Ns rejection: 123909\nINFO:root:remove_low_quality(): Done preprocessing sequences.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\n\n\nProcessing feature types: 17%|██▌ | 1/6 [46:10<3:50:52, 2770.60s/it]\u001b[A\u001b[AINFO:root:find_sequences(): Going to find sequences based on genomic loci.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:find_sequences(): Done finding sequences.\nINFO:root:remove_low_quality(): Going to preprocess sequences.\nINFO:root:remove_low_quality(): Original number of sequences: 177376\nINFO:root:remove_low_quality(): Number of sequences after contigs rejection: 177376\nINFO:root:remove_low_quality(): Number of sequences after outlier rejection: 152129\nINFO:root:remove_low_quality(): Number of sequences after Ns rejection: 152106\nINFO:root:remove_low_quality(): Done preprocessing sequences.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\n\n\nProcessing feature types: 33%|████▎ | 2/6 [1:17:19<2:29:21, 2240.25s/it]\u001b[A\u001b[AINFO:root:find_sequences(): Going to find sequences based on genomic loci.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:find_sequences(): Done finding sequences.\nINFO:root:remove_low_quality(): Going to preprocess sequences.\nINFO:root:remove_low_quality(): Original number of sequences: 132592\nINFO:root:remove_low_quality(): Number of sequences after contigs rejection: 132592\nINFO:root:remove_low_quality(): Number of sequences after outlier rejection: 106894\nINFO:root:remove_low_quality(): Number of sequences after Ns rejection: 106890\nINFO:root:remove_low_quality(): Done preprocessing sequences.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\n\n\nProcessing feature types: 50%|██████▌ | 3/6 [1:37:08<1:28:00, 1760.04s/it]\u001b[A\u001b[AINFO:root:find_sequences(): Going to find sequences based on genomic loci.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:find_sequences(): Done finding sequences.\nINFO:root:remove_low_quality(): Going to preprocess sequences.\nINFO:root:remove_low_quality(): Original number of sequences: 97099\nINFO:root:remove_low_quality(): Number of sequences after contigs rejection: 96572\nINFO:root:remove_low_quality(): Number of sequences after outlier rejection: 87381\nINFO:root:remove_low_quality(): Number of sequences after Ns rejection: 87378\nINFO:root:remove_low_quality(): Done preprocessing sequences.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\n\n\nProcessing feature types: 67%|██████████ | 4/6 [1:51:27<46:48, 1404.49s/it]\u001b[A\u001b[AINFO:root:find_sequences(): Going to find sequences based on genomic loci.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:find_sequences(): Done finding sequences.\nINFO:root:remove_low_quality(): Going to preprocess sequences.\nINFO:root:remove_low_quality(): Original number of sequences: 35191\nINFO:root:remove_low_quality(): Number of sequences after contigs rejection: 35191\nINFO:root:remove_low_quality(): Number of sequences after outlier rejection: 32260\nINFO:root:remove_low_quality(): Number of sequences after Ns rejection: 32258\nINFO:root:remove_low_quality(): Done preprocessing sequences.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\n\n\nProcessing feature types: 83%|█████████████▎ | 5/6 [1:54:37<16:06, 966.54s/it]\u001b[A\u001b[AINFO:root:find_sequences(): Going to find sequences based on genomic loci.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:find_sequences(): Done finding sequences.\nINFO:root:remove_low_quality(): Going to preprocess sequences.\nINFO:root:remove_low_quality(): Original number of sequences: 30436\nINFO:root:remove_low_quality(): Number of sequences after contigs rejection: 29820\nINFO:root:remove_low_quality(): Number of sequences after outlier rejection: 25816\nINFO:root:remove_low_quality(): Number of sequences after Ns rejection: 25816\nINFO:root:remove_low_quality(): Done preprocessing sequences.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\nINFO:root:download_2bit_file(): Going to download 2bit file hg38\nINFO:root:download_2bit_file(): File for hg38 already exists. Not going to download.\n\n\nProcessing feature types: 100%|███████████████| 6/6 [1:56:39<00:00, 1166.62s/it]\u001b[A\u001b[A\n\nProcessing feature files: 100%|███████████████| 1/1 [1:56:59<00:00, 7019.15s/it]\u001b[A\nProcessing organisms: 100%|███████████████████| 1/1 [1:56:59<00:00, 7019.15s/it]\n" ] ], [ [ "# Reformating", "_____no_output_____" ] ], [ [ "!mkdir -p ../../datasets/human_ensembl_regulatory/train\n!mkdir -p ../../datasets/human_ensembl_regulatory/test", "_____no_output_____" ], [ "!mv ../../datasets/homo_sapiens_regulatory_feature_open_chromatin_region/train/positive.csv ../../datasets/human_ensembl_regulatory/train/ocr.csv\n!mv ../../datasets/homo_sapiens_regulatory_feature_open_chromatin_region/test/positive.csv ../../datasets/human_ensembl_regulatory/test/ocr.csv\n\n!mv ../../datasets/homo_sapiens_regulatory_feature_promoter/train/positive.csv ../../datasets/human_ensembl_regulatory/train/promoter.csv\n!mv ../../datasets/homo_sapiens_regulatory_feature_promoter/test/positive.csv ../../datasets/human_ensembl_regulatory/test/promoter.csv\n\n!mv ../../datasets/homo_sapiens_regulatory_feature_enhancer/train/positive.csv ../../datasets/human_ensembl_regulatory/train/enhancer.csv\n!mv ../../datasets/homo_sapiens_regulatory_feature_enhancer/test/positive.csv ../../datasets/human_ensembl_regulatory/test/enhancer.csv", "_____no_output_____" ], [ "def chop_sequnces(file_path, max_len):\n\n df = pd.read_csv(file_path)\n df_array = df.values\n\n new_df_array = []\n index = 0\n for row in df_array:\n splits = ((row[3] - row[2]) // max_len) + 1\n\n if splits == 1:\n new_df_array.append([index, row[1], row[2], row[3], row[4]])\n index += 1\n\n elif splits == 2:\n length = (row[3] - row[2]) // 2\n new_df_array.append([\n index,\n row[1],\n row[2],\n row[2] + length,\n row[4] \n ])\n index += 1\n new_df_array.append([\n index,\n row[1],\n row[2] + length + 1,\n row[3],\n row[4] \n ])\n index += 1\n else:\n length = (row[3] - row[2]) // splits\n new_df_array.append([\n index,\n row[1],\n row[2],\n row[2] + length,\n row[4] \n ])\n index += 1\n for i in range(1, splits - 1):\n new_df_array.append([\n index,\n row[1],\n row[2] + i*length + 1,\n row[2] + (i + 1)*length,\n row[4] \n ])\n index += 1\n\n new_df_array.append([\n index,\n row[1],\n row[2] + (splits - 1)*length + 1,\n row[3],\n row[4] \n ])\n index += 1\n new_df = pd.DataFrame(new_df_array, columns=df.columns)\n new_df.to_csv(file_path, index=False)", "_____no_output_____" ], [ "chop_sequnces(\"../../datasets/human_ensembl_regulatory/train/promoter.csv\", 700)\nchop_sequnces(\"../../datasets/human_ensembl_regulatory/test/promoter.csv\", 700)", "_____no_output_____" ], [ "!find ../../datasets/human_ensembl_regulatory/ -type f -name \"*.csv\" -exec gzip {} \\;", "_____no_output_____" ], [ "!mv ../../datasets/homo_sapiens_regulatory_feature_enhancer/metadata.yaml ../../datasets/human_ensembl_regulatory/metadata.yaml", "_____no_output_____" ], [ "with open(\"../../datasets/human_ensembl_regulatory/metadata.yaml\", \"r\") as stream:\n try:\n config = yaml.safe_load(stream)\n except yaml.YAMLError as exc:\n print(exc)\n\nconfig", "_____no_output_____" ], [ "new_config = {\n 'classes' : {\n 'ocr': {\n 'type': config['classes']['positive']['type'],\n 'url': config['classes']['positive']['url'],\n 'extra_processing': 'ENSEMBL_HUMAN_GENOME'\n },\n 'promoter': {\n 'type': config['classes']['positive']['type'],\n 'url': config['classes']['positive']['url'],\n 'extra_processing': 'ENSEMBL_HUMAN_GENOME'\n },\n 'enhancer': {\n 'type': config['classes']['positive']['type'],\n 'url': config['classes']['positive']['url'],\n 'extra_processing': 'ENSEMBL_HUMAN_GENOME'\n }\n },\n 'version': config['version']\n}\nnew_config ", "_____no_output_____" ], [ "with open(\"../../datasets/human_ensembl_regulatory/metadata.yaml\", 'w') as handle:\n yaml.dump(new_config, handle)", "_____no_output_____" ] ], [ [ "# Cleaning", "_____no_output_____" ] ], [ [ "!rm user_config.yaml\n!rm -rf ../../datasets/tmp/\n\n!rm -rf ../../datasets/homo_sapiens_regulatory_feature_CTCF_binding_site\n!rm -rf ../../datasets/homo_sapiens_regulatory_feature_enhancer\n!rm -rf ../../datasets/homo_sapiens_regulatory_feature_promoter\n!rm -rf ../../datasets/homo_sapiens_regulatory_feature_promoter_flanking_region\n!rm -rf ../../datasets/homo_sapiens_regulatory_feature_TF_binding_site\n!rm -rf ../../datasets/homo_sapiens_regulatory_feature_open_chromatin_region\n", "_____no_output_____" ] ], [ [ "# Testing", "_____no_output_____" ] ], [ [ "from genomic_benchmarks.loc2seq import download_dataset\n\ndownload_dataset(\"human_ensembl_regulatory\", local_repo=True)", "Reference /home/katarina/.genomic_benchmarks/fasta/Homo_sapiens.GRCh38.dna.toplevel.fa.gz already exists. Skipping.\n" ], [ "from genomic_benchmarks.data_check import info\n\ninfo(\"human_ensembl_regulatory\", 0, local_repo=True)", "Dataset `human_ensembl_regulatory` has 3 classes: enhancer, ocr, promoter.\n\nThe length of genomic intervals ranges from 71 to 802, with average 429.91753643694585 and median 401.0.\n\nTotally 289061 sequences have been found, 231348 for training and 57713 for testing.\n" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ] ]

[ "BSD-3-Clause" ]

[ "BSD-3-Clause" ]

[ "BSD-3-Clause" ]

[ [ [ "# Source detection with Gammapy\n\n## Context\n\nThe first task in a source catalogue production is to identify significant excesses in the data that can be associated to unknown sources and provide a preliminary parametrization in term of position, extent, and flux. In this notebook we will use Fermi-LAT data to illustrate how to detect candidate sources in counts images with known background.\n\n**Objective: build a list of significant excesses in a Fermi-LAT map**\n\n\n## Proposed approach \n\nThis notebook show how to do source detection with Gammapy using the methods available in `~gammapy.detect`.\nWe will use images from a Fermi-LAT 3FHL high-energy Galactic center dataset to do this:\n\n* perform adaptive smoothing on counts image\n* produce 2-dimensional test-statistics (TS)\n* run a peak finder to detect point-source candidates\n* compute Li & Ma significance images\n* estimate source candidates radius and excess counts\n\nNote that what we do here is a quick-look analysis, the production of real source catalogs use more elaborate procedures.\n\nWe will work with the following functions and classes:\n\n* `~gammapy.maps.WcsNDMap`\n* `~gammapy.detect.ASmooth`\n* `~gammapy.detect.TSMapEstimator`\n* `~gammapy.detect.find_peaks`\n* `~gammapy.detect.compute_lima_image`\n", "_____no_output_____" ], [ "## Setup\n\nAs always, let's get started with some setup ...", "_____no_output_____" ] ], [ [ "%matplotlib inline\nimport matplotlib.pyplot as plt\nimport matplotlib.cm as cm\nfrom gammapy.maps import Map\nfrom gammapy.detect import (\n ASmooth,\n TSMapEstimator,\n find_peaks,\n compute_lima_image,\n)\nfrom gammapy.catalog import SOURCE_CATALOGS\nfrom gammapy.cube import PSFKernel\nfrom gammapy.stats import significance\nfrom astropy.coordinates import SkyCoord\nfrom astropy.convolution import Tophat2DKernel\nimport astropy.units as u\nimport numpy as np", "_____no_output_____" ], [ "# defalut matplotlib colors without grey\ncolors = [\n u\"#1f77b4\",\n u\"#ff7f0e\",\n u\"#2ca02c\",\n u\"#d62728\",\n u\"#9467bd\",\n u\"#8c564b\",\n u\"#e377c2\",\n u\"#bcbd22\",\n u\"#17becf\",\n]", "_____no_output_____" ] ], [ [ "## Read in input images\n\nWe first read in the counts cube and sum over the energy axis:", "_____no_output_____" ] ], [ [ "counts = Map.read(\"$GAMMAPY_DATA/fermi-3fhl-gc/fermi-3fhl-gc-counts.fits.gz\")\nbackground = Map.read(\n \"$GAMMAPY_DATA/fermi-3fhl-gc/fermi-3fhl-gc-background.fits.gz\"\n)\nexposure = Map.read(\n \"$GAMMAPY_DATA/fermi-3fhl-gc/fermi-3fhl-gc-exposure.fits.gz\"\n)\n\nmaps = {\"counts\": counts, \"background\": background, \"exposure\": exposure}\n\nkernel = PSFKernel.read(\n \"$GAMMAPY_DATA/fermi-3fhl-gc/fermi-3fhl-gc-psf.fits.gz\"\n)", "_____no_output_____" ] ], [ [ "## Adaptive smoothing\n\nFor visualisation purpose it can be nice to look at a smoothed counts image. This can be performed using the adaptive smoothing algorithm from [Ebeling et al. (2006)](https://ui.adsabs.harvard.edu/abs/2006MNRAS.368...65E/abstract).\nIn the following example the `threshold` argument gives the minimum significance expected, values below are clipped.\n", "_____no_output_____" ] ], [ [ "%%time\nscales = u.Quantity(np.arange(0.05, 1, 0.05), unit=\"deg\")\nsmooth = ASmooth(threshold=3, scales=scales)\nimages = smooth.run(**maps)", "_____no_output_____" ], [ "plt.figure(figsize=(15, 5))\nimages[\"counts\"].plot(add_cbar=True, vmax=10);", "_____no_output_____" ] ], [ [ "## TS map estimation\n\nThe Test Statistic, TS = 2 ∆ log L ([Mattox et al. 1996](https://ui.adsabs.harvard.edu/abs/1996ApJ...461..396M/abstract)), compares the likelihood function L optimized with and without a given source.\nThe TS map is computed by fitting by a single amplitude parameter on each pixel as described in Appendix A of [Stewart (2009)](https://ui.adsabs.harvard.edu/abs/2009A%26A...495..989S/abstract). The fit is simplified by finding roots of the derivative of the fit statistics (default settings use [Brent's method](https://en.wikipedia.org/wiki/Brent%27s_method)).", "_____no_output_____" ] ], [ [ "%%time\nestimator = TSMapEstimator()\nimages = estimator.run(maps, kernel.data)", "_____no_output_____" ] ], [ [ "### Plot resulting images", "_____no_output_____" ] ], [ [ "plt.figure(figsize=(15, 5))\nimages[\"sqrt_ts\"].plot(add_cbar=True);", "_____no_output_____" ], [ "plt.figure(figsize=(15, 5))\nimages[\"flux\"].plot(add_cbar=True, stretch=\"sqrt\", vmin=0);", "_____no_output_____" ], [ "plt.figure(figsize=(15, 5))\nimages[\"niter\"].plot(add_cbar=True);", "_____no_output_____" ] ], [ [ "## Source candidates\n\nLet's run a peak finder on the `sqrt_ts` image to get a list of point-sources candidates (positions and peak `sqrt_ts` values).\nThe `find_peaks` function performs a local maximun search in a sliding window, the argument `min_distance` is the minimum pixel distance between peaks (smallest possible value and default is 1 pixel).", "_____no_output_____" ] ], [ [ "sources = find_peaks(images[\"sqrt_ts\"], threshold=8, min_distance=1)\nnsou = len(sources)\nsources", "_____no_output_____" ], [ "# Plot sources on top of significance sky image\nplt.figure(figsize=(15, 5))\n\n_, ax, _ = images[\"sqrt_ts\"].plot(add_cbar=True)\n\nax.scatter(\n sources[\"ra\"],\n sources[\"dec\"],\n transform=plt.gca().get_transform(\"icrs\"),\n color=\"none\",\n edgecolor=\"w\",\n marker=\"o\",\n s=600,\n lw=1.5,\n);", "_____no_output_____" ] ], [ [ "Note that we used the instrument point-spread-function (PSF) as kernel, so the hypothesis we test is the presence of a point source. In order to test for extended sources we would have to use as kernel an extended template convolved by the PSF. Alternatively, we can compute the significance of an extended excess using the Li & Ma formalism, which is faster as no fitting is involve.", "_____no_output_____" ], [ "## Li & Ma significance maps\n\nWe can compute significance for an observed number of counts and known background using an extension of equation (17) from the [Li & Ma (1983)](https://ui.adsabs.harvard.edu/abs/1983ApJ...272..317L/abstract) (see `gammapy.stats.significance` for details). We can perform this calculation intergating the counts within different radius. To do so we use an astropy Tophat kernel with the `compute_lima_image` function.\n\n", "_____no_output_____" ] ], [ [ "%%time\nradius = np.array([0.05, 0.1, 0.15, 0.2, 0.25, 0.3, 0.4, 0.5])\npixsize = counts.geom.pixel_scales[0].value\nnr = len(radius)\nsigni = np.zeros((nsou, nr))\nexcess = np.zeros((nsou, nr))\nfor kr in range(nr):\n npixel = radius[kr] / pixsize\n kernel = Tophat2DKernel(npixel)\n result = compute_lima_image(counts, background, kernel)\n signi[:, kr] = result[\"significance\"].data[sources[\"y\"], sources[\"x\"]]\n excess[:, kr] = result[\"excess\"].data[sources[\"y\"], sources[\"x\"]]", "_____no_output_____" ] ], [ [ "For simplicity we saved the significance and excess at the position of the candidates found previously on the TS map, but we could aslo have applied the peak finder on these significances maps for each scale, or alternatively implemented a 3D peak detection (in longitude, latitude, radius). Now let's look at the significance versus integration radius:", "_____no_output_____" ] ], [ [ "plt.figure()\nfor ks in range(nsou):\n plt.plot(radius, signi[ks, :], color=colors[ks])\nplt.xlabel(\"Radius\")\nplt.ylabel(\"Li & Ma Significance\")\nplt.title(\"Guessing optimal radius of each candidate\");", "_____no_output_____" ] ], [ [ "We can add the optimal radius guessed and the corresdponding excess to the source candidate properties table.", "_____no_output_____" ] ], [ [ "# rename the value key to sqrt(TS)_PS\nsources.rename_column(\"value\", \"sqrt(TS)_PS\")\n\nindex = np.argmax(signi, axis=1)\nsources[\"significance\"] = signi[range(nsou), index]\nsources[\"radius\"] = radius[index]\nsources[\"excess\"] = excess[range(nsou), index]\nsources", "_____no_output_____" ], [ "# Plot candidates sources on top of significance sky image with radius guess\nplt.figure(figsize=(15, 5))\n\n_, ax, _ = images[\"sqrt_ts\"].plot(add_cbar=True, cmap=cm.Greys_r)\n\nphi = np.arange(0, 2 * np.pi, 0.01)\nfor ks in range(nsou):\n x = sources[\"x\"][ks] + sources[\"radius\"][ks] / pixsize * np.cos(phi)\n y = sources[\"y\"][ks] + sources[\"radius\"][ks] / pixsize * np.sin(phi)\n ax.plot(x, y, \"-\", color=colors[ks], lw=1.5);", "_____no_output_____" ] ], [ [ "Note that the optimal radius of nested sources is likely overestimated due to their neighbor. We limited this example to only the most significant source above ~8 sigma. When lowering the detection threshold the number of candidated increase together with the source confusion.", "_____no_output_____" ], [ "## What next?\n\nIn this notebook, we have seen how to work with images and compute TS and significance images from counts data, if a background estimate is already available.\n\nHere's some suggestions what to do next:\n\n- Look how background estimation is performed for IACTs with and without the high-level interface in [analysis_1](analysis_1.ipynb) and [analysis_2](analysis_2.ipynb) notebooks, respectively\n- Learn about 2D model fitting in the [image_analysis](image_analysis.ipynb) notebook\n- find more about Fermi-LAT data analysis in the [fermi_lat](fermi_lat.ipynb) notebook\n- Use source candidates to build a model and perform a 3D fitting (see [analysis_3d](analysis_3d.ipynb), [analysis_mwl](analysis_mwl) notebooks for some hints)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "# Quasi-Laplace approximation for Poisson data\n\n- toc: true \n- badges: true\n- comments: true\n- categories: [jupyter]", "_____no_output_____" ], [ "### About\n\nThe [quasi-Laplace approximation]({% post_url 2020-06-22-intuition-for-quasi-Laplace %}) may be extended to approximate the posterior of a Poisson distribution with a Gaussian, as we will see here.\nWe approximate the regularized likelihood \n$\\mathscr{L}_{\\mathrm{reg}}(\\boldsymbol{\\beta})$\ndefined as the product of the likelihood and a Gaussian *regularizer*,\n\\begin{equation*}\n\\mathscr{L}_{\\mathrm{reg}}(\\boldsymbol{\\beta}) \\triangleq p\\left(\\mathbf{y} \\mid \\mathbf{X}, \\boldsymbol{\\beta}\\right)\n\\mathcal{N}\\left( \\boldsymbol{\\beta} \\mid \\mathbf{0}, \\mathbf{\\Lambda}^{-1} \\right) \\propto\n\\mathcal{N}\\left( \\boldsymbol{\\beta} \\mid \\mathbf{m}, \\mathbf{S} \\right)\n\\end{equation*}\nsuch that the mode of the regularized likelihood is near the mode of the posterior.\n \nThe prior $p\\left(\\boldsymbol{\\beta}\\mid g \\right)$ is a mixture of Gaussians with known variances but unknown mixture proportions.\nThe precision matrix $\\mathbf{\\Lambda}$ is defined as a diagonal matrix, $\\mathbf{\\Lambda} \\triangleq \\mathrm{diag}\\left(\\boldsymbol{\\lambda}\\right)$, whose elements $\\lambda_j$ are roughly set to some expected value to ensure that the regularized likelihood is centered at the mode of the posterior.", "_____no_output_____" ] ], [ [ "#collapse_hide\n\nimport numpy as np\nnp.set_printoptions(precision = 4, suppress=True)\n\nfrom scipy import optimize\nfrom scipy.special import gammaln\nimport matplotlib.pyplot as plt\nfrom matplotlib import cm\nfrom matplotlib import ticker as plticker\nfrom mpl_toolkits.axes_grid1 import make_axes_locatable\n\nimport sys\nsys.path.append(\"../../utils/\")\nimport mpl_stylesheet\nmpl_stylesheet.banskt_presentation(fontfamily = 'latex-clearsans', fontsize = 18, colors = 'banskt', dpi = 72)", "_____no_output_____" ] ], [ [ "### Generate toy data\n\nLet us consider a Poisson model with sparse coefficients, so that the number of causal variables `ncausal` is much less than the number of variables `nvar` in the model. This is ensured by sampling the betas from a Gaussian mixture prior of `nGcomp` components with variances given by $\\sigma_k^2$ (`sigmak2`) and the mixture proportions given by `probk`. The sparsity is controlled by the variable `sparsity` that specifies the mixture proportion of the zero-th component $\\mathcal{N}(0, 0)$ (or the delta function).", "_____no_output_____" ] ], [ [ "nsample = 20\nnvar = 30\nnGcomp = 3\nsparsity = 0.8\nprior_strength = 20\nnum_inf = 1e4 # a large number for 1/sigma_k^2 when sigma_k^2 = 0\n\nprobk = np.zeros(nGcomp)\nprobk[0] = sparsity\nprobk[1:(nGcomp - 1)] = (1 - sparsity) / (nGcomp - 1)\nprobk[nGcomp - 1] = 1 - np.sum(probk)\nsigmak2 = np.array([prior_strength * np.square(np.power(2, (i)/nGcomp) - 1) for i in range(nGcomp)])", "_____no_output_____" ] ], [ [ "We use the exponential link function $\\lambda_i = \\exp\\left(\\mathbf{X}_i\\boldsymbol{\\beta}\\right)$ to generate the response variable $y_i$ for $i = 1, \\ldots, N$ samples using the Poisson probability distribution\n\\begin{equation*}\np\\left(y_i \\mid \\mathbf{X}_i, \\boldsymbol{\\beta}\\right) = \\frac{\\lambda_i^{y_i}e^{-\\lambda_i}}{y_i!}\n\\end{equation*}\n$\\mathbf{X}$ is centered and scaled such that for each variable $j$, the variance is $\\mathrm{var}(\\mathbf{x}_j) = 1$.", "_____no_output_____" ] ], [ [ "# collapse-hide\n\ndef standardize(X):\n Xnorm = (X - np.mean(X, axis = 0)) \n Xstd = Xnorm / np.std(Xnorm, axis = 0)\n return Xstd\n\ndef poisson_data(X, beta):\n Xbeta = np.dot(X, beta)\n pred = np.exp(Xbeta)\n Y = np.random.poisson(pred)\n return Y\n\n\nX = np.random.rand(nsample * nvar).reshape(nsample, nvar)\nX = standardize(X)\n\ngammajk = np.random.multinomial(1, probk, size = nvar)\nbeta = np.zeros(nvar)\nfor j in range(nvar):\n if gammajk[j, 0] != 1:\n kidx = np.where(gammajk[j, :] == 1)[0][0]\n kstd = np.sqrt(sigmak2[kidx])\n beta[j] = np.random.normal(loc = 0., scale = kstd)\n\nncausal = beta[beta != 0].shape[0]\nbetavar = np.var(beta[beta != 0])\n\nY = poisson_data(X, beta)", "_____no_output_____" ] ], [ [ "Let us have a look at the generated data.", "_____no_output_____" ] ], [ [ "# collapse-hide\n\nprint(f'There are {ncausal} non-zero coefficients with variance {betavar:.4f}')\n\nfig = plt.figure(figsize = (12,6))\nax1 = fig.add_subplot(121)\nax2 = fig.add_subplot(122)\n\nXbeta = np.dot(X, beta)\nax1.scatter(Xbeta, np.log(Y+1), s = 10)\nax2.hist(Y)\n\nax1.set_xlabel(r'$\\sum_i X_{ni} \\beta_i$')\nax1.set_ylabel(r'$\\log(Y_n + 1)$')\nax2.set_xlabel(r'$Y_n$')\nax2.set_ylabel('Number')\nplt.tight_layout(pad = 2.0)\nplt.show()", "There are 5 non-zero coefficients with variance 3.2003\n" ] ], [ [ "### True posterior vs quasi-Laplace posterior\n\nWe select two causal variables (with maximum effect size) and fix all the others to optimum values to understand how the likelihood and posterior depends on these two chosen variables. To avoid the sum over the indicator variables, we use the joint prior $p\\left(\\boldsymbol{\\beta}, \\boldsymbol{\\gamma} \\mid g\\right)$.", "_____no_output_____" ], [ "Some useful function definitions:", "_____no_output_____" ] ], [ [ "# collapse-hide\n\ndef get_log_likelihood(Y, X, beta):\n Xbeta = np.dot(X, beta)\n logL = np.sum(Y * Xbeta - np.exp(Xbeta))# - gammaln(Y+1))\n return logL\n\ndef get_log_prior(beta, gammajk, probk, sigmak2):\n logprior = 0\n for j, b in enumerate(beta):\n k = np.where(gammajk[j, :] == 1)[0][0]\n logprior += np.log(probk[k])\n if k > 0:\n logprior += - 0.5 * (np.log(2 * np.pi) + np.log(sigmak2[k]) + b * b / sigmak2[k]) \n return logprior\n\ndef plot_contours(ax, X, Y, Z, beta, norm, cstep = 10, zlabel = \"\"):\n zmin = np.min(Z) - 1 * np.std(Z)\n zmax = np.max(Z) + 1 * np.std(Z)\n ind = np.unravel_index(np.argmax(Z, axis=None), Z.shape)\n\n levels = np.linspace(zmin, zmax, 200)\n clevels = np.linspace(zmin, zmax, 20)\n cmap = cm.YlOrRd_r\n\n if norm:\n cset1 = ax.contourf(X, Y, Z, levels, norm = norm,\n cmap=cm.get_cmap(cmap, len(levels) - 1))\n else:\n cset1 = ax.contourf(X, Y, Z, levels,\n cmap=cm.get_cmap(cmap, len(levels) - 1))\n cset2 = ax.contour(X, Y, Z, clevels, colors='k')\n for c in cset2.collections:\n c.set_linestyle('solid')\n\n\n ax.set_aspect(\"equal\")\n ax.scatter(beta[0], beta[1], color = 'blue', s = 100)\n ax.scatter(X[ind[1]], Y[ind[0]], color = 'k', s = 100)\n\n divider = make_axes_locatable(ax)\n cax = divider.append_axes(\"right\", size=\"5%\", pad=0.2)\n cbar = plt.colorbar(cset1, cax=cax)\n ytickpos = np.arange(int(zmin / cstep) * cstep, zmax, cstep)\n cbar.set_ticks(ytickpos)\n if zlabel:\n cax.set_ylabel(zlabel)\n\n #loc = plticker.AutoLocator()\n #ax.xaxis.set_major_locator(loc)\n #ax.yaxis.set_major_locator(loc)\n \ndef regularized_log_likelihood(beta, X, Y, L):\n nvar = beta.shape[0]\n Xbeta = np.dot(X, beta)\n\n ## Function\n llb = np.sum(Y * Xbeta - np.exp(Xbeta))# - gammaln(Y+1))\n #reg = 0.5 * np.sum(np.log(L)) - 0.5 * np.einsum('i,i->i', np.square(beta), L)\n reg = - 0.5 * np.einsum('i,i->i', np.square(beta), L)\n loglik = llb + reg\n \n ## Gradient\n pred = 1 / (1 + np.exp(-Xbeta))\n der = np.einsum('i,ij->j', Y, X) - np.einsum('ij, i->j', X, np.exp(Xbeta)) - np.multiply(beta, L)\n \n return -loglik, -der\n\ndef precisionLL(X, beta, L):\n nvar = X.shape[1]\n Xbeta = np.dot(X, beta)\n pred = np.exp(Xbeta)\n hess = - np.einsum('i,ij,ik->jk', pred, X, X)\n hess[np.diag_indices(nvar)] -= L\n return -hess\n\ndef get_mS(X, Y, beta0, L):\n nvar = X.shape[1]\n args = X, Y, L\n\n gmode = optimize.minimize(regularized_log_likelihood,\n beta0,\n args=args,\n method='L-BFGS-B',\n jac=True,\n bounds=None,\n options={'maxiter': 20000000,\n 'maxfun': 20000000,\n 'ftol': 1e-9,\n 'gtol': 1e-9\n #'disp': True\n })\n \n M = gmode.x\n Sinv = precisionLL(X, M, L)\n return M, Sinv\n\ndef get_qL_log_posterior(beta, L, M, Sinv, logdetSinv, logprior):\n blessM = beta - M\n bMSbM = np.dot(blessM.T, np.dot(Sinv, blessM))\n bLb = np.einsum('i, i', np.square(beta), L)\n logdetLinv = - np.sum(np.log(L))\n logposterior = 0.5 * (logdetSinv + logdetLinv - bMSbM + bLb)\n logposterior += logprior\n return logposterior", "_____no_output_____" ] ], [ [ "We calculate the likelihood, prior, *true* posterior and the quasi-Laplace posterior. Note that the posteriors are not normalized. We apply quasi-Laplace (QL) approximation with some $\\mathbf{\\Lambda}$ and show QL posterior distribution, which is given by\n\\begin{equation*}\np\\left(\\mathbf{y} \\mid \\mathbf{X}, \\boldsymbol{\\beta}\\right) p\\left(\\boldsymbol{\\beta}\\mid g \\right)\n\\propto\n\\frac{\\mathcal{N}\\left( \\boldsymbol{\\beta} \\mid \\mathbf{m}, \\mathbf{S} \\right)\n p\\left(\\boldsymbol{\\beta}\\mid g \\right)\n }{\n \\mathcal{N}\\left( \\boldsymbol{\\beta} \\mid \\mathbf{0}, \\mathbf{\\Lambda}^{-1} \\right)\n }\n\\end{equation*}\nHere, we assume that we know $\\mathbf{\\Lambda}$. In reality, we will not know $\\mathbf{\\Lambda}$ but will have to learn it from the data or make some educated guess from the prior choice.", "_____no_output_____" ] ], [ [ "np.max(beta)", "_____no_output_____" ], [ "# collapse-hide\n\nbchoose = np.argsort(abs(beta))[-2:]\nnplotx = 20\nnploty = 20\n\nb1min = -0.5\nb1max = 4\nb2min = -4\nb2max = 0.5\nbeta1 = np.linspace(b1min, b1max, nplotx)\nbeta2 = np.linspace(b2min, b2max, nploty)\nlogL = np.zeros((nploty, nplotx))\nlogPr = np.zeros((nploty, nplotx))\nlogPs = np.zeros((nploty, nplotx))\nlogQL = np.zeros((nploty, nplotx))\n\nthisbeta = beta.copy()\nmask = np.ones(nvar, bool)\nmask[bchoose] = False\n\ntrue_pi = np.sum(gammajk, axis = 0) / np.sum(gammajk)\n#reg = 1 / np.einsum('i,i', true_pi, sigmak2)\n#regL = np.repeat(reg, nvar)\nregL = np.repeat(num_inf, nvar)\nfor j, b in enumerate(beta):\n k = np.where(gammajk[j, :] == 1)[0][0]\n if k > 0:\n regL[j] = 1 / sigmak2[k]\nM, Sinv = get_mS(X, Y, beta, regL)\nsgndetSinv, logdetSinv = np.linalg.slogdet(Sinv)\n\nfor i, b1 in enumerate(beta1):\n for j, b2 in enumerate(beta2):\n thisbeta[bchoose] = np.array([b1, b2])\n logL[j, i] = get_log_likelihood(Y, X, thisbeta)\n logPr[j, i] = get_log_prior(thisbeta, gammajk, probk, sigmak2)\n logQL[j, i] = get_qL_log_posterior(thisbeta, regL, M, Sinv, logdetSinv, logPr[j, i])\nlogPs = logL + logPr", "_____no_output_____" ] ], [ [ "Here, we plot the contour maps. The x and y-axis show the two coefficients $\\beta_1$ and $\\beta_2$, which we chose to vary. The blue dot shows the coordinates of the true values of $\\{\\beta_1, \\beta_2\\}$ and the black dot shows the maximum of the log probabilities. Note how the non-Gaussian *true* posterior is now approximated by a Gaussian QL posterior.", "_____no_output_____" ] ], [ [ "fig = plt.figure(figsize = (12, 8))\nax1 = fig.add_subplot(221)\nax2 = fig.add_subplot(222)\nax3 = fig.add_subplot(223)\nax4 = fig.add_subplot(224)\n\nnorm = cm.colors.Normalize(vmin=np.min(logPs), vmax=np.max(logPs))\n\nplot_contours(ax1, beta1, beta2, logL, beta[bchoose], None, cstep = 5000, zlabel = \"Log Likelihood\")\nplot_contours(ax2, beta1, beta2, logPr, beta[bchoose], None, cstep = 5, zlabel = \"Log Prior\")\nplot_contours(ax3, beta1, beta2, logPs, beta[bchoose], None, cstep = 5000, zlabel = \"Log Posterior\")\nplot_contours(ax4, beta1, beta2, logQL, beta[bchoose], None, cstep = 5000, zlabel = \"Log QL Posterior\")\n\nplt.tight_layout()\nplt.show()", "_____no_output_____" ] ], [ [ "### Effect of regularizer\n\nLet us assume that we do not know $\\mathbf{\\Lambda}$ and all $\\lambda_j$'s are equal. Here we look at how the QL posterior changes with varying $\\beta_2$ for different values of $\\lambda_j$.\n\nHere we define a function for calculating the QL posterior and true posterior for changing the coefficients of a single variable:", "_____no_output_____" ] ], [ [ "# collapse-hide\n\ndef get_logQL_logPs_single_variable(X, Y, beta, regvar, betavar, bidx,\n regL, gammajk, probk, sigmak2):\n\n nvar = beta.shape[0]\n nplotrv = regvar.shape[0]\n nplotx = betavar.shape[0]\n logL = np.zeros(nplotx)\n logPr = np.zeros(nplotx)\n logPs = np.zeros(nplotx)\n logQL = np.zeros((nplotrv, nplotx))\n\n thisbeta = beta.copy()\n thisL = regL.copy()\n\n for j, b2 in enumerate(betavar):\n thisbeta[bidx] = b2\n logL[j] = get_log_likelihood(Y, X, thisbeta)\n logPr[j] = get_log_prior(thisbeta, gammajk, probk, sigmak2)\n logPs = logL + logPr\n\n for i, r1 in enumerate(regvar):\n thisL = np.repeat(r1, nvar)\n #thisL[bidx] = r1\n M, Sinv = get_mS(X, Y, beta, thisL)\n sgndetSinv, logdetSinv = np.linalg.slogdet(Sinv)\n for j, b2 in enumerate(betavar):\n thisbeta[bidx] = b2\n logQL[i, j] = get_qL_log_posterior(thisbeta, thisL, M, Sinv, logdetSinv, logPr[j])\n\n return logPs, logQL", "_____no_output_____" ] ], [ [ "And then look at the posteriors for $\\beta_2$.", "_____no_output_____" ] ], [ [ "#collapse-hide\n\nnplotx_rv = 200\nnplotrv = 4\nbmin = -4\nbmax = 0\nrvmin = 1\nrvmax = 100\nbidx = bchoose[1]\n\nfig = plt.figure(figsize = (8, 8))\nax1 = fig.add_subplot(111)\n\nbetavals = np.linspace(bmin, bmax, nplotx_rv)\nregvals = np.linspace(rvmin, rvmax, nplotrv)\nlogPs_rv, logQL_rv = get_logQL_logPs_single_variable(X, Y, beta, regvals, betavals, bidx,\n regL, gammajk, probk, sigmak2)\n\nax1.plot(betavals, logPs_rv, lw = 3, zorder = 2, label = 'True Posterior')\nax1.scatter(betavals[np.argmax(logPs_rv)], logPs_rv[np.argmax(logPs_rv)], s = 40, zorder = 10, color = 'black')\nfor i, r1 in enumerate(regvals):\n ax1.plot(betavals, logQL_rv[i, :], lw = 2, zorder = 5, label = f'$\\lambda =${r1:.2f}')\n ix = np.argmax(logQL_rv[i, :])\n ax1.scatter(betavals[ix], logQL_rv[i, ix], s = 40, zorder = 10, color = 'black')\nax1.axvline(beta[bidx], ls = 'dotted', zorder = 1)\n\nax1.legend(handlelength = 1.5)\nax1.set_xlabel(r'$\\beta$')\nax1.set_ylabel('Log Posterior')\nplt.tight_layout()\nplt.show()", "_____no_output_____" ] ], [ [ "What happens to the QL posterior for other variables? Let us look at every $\\beta_j$ and their individual maximum posterior, while all others are kept fixed at optimum values. Here, we have arranged the $\\beta_j$ in ascending order.", "_____no_output_____" ] ], [ [ "#collapse-hide\n\nbmin = -5\nbmax = 5\n\nbidx_sorted = np.argsort(beta)\nbidx_sorted_nz = bidx_sorted #bidx_sorted[beta[bidx_sorted]!=0]\n\nbetavals = np.linspace(bmin, bmax, nplotx_rv)\nregvals = np.linspace(rvmin, rvmax, nplotrv)\n\nmaxQL = np.zeros((nplotrv, len(bidx_sorted_nz)))\nmaxPs = np.zeros(len(bidx_sorted_nz))\nfor i, bidx in enumerate(bidx_sorted_nz):\n _logPs, _logQL = get_logQL_logPs_single_variable(X, Y, beta, regvals, betavals, bidx,\n regL, gammajk, probk, sigmak2)\n maxPs[i] = betavals[np.argmax(_logPs)]\n for j in range(len(regvals)):\n maxQL[j, i] = betavals[np.argmax(_logQL[j, :])]\n \nfig = plt.figure(figsize = (16, 8))\nax1 = fig.add_subplot(111)\n\nxvals = np.arange(len(bidx_sorted_nz))\n\nax1.plot(xvals, maxPs, lw = 2, zorder = 2, label = 'True Posterior')\nfor i, r1 in enumerate(regvals):\n ax1.plot(xvals, maxQL[i, :], lw = 2, zorder = 5, label = f'$\\lambda =${r1:.2f}')\n \nax1.scatter(xvals, maxPs, s = 20, zorder = 1)\nfor i, r1 in enumerate(regvals):\n ax1.scatter(xvals, maxQL[i, :], s = 20, zorder = 1)\nax1.scatter(xvals, beta[bidx_sorted_nz], s = 80, zorder = 10, color = 'maroon', label = 'Input')\n\nax1.legend(handlelength = 1.5)\nax1.set_xlabel(r'Index of $\\beta$')\nax1.set_ylabel(r'$\\beta$ at maximum log posterior')\nax1.set_xticks(xvals)\nax1.set_xticklabels([f'{idx}' for idx in bidx_sorted_nz])\n\nplt.tight_layout()\nplt.show()", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# Modelo Ciclo de Carnot", "_____no_output_____" ], [ "Importar librerias", "_____no_output_____" ] ], [ [ "import numpy as np\nimport matplotlib.pyplot as plt", "_____no_output_____" ] ], [ [ "## Definir funciones\n\nPara la presión de un gas ideal isotérmico y adiabático", "_____no_output_____" ] ], [ [ "# Valores dados para que calce el gráfico\nR = 1\ngamma = 10\n\ndef pressure_ideal(T, v):\n return R*T / v\n\ndef pressure_adiabatic(v, P1, v1):\n return P1*np.power(v1/v, gamma)", "_____no_output_____" ] ], [ [ "## Definir valores a gráficar", "_____no_output_____" ] ], [ [ "Th = 850\nTc = 310\n\ndelta_v = 0.05", "_____no_output_____" ] ], [ [ "### Proceso Isotérmico", "_____no_output_____" ] ], [ [ "vC = 2.5\nvD = 2.0", "_____no_output_____" ] ], [ [ "Calculo de vA y vB", "_____no_output_____" ] ], [ [ "PC = pressure_ideal(Tc, vC)\nPD = pressure_ideal(Tc, vD)\n\nvA = ((PD*(vD**gamma))/(R*Th))**(1/(gamma-1))\nvB = ((PC*(vC**gamma))/(R*Th))**(1/(gamma-1))\n\nv = np.linspace(vA-delta_v, vC+delta_v) # v1, 3", "_____no_output_____" ] ], [ [ "### Proceso adiabático", "_____no_output_____" ] ], [ [ "v_2 = np.linspace(vB-delta_v, vC+delta_v)\nv_4 = np.linspace(vA-delta_v, vD+delta_v)", "_____no_output_____" ], [ "_, ax = plt.subplots(figsize=(8, 6))\n\nax.plot(v, pressure_ideal(Tc, v), '--b', label=\"Isoterma\")\nax.plot(v, pressure_ideal(Th, v), '--b')\nax.plot(v_2, pressure_adiabatic(v_2, PC, vC), '-.r', label=\"Adiabática\")\nax.plot(v_4, pressure_adiabatic(v_4, PD, vD), '-.r')\nax.set_title(\"Ciclo de Carnot\\n\", fontsize=16)\nax.set_xlabel(\"v\", fontsize=14)\nax.set_ylabel(\"P\", fontsize=14)\nax.tick_params(\n which='both',\n bottom=False,\n top=False,\n right=False,\n left=False,\n labelbottom=False,\n labelleft=False)\nax.grid()\nax.legend(fontsize=14)\n\nplt.savefig(\"../Matplotlib/ciclo_carnot.png\")", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# G4: Standard 4D Model\n\n# Make SymPy available to this program:\nimport sympy \nfrom sympy import *\n\n# Make GAlgebra available to this program:\nfrom galgebra.ga import * \nfrom galgebra.mv import *\nfrom galgebra.printer import Fmt, GaPrinter, Format\n # Fmt: sets the way that a multivector's basis expansion is output.\n # GaPrinter: makes GA output a little more readable.\n # Format: turns on latex printer.\nfrom galgebra.gprinter import gFormat, gprint\ngFormat()", "_____no_output_____" ], [ "# g4: R^4 using cartesian coordiantes\ng4coords = (x,y,z,w) = symbols('x y z w', real=True)\ng4 = Ga('e', g=[1,1,1,1], coords=g4coords)\n(ex, ey, ez, ew) = g4.mv()", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# App and Process Controller\n\nThis notebook is used to start client-Processes", "_____no_output_____" ] ], [ [ "# from subprocess import call\r\n# import os", "_____no_output_____" ] ], [ [ "## Test if this notebook works", "_____no_output_____" ] ], [ [ "! start powershell python ./test/testConsoleScript.py", "_____no_output_____" ] ], [ [ "## Start a server", "_____no_output_____" ] ], [ [ "! start powershell python ./server.py", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# Phase Proportions Over Time\n\nGoal is to create a figure like that in Paul et al. [1].\n\n[1] Michael J. Paul, Ryen W. White, and Eric Horvitz. 2015. Diagnoses, Decisions, and Outcomes: Web Search as Decision Support for Cancer. In Proceedings of the 24th International Conference on World Wide Web - WWW ’15 (WWW’15), 831–841. DOI:https://doi.org/10.1145/2736277.2741662\n", "_____no_output_____" ] ], [ [ "%reload_ext autoreload\n%autoreload 2\n%matplotlib inline", "_____no_output_____" ], [ "import sys\nsys.path.append(\"../../annotation_data\")", "_____no_output_____" ], [ "from phase import *", "_____no_output_____" ], [ "import pandas as pd\nimport numpy as np\nimport sklearn\nimport sklearn.metrics\nimport subprocess\nfrom tqdm import tqdm", "_____no_output_____" ], [ "import matplotlib.pyplot as plt\nimport matplotlib.dates as md\nimport matplotlib\nimport pylab as pl", "_____no_output_____" ], [ "vw_working_dir = \"/home/srivbane/shared/caringbridge/data/projects/qual-health-journeys/classification/phases/vw\"", "_____no_output_____" ] ], [ [ "## Load VW phase labels", "_____no_output_____" ] ], [ [ "all_predictions_filepath = os.path.join(vw_working_dir, \"vw_all_preds.pkl\")\npred_df = pd.read_pickle(all_predictions_filepath)\nlen(pred_df)", "_____no_output_____" ], [ "pred_df.head()", "_____no_output_____" ], [ "for site_id, site_preds in tqdm(pred_df.groupby(by='site_id', sort=False)):\n journal_creation_times = site_preds.loc[:, 'created_at']\n journal_labels = site_preds.loc[:,[phase_label + \"_pred_label\" for phase_label in phase_labels]]\n # do something with the labels and the creation time...", "100%|██████████| 4977/4977 [00:05<00:00, 950.92it/s]\n" ] ] ]

[ "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import nltk\nimport pandas as pd\nimport numpy as np\nimport matplotlib.pyplot as plt\nfrom nltk.stem import WordNetLemmatizer\nfrom nltk.corpus import wordnet\nimport re, collections\nfrom collections import defaultdict\nfrom sklearn.feature_extraction.text import CountVectorizer\nfrom sklearn.metrics import mean_squared_error, r2_score, cohen_kappa_score\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.linear_model import LinearRegression, Ridge, Lasso\nfrom sklearn.svm import SVR\nfrom sklearn.ensemble import RandomForestRegressor\nfrom sklearn.ensemble import AdaBoostRegressor\nfrom spellchecker import SpellChecker\nfrom nltk.tokenize import word_tokenize\nimport string\nfrom sklearn.metrics import classification_report\nfrom sklearn import svm\nfrom sklearn.model_selection import cross_val_score\nfrom sklearn.metrics import classification_report\nfrom sklearn.preprocessing import MinMaxScaler\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nfrom matplotlib import rcParams", "_____no_output_____" ] ], [ [ "## Loading data", "_____no_output_____" ] ], [ [ "#dataframe = pd.read_csv('all_essaysets.csv', encoding = 'latin-1')\ndataframe = pd.read_csv('training.tsv', encoding = 'latin-1', sep='\\t')\ndataframe.describe()", "_____no_output_____" ], [ "dataframe.head()", "_____no_output_____" ] ], [ [ "## Methods", "_____no_output_____" ] ], [ [ "# selecting which set to be used 1-8\n# in order to combine them all assign set number to 9\ndef select_set(dataframe,setNumber):\n if setNumber == 9:\n dataframe2 = dataframe[dataframe.essay_set ==1]\n texts = dataframe2['essay']\n scores = dataframe2['domain1_score']\n scores = scores.apply(lambda x: (x*3)/scores.max())\n for i in range(1,9):\n dataframe2 = dataframe[dataframe.essay_set == i]\n texts = texts.append(dataframe2['essay'])\n s = dataframe2['domain1_score']\n s = s.apply(lambda x: (x*3)/s.max())\n scores = scores.append(s)\n else:\n dataframe2 = dataframe[dataframe.essay_set ==setNumber]\n texts = dataframe2['essay']\n scores = dataframe2['domain1_score']\n scores = scores.apply(lambda x: (x*3)/scores.max())\n return texts, scores", "_____no_output_____" ], [ "# get histogram plot of scores and average score\ndef get_hist_avg(scores,bin_count):\n print(sum(scores)/len(scores))\n scores.hist(bins=bin_count)", "_____no_output_____" ], [ "#average word length for a text\ndef avg_word_len(text):\n clean_essay = re.sub(r'\\W', ' ', text)\n words = nltk.word_tokenize(clean_essay)\n total = 0\n for word in words:\n total = total + len(word)\n average = total / len(words)\n \n return average\n\n# word count in a given text\ndef word_count(text):\n clean_essay = re.sub(r'\\W', ' ', text)\n return len(nltk.word_tokenize(clean_essay))\n\n# char count in a given text\ndef char_count(text):\n return len(re.sub(r'\\s', '', str(text).lower()))\n\n# sentence count in a given text\ndef sent_count(text):\n return len(nltk.sent_tokenize(text))\n\n#tokenization of texts to sentences\ndef sent_tokenize(text):\n stripped_essay = text.strip()\n \n tokenizer = nltk.data.load('tokenizers/punkt/english.pickle')\n raw_sentences = tokenizer.tokenize(stripped_essay)\n \n tokenized_sentences = []\n for raw_sentence in raw_sentences:\n if len(raw_sentence) > 0:\n clean_sentence = re.sub(\"[^a-zA-Z0-9]\",\" \", raw_sentence)\n tokens = nltk.word_tokenize(clean_sentence)\n tokenized_sentences.append(tokens)\n return tokenized_sentences\n\n\n# lemma, noun, adjective, verb, adverb count for a given text\n\ndef count_lemmas(text):\n \n noun_count = 0\n adj_count = 0\n verb_count = 0\n adv_count = 0 \n lemmas = []\n lemmatizer = WordNetLemmatizer()\n tokenized_sentences = sent_tokenize(text)\n \n for sentence in tokenized_sentences:\n tagged_tokens = nltk.pos_tag(sentence) \n \n for token_tuple in tagged_tokens:\n pos_tag = token_tuple[1]\n \n if pos_tag.startswith('N'): \n noun_count += 1\n pos = wordnet.NOUN\n lemmas.append(lemmatizer.lemmatize(token_tuple[0], pos))\n elif pos_tag.startswith('J'):\n adj_count += 1\n pos = wordnet.ADJ\n lemmas.append(lemmatizer.lemmatize(token_tuple[0], pos))\n elif pos_tag.startswith('V'):\n verb_count += 1\n pos = wordnet.VERB\n lemmas.append(lemmatizer.lemmatize(token_tuple[0], pos))\n elif pos_tag.startswith('R'):\n adv_count += 1\n pos = wordnet.ADV\n lemmas.append(lemmatizer.lemmatize(token_tuple[0], pos))\n else:\n pos = wordnet.NOUN\n lemmas.append(lemmatizer.lemmatize(token_tuple[0], pos))\n \n lemma_count = len(set(lemmas))\n \n return noun_count, adj_count, verb_count, adv_count, lemma_count", "_____no_output_____" ], [ "def token_word(text):\n text = \"\".join([ch.lower() for ch in text if ch not in string.punctuation])\n tokens = nltk.word_tokenize(text)\n return tokens", "_____no_output_____" ], [ "def misspell_count(text):\n spell = SpellChecker()\n # find those words that may be misspelled\n misspelled = spell.unknown(token_word(text))\n #print(misspelled)\n return len(misspelled)", "_____no_output_____" ], [ "def create_features(texts):\n data = pd.DataFrame(columns=('Average_Word_Length','Sentence_Count','Word_Count',\n 'Character_Count', 'Noun_Count','Adjective_Count',\n 'Verb_Count', 'Adverb_Count', 'Lemma_Count' , 'Misspell_Count'\n ))\n\n data['Average_Word_Length'] = texts.apply(avg_word_len)\n data['Sentence_Count'] = texts.apply(sent_count)\n data['Word_Count'] = texts.apply(word_count)\n data['Character_Count'] = texts.apply(char_count)\n temp=texts.apply(count_lemmas)\n noun_count,adj_count,verb_count,adverb_count,lemma_count = zip(*temp)\n data['Noun_Count'] = noun_count\n data['Adjective_Count'] = adj_count\n data['Verb_Count'] = verb_count\n data['Adverb_Count'] = adverb_count\n data['Lemma_Count'] = lemma_count\n data['Misspell_Count'] = texts.apply(misspell_count)\n return data", "_____no_output_____" ], [ "def data_prepare(texts,scores):\n #create features from the texts and clean non graded essays\n data = create_features(texts)\n data.describe()\n t1=np.where(np.asanyarray(np.isnan(scores)))\n scores=scores.drop(scores.index[t1])\n data=data.drop(scores.index[t1])\n \n #scaler = MinMaxScaler()\n #data = scaler.fit_transform(data)\n\n #train test split\n X_train, X_test, y_train, y_test = train_test_split(data, scores, test_size = 0.3)\n\n #checking is there any nan cells\n print(np.any(np.isnan(scores)))\n print(np.all(np.isfinite(scores)))\n return X_train, X_test, y_train, y_test, data", "_____no_output_____" ], [ "def lin_regression(X_train,y_train,X_test,y_test):\n regr = LinearRegression()\n regr.fit(X_train, y_train)\n y_pred = regr.predict(X_test)\n\n # The mean squared error\n mse=mean_squared_error(y_test, y_pred)\n mse_per= 100*mse/3\n print(\"Mean squared error: {}\".format(mse))\n print(\"Mean squared error in percentage: {}\".format(mse_per))\n #explained variance score\n print('Variance score: {}'.format(regr.score(X_test, y_test)))", "_____no_output_____" ], [ "def adaBoost_reg(X_train,y_train,X_test,y_test):\n #regr = RandomForestRegressor(max_depth=2, n_estimators=300)\n #regr = SVR(gamma='scale', C=1, kernel='linear')\n regr = AdaBoostRegressor()\n regr.fit(X_train, y_train)\n y_pred = regr.predict(X_test)\n # The mean squared error\n mse=mean_squared_error(y_test, y_pred)\n mse_per= 100*mse/3\n print(\"Mean squared error: {}\".format(mse))\n print(\"Mean squared error in percentage: {}\".format(mse_per))\n #explained variance score\n print('Variance score: {}'.format(regr.score(X_test, y_test)))\n\n feature_importance = regr.feature_importances_\n\n # make importances relative to max importance\n feature_importance = 100.0 * (feature_importance / feature_importance.max())\n feature_names = list(('Average_Word_Length','Sentence_Count','Word_Count',\n 'Character_Count', 'Noun_Count','Adjective_Count',\n 'Verb_Count', 'Adverb_Count', 'Lemma_Count' ,'Misspell_Count'\n ))\n feature_names = np.asarray(feature_names)\n sorted_idx = np.argsort(feature_importance)\n pos = np.arange(sorted_idx.shape[0]) + .5\n plt.subplot(1, 2, 2)\n plt.barh(pos, feature_importance[sorted_idx], align='center')\n plt.yticks(pos, feature_names[sorted_idx])\n plt.xlabel('Relative Importance')\n plt.title('Variable Importance')\n plt.show()", "_____no_output_____" ], [ "# convert numerical scores to labels\n# (0-1.5) bad (1.5-2.3) average (2.3-3) good\n# bad: '0'\n# average '1'\n# good '2'\ndef convert_scores(scores):\n def mapping(x):\n if x < np.percentile(scores,25):\n return 0\n elif x < np.percentile(scores,75):\n return 1\n else:\n return 2\n return scores.apply(mapping)", "_____no_output_____" ], [ "# selecting which set to be used 1-8\n# in order to combine them all assign set number to 9\ndef select_set_classification(dataframe,setNumber):\n if setNumber == 9:\n dataframe2 = dataframe[dataframe.essay_set ==1]\n texts = dataframe2['essay']\n scores = dataframe2['domain1_score']\n scores = scores.apply(lambda x: (x*3)/scores.max())\n scores = convert_scores(scores)\n for i in range(1,9):\n dataframe2 = dataframe[dataframe.essay_set == i]\n texts = texts.append(dataframe2['essay'])\n s = dataframe2['domain1_score']\n s = s.apply(lambda x: (x*3)/s.max())\n s = convert_scores(s)\n scores = scores.append(s)\n else:\n dataframe2 = dataframe[dataframe.essay_set ==setNumber]\n texts = dataframe2['essay']\n scores = dataframe2['domain1_score']\n scores = scores.apply(lambda x: (x*3)/scores.max())\n scores = convert_scores(scores)\n return texts, scores", "_____no_output_____" ] ], [ [ "## Dataset selection", "_____no_output_____" ] ], [ [ "# 1-8\n# 9:all sets combined\ntexts, scores = select_set(dataframe,1)\nget_hist_avg(scores,5)\nX_train, X_test, y_train, y_test, data = data_prepare(texts,scores)", "_____no_output_____" ] ], [ [ "## Regression Analysis", "_____no_output_____" ] ], [ [ "print('Testing for Linear Regression \\n')\nlin_regression(X_train,y_train,X_test,y_test)\nprint('Testing for Adaboost Regression \\n')\nadaBoost_reg(X_train,y_train,X_test,y_test)", "_____no_output_____" ] ], [ [ "## Dataset selection 2", "_____no_output_____" ] ], [ [ "# 1-8\n# 9:all sets combined\ntexts, scores = select_set_classification(dataframe,1)\nX_train, X_test, y_train, y_test, data = data_prepare(texts,scores)", "_____no_output_____" ] ], [ [ "## Classification analysis", "_____no_output_____" ] ], [ [ "a=[0.1,1,10,100,500,1000]\nfor b in a:\n clf = svm.SVC(C=b, gamma=0.00001)\n clf.fit(X_train, y_train) \n y_pred = clf.predict(X_test)\n print (b)\n print (clf.score(X_test,y_test))\n print (np.mean(cross_val_score(clf, X_train, y_train, cv=3)))", "_____no_output_____" ], [ "clf = svm.SVC(C=100, gamma=0.00001)\nclf.fit(X_train, y_train) \ny_pred = clf.predict(X_test)\nprint('Cohen’s kappa score: {}'.format(cohen_kappa_score(y_test,y_pred)))", "_____no_output_____" ], [ "print(classification_report(y_test, y_pred))", "_____no_output_____" ] ], [ [ "## Data Analysis", "_____no_output_____" ] ], [ [ "sns.countplot(scores)", "_____no_output_____" ], [ "zero = data[(data[\"Character_Count\"] > 0) & (scores == 0)]\none = data[(data[\"Character_Count\"] > 0) & (scores == 1)]\ntwo = data[(data[\"Character_Count\"] > 0) & (scores == 2)]\nsns.distplot(zero[\"Character_Count\"], bins=10, color='r')\nsns.distplot(one[\"Character_Count\"], bins=10, color='g')\nsns.distplot(two[\"Character_Count\"], bins=10, color='b')\nplt.title(\"Score Distribution with respect to Character_Count\",fontsize=20)\nplt.xlabel(\"Character_Count\",fontsize=15)\nplt.ylabel(\"Distribuition of the scores\",fontsize=15)\nplt.show()", "_____no_output_____" ], [ "zero = data[(data[\"Lemma_Count\"] > 0) & (scores == 0)]\none = data[(data[\"Lemma_Count\"] > 0) & (scores == 1)]\ntwo = data[(data[\"Lemma_Count\"] > 0) & (scores == 2)]\nsns.distplot(zero[\"Lemma_Count\"], bins=10, color='r')\nsns.distplot(one[\"Lemma_Count\"], bins=10, color='g')\nsns.distplot(two[\"Lemma_Count\"], bins=10, color='b')\nplt.title(\"Score Distribution with respect to lemma count\",fontsize=20)\nplt.xlabel(\"Lemma Count\",fontsize=15)\nplt.ylabel(\"Distribuition of the scores\",fontsize=15)\nplt.show()", "_____no_output_____" ], [ "zero = data[(data[\"Sentence_Count\"] > 0) & (scores == 0)]\none = data[(data[\"Sentence_Count\"] > 0) & (scores == 1)]\ntwo = data[(data[\"Sentence_Count\"] > 0) & (scores == 2)]\nsns.distplot(zero[\"Sentence_Count\"], bins=10, color='r')\nsns.distplot(one[\"Sentence_Count\"], bins=10, color='g')\nsns.distplot(two[\"Sentence_Count\"], bins=10, color='b')\nplt.title(\"Score Distribution with respect to sentence count\",fontsize=20)\nplt.xlabel(\"Sentence Count\",fontsize=15)\nplt.ylabel(\"Distribuition of the scores\",fontsize=15)\nplt.show()", "_____no_output_____" ], [ "zero = data[(data[\"Word_Count\"] > 0) & (scores == 0)]\none = data[(data[\"Word_Count\"] > 0) & (scores == 1)]\ntwo = data[(data[\"Word_Count\"] > 0) & (scores == 2)]\nsns.distplot(zero[\"Word_Count\"], bins=10, color='r')\nsns.distplot(one[\"Word_Count\"], bins=10, color='g')\nsns.distplot(two[\"Word_Count\"], bins=10, color='b')\nplt.title(\"Score Distribution with respect to word count\",fontsize=20)\nplt.xlabel(\"Word_Count\",fontsize=15)\nplt.ylabel(\"Distribuition of the scores\",fontsize=15)\nplt.show()", "_____no_output_____" ], [ "zero = data[(data[\"Average_Word_Length\"] > 0) & (scores == 0)]\none = data[(data[\"Average_Word_Length\"] > 0) & (scores == 1)]\ntwo = data[(data[\"Average_Word_Length\"] > 0) & (scores == 2)]\nsns.distplot(zero[\"Average_Word_Length\"], bins=10, color='r')\nsns.distplot(one[\"Average_Word_Length\"], bins=10, color='g')\nsns.distplot(two[\"Average_Word_Length\"], bins=10, color='b')\nplt.title(\"Score Distribution with respect to Average_Word_Length\",fontsize=20)\nplt.xlabel(\"Average_Word_Length\",fontsize=15)\nplt.ylabel(\"Distribuition of the scores\",fontsize=15)\nplt.show()", "_____no_output_____" ] ], [ [ "### Kappa Score Reliability\nAccording to Cohen's original article, values ≤ 0 as indicating no agreement and 0.01–0.20 as none to slight, 0.21–0.40 as fair, 0.41– 0.60 as moderate, 0.61–0.80 as substantial, and 0.81–1.00 as almost perfect agreement. McHugh says that many texts recommend 80% agreement as the minimum acceptable interrater agreement.", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "# Hello Segmentation\n\nA very basic introduction to using segmentation models with OpenVINO.\n\nWe use the pre-trained [road-segmentation-adas-0001](https://docs.openvinotoolkit.org/latest/omz_models_model_road_segmentation_adas_0001.html) model from the [Open Model Zoo](https://github.com/openvinotoolkit/open_model_zoo/). ADAS stands for Advanced Driver Assistance Services. The model recognizes four classes: background, road, curb and mark.", "_____no_output_____" ], [ "## Imports", "_____no_output_____" ] ], [ [ "import cv2\nimport matplotlib.pyplot as plt\nimport numpy as np\nimport sys\nfrom openvino.inference_engine import IECore\n\nsys.path.append(\"../utils\")\nfrom notebook_utils import segmentation_map_to_image", "_____no_output_____" ] ], [ [ "## Load the Model", "_____no_output_____" ] ], [ [ "ie = IECore()\n\nnet = ie.read_network(\n model=\"model/road-segmentation-adas-0001.xml\")\nexec_net = ie.load_network(net, \"CPU\")\n\noutput_layer_ir = next(iter(exec_net.outputs))\ninput_layer_ir = next(iter(exec_net.input_info))", "_____no_output_____" ] ], [ [ "## Load an Image\nA sample image from the [Mapillary Vistas](https://www.mapillary.com/dataset/vistas) dataset is provided. ", "_____no_output_____" ] ], [ [ "# The segmentation network expects images in BGR format\nimage = cv2.imread(\"data/empty_road_mapillary.jpg\")\n\nrgb_image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)\nimage_h, image_w, _ = image.shape\n\n# N,C,H,W = batch size, number of channels, height, width\nN, C, H, W = net.input_info[input_layer_ir].tensor_desc.dims\n\n# OpenCV resize expects the destination size as (width, height)\nresized_image = cv2.resize(image, (W, H))\n\n# reshape to network input shape\ninput_image = np.expand_dims(\n resized_image.transpose(2, 0, 1), 0\n) \nplt.imshow(rgb_image)", "_____no_output_____" ] ], [ [ "## Do Inference", "_____no_output_____" ] ], [ [ "# Run the infernece\nresult = exec_net.infer(inputs={input_layer_ir: input_image})\nresult_ir = result[output_layer_ir]\n\n# Prepare data for visualization\nsegmentation_mask = np.argmax(result_ir, axis=1)\nplt.imshow(segmentation_mask[0])", "_____no_output_____" ] ], [ [ "## Prepare Data for Visualization", "_____no_output_____" ] ], [ [ "# Define colormap, each color represents a class\ncolormap = np.array([[68, 1, 84], [48, 103, 141], [53, 183, 120], [199, 216, 52]])\n\n# Define the transparency of the segmentation mask on the photo\nalpha = 0.3\n\n# Use function from notebook_utils.py to transform mask to an RGB image\nmask = segmentation_map_to_image(segmentation_mask, colormap)\nresized_mask = cv2.resize(mask, (image_w, image_h))\n\n# Create image with mask put on\nimage_with_mask = cv2.addWeighted(resized_mask, alpha, rgb_image, 1 - alpha, 0)", "_____no_output_____" ] ], [ [ "## Visualize data", "_____no_output_____" ] ], [ [ "# Define titles with images\ndata = {\"Base Photo\": rgb_image, \"Segmentation\": mask, \"Masked Photo\": image_with_mask}\n\n# Create subplot to visualize images\nf, axs = plt.subplots(1, len(data.items()), figsize=(15, 10))\n\n# Fill subplot\nfor ax, (name, image) in zip(axs, data.items()):\n ax.axis('off')\n ax.set_title(name)\n ax.imshow(image)\n\n# Display image\nplt.show(f)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# TD - Implémentation des arbres en POO\n\nOn va dans ce TD créer une classe arbre binaire qui va nous permettre d'implémenter cette structure de données avec toutes ses caractéristiques.\n\nOn se souvient du cours sur les arbres https://pixees.fr/informatiquelycee/n_site/nsi_term_structDo_arbre.html dans lequel on définit le noeud racine, les sous-arbres gauche et droit.", "_____no_output_____" ] ], [ [ "class ArbreBinaire:\n def __init__(self, valeur): #un arbre est un noeud contenant une valeur et deux enfants gauche et droit\n self.valeur = valeur #le noeud contient la valeur passée en paramètre\n self.enfant_gauche = None #l'enfant gauche est vide au départ\n self.enfant_droit = None #l'enfant droit est vide au départ\n\n def insert_gauche(self, valeur): #on insere valeur à la racine du sous-arbre de gauche\n if self.enfant_gauche == None:\n self.enfant_gauche = ArbreBinaire(valeur)\n else:\n new_node = ArbreBinaire(valeur)\n new_node.enfant_gauche = self.enfant_gauche\n self.enfant_gauche = new_node\n\n def insert_droit(self, valeur): #on insere valeur à la racine du sous-arbre de droite\n if self.enfant_droit == None:\n self.enfant_droit = ArbreBinaire(valeur)\n else:\n new_node = ArbreBinaire(valeur)\n new_node.enfant_droit = self.enfant_droit\n self.enfant_droit = new_node\n \n def get_valeur(self): #on renvoit la valeur du noeud racine\n return self.valeur\n\n def get_gauche(self): #on renvoit le sous arbre gauche\n return self.enfant_gauche\n\n def get_droit(self): #on renvoit le sous arbre droit\n return self.enfant_droit\n \n def __str__(self): #affichage d'un arbre avec print()\n if self!= None: #on écrit (enfantgauche)valeur(enfantdroit)\n return \"(\"+self.enfant_gauche.__str__()+\")\"+str(self.valeur)+\"(\"+self.enfant_droit.__str__()+\")\"\n else:\n return \"\"", "_____no_output_____" ] ], [ [ "Il est important de comprendre comment la structure de donnée est implémentée : un arbre est défini par une valeur et deux sous arbres gauches et droits qui sont tous les deux initialiés à <code>None</code> au démarrage.\n\nLors de l'insertion d'une valeur à gauche ou à droite, on crée un nouvel arbre que l'on insère à la racine du sous-arbre de gauche (dans le cas de <code>insert_gauche()</code>) ou droite (dans le cas de <code>insert_droit()</code>)\n\n## Exercice 1\nSur papier, dessinez ce qui se passe étape par étape quand vous éxécutez le code suivant :", "_____no_output_____" ] ], [ [ "arbre=ArbreBinaire(3)\narbre.insert_gauche(1)\narbre.insert_gauche(2) #on insère 2 à gauche, la valeur 1 se décale à gauche du 2\nprint(arbre)", "(((None)1(None))2(None))3(None)\n" ] ], [ [ "## Exercice 2\nÉcrire le code permettant de créer l'arbre suivant\n<img src=\"nsi_term_algo_arbre_1.png\">\n\nVoila le début du code ci-dessous pour vous aider à démarrer...", "_____no_output_____" ] ], [ [ "arbre=ArbreBinaire(\"A\")\narbre.insert_gauche(\"B\")\narbreg=arbre.get_gauche()\narbreg.insert_gauche(\"C\") #les deux lignes précédentes peuvent-être abbrégées en arbre.get_gauche().insert_gauche(\"C\")\n# à suivre...\nprint(arbre)", "(((None)C(None))B(None))A(None)\n" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "from __future__ import print_function\n\nimport math\n\nfrom IPython import display\nfrom matplotlib import cm\nfrom matplotlib import gridspec\nfrom matplotlib import pyplot as plt\nimport numpy as np\nimport pandas as pd\nfrom sklearn import metrics\nimport tensorflow as tf\nfrom tensorflow.python.data import Dataset\n\ntf.logging.set_verbosity(tf.logging.ERROR)\npd.options.display.max_rows = 10\npd.options.display.float_format = '{:.1f}'.format", "/home/jimut/anaconda3/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n from ._conv import register_converters as _register_converters\n" ], [ "ls\n", "LUND.ipynb README.md Test1_Regressor.ipynb\r\nPandas.ipynb Regressor.ipynb test.csv\r\nprices.csv reinforcement_q_learning.ipynb Untitled.ipynb\r\nprices_CSV_REG.ipynb Stock_Kaggle_1.ipynb WIKI-PRICES.csv\r\n" ], [ "dataset_all = pd.read_csv('prices.csv')", "_____no_output_____" ], [ "dataset_all.head()", "_____no_output_____" ], [ "dataset_all", "_____no_output_____" ], [ "def my_input_fn(features, targets, batch_size=1, shuffle=True, num_epochs=None):\n \"\"\"Trains a linear regression model of one feature.\n \n Args:\n features: pandas DataFrame of features\n targets: pandas DataFrame of targets\n batch_size: Size of batches to be passed to the model\n shuffle: True or False. Whether to shuffle the data.\n num_epochs: Number of epochs for which data should be repeated. None = repeat indefinitely\n Returns:\n Tuple of (features, labels) for next data batch\n \"\"\"\n \n # Convert pandas data into a dict of np arrays.\n features = {key:np.array(value) for key,value in dict(features).items()} \n \n # Construct a dataset, and configure batching/repeating.\n ds = Dataset.from_tensor_slices((features,targets)) # warning: 2GB limit\n ds = ds.batch(batch_size).repeat(num_epochs)\n \n # Shuffle the data, if specified.\n if shuffle:\n ds = ds.shuffle(buffer_size=10000)\n \n # Return the next batch of data.\n features, labels = ds.make_one_shot_iterator().get_next()\n return features, labels", "_____no_output_____" ], [ "def train_model(learning_rate, steps, batch_size, input_feature=\"close\"):\n \"\"\"Trains a linear regression model of one feature.\n \n Args:\n learning_rate: A `float`, the learning rate.\n steps: A non-zero `int`, the total number of training steps. A training step\n consists of a forward and backward pass using a single batch.\n batch_size: A non-zero `int`, the batch size.\n input_feature: A `string` specifying a column from `california_housing_dataframe`\n to use as input feature.\n \"\"\"\n \n periods = 10\n steps_per_period = steps / periods\n\n my_feature = input_feature\n my_feature_data = dataset_all[[my_feature]]\n my_label = \"open\"\n targets = dataset_all[my_label]\n\n # Create feature columns.\n feature_columns = [tf.feature_column.numeric_column(my_feature)]\n \n # Create input functions.\n training_input_fn = lambda:my_input_fn(my_feature_data, targets, batch_size=batch_size)\n prediction_input_fn = lambda: my_input_fn(my_feature_data, targets, num_epochs=1, shuffle=False)\n \n # Create a linear regressor object.\n my_optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)\n my_optimizer = tf.contrib.estimator.clip_gradients_by_norm(my_optimizer, 5.0)\n linear_regressor = tf.estimator.LinearRegressor(\n feature_columns=feature_columns,\n optimizer=my_optimizer\n )\n\n # Set up to plot the state of our model's line each period.\n plt.figure(figsize=(15, 6))\n plt.subplot(1, 2, 1)\n plt.title(\"Learned Line by Period\")\n plt.ylabel(my_label)\n plt.xlabel(my_feature)\n sample = dataset_all.sample(n=300)\n plt.scatter(sample[my_feature], sample[my_label])\n colors = [cm.coolwarm(x) for x in np.linspace(-1, 1, periods)]\n\n # Train the model, but do so inside a loop so that we can periodically assess\n # loss metrics.\n print(\"Training model...\")\n print(\"RMSE (on training data):\")\n root_mean_squared_errors = []\n for period in range (0, periods):\n # Train the model, starting from the prior state.\n linear_regressor.train(\n input_fn=training_input_fn,\n steps=steps_per_period\n )\n # Take a break and compute predictions.\n predictions = linear_regressor.predict(input_fn=prediction_input_fn)\n predictions = np.array([item['predictions'][0] for item in predictions])\n \n # Compute loss.\n root_mean_squared_error = math.sqrt(\n metrics.mean_squared_error(predictions, targets))\n # Occasionally print the current loss.\n print(\" period %02d : %0.2f\" % (period, root_mean_squared_error))\n # Add the loss metrics from this period to our list.\n root_mean_squared_errors.append(root_mean_squared_error)\n # Finally, track the weights and biases over time.\n # Apply some math to ensure that the data and line are plotted neatly.\n y_extents = np.array([0, sample[my_label].max()])\n \n weight = linear_regressor.get_variable_value('linear/linear_model/%s/weights' % input_feature)[0]\n bias = linear_regressor.get_variable_value('linear/linear_model/bias_weights')\n\n x_extents = (y_extents - bias) / weight\n x_extents = np.maximum(np.minimum(x_extents,\n sample[my_feature].max()),\n sample[my_feature].min())\n y_extents = weight * x_extents + bias\n plt.plot(x_extents, y_extents, color=colors[period]) \n print(\"Model training finished.\")\n\n # Output a graph of loss metrics over periods.\n plt.subplot(1, 2, 2)\n plt.ylabel('RMSE')\n plt.xlabel('Periods')\n plt.title(\"Root Mean Squared Error vs. Periods\")\n plt.tight_layout()\n plt.plot(root_mean_squared_errors)\n\n # Output a table with calibration data.\n calibration_data = pd.DataFrame()\n calibration_data[\"predictions\"] = pd.Series(predictions)\n calibration_data[\"targets\"] = pd.Series(targets)\n display.display(calibration_data.describe())\n\n print(\"Final RMSE (on training data): %0.2f\" % root_mean_squared_error)", "_____no_output_____" ], [ "train_model(\n learning_rate=0.002,\n steps=100,\n batch_size=10,\n input_feature=\"close\"\n)", "Training model...\nRMSE (on training data):\n period 00 : 98.68\n period 01 : 87.72\n period 02 : 76.76\n period 03 : 65.80\n period 04 : 54.83\n period 05 : 43.87\n period 06 : 32.92\n period 07 : 21.97\n period 08 : 11.05\n period 09 : 2.62\nModel training finished.\n" ], [ "def train_model(learning_rate, steps, batch_size, input_feature=\"close\"):\n \"\"\"Trains a linear regression model of one feature.\n \n Args:\n learning_rate: A `float`, the learning rate.\n steps: A non-zero `int`, the total number of training steps. A training step\n consists of a forward and backward pass using a single batch.\n batch_size: A non-zero `int`, the batch size.\n input_feature: A `string` specifying a column from `california_housing_dataframe`\n to use as input feature.\n \"\"\"\n \n periods = 10\n steps_per_period = steps / periods\n\n my_feature = input_feature\n my_feature_data = dataset_all[[my_feature]]\n my_label = \"open\"\n targets = dataset_all[my_label]\n\n # Create feature columns.\n feature_columns = [tf.feature_column.numeric_column(my_feature)]\n \n # Create input functions.\n training_input_fn = lambda:my_input_fn(my_feature_data, targets, batch_size=batch_size)\n prediction_input_fn = lambda: my_input_fn(my_feature_data, targets, num_epochs=1, shuffle=False)\n \n # Create a linear regressor object.\n my_optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate)\n my_optimizer = tf.contrib.estimator.clip_gradients_by_norm(my_optimizer, 5.0)\n linear_regressor = tf.estimator.LinearRegressor(\n feature_columns=feature_columns,\n optimizer=my_optimizer\n )\n\n # Set up to plot the state of our model's line each period.\n plt.figure(figsize=(15, 6))\n plt.subplot(1, 2, 1)\n plt.title(\"Learned Line by Period\")\n plt.ylabel(my_label)\n plt.xlabel(my_feature)\n sample = dataset_all.sample(n=300)\n plt.scatter(sample[my_feature], sample[my_label])\n colors = [cm.coolwarm(x) for x in np.linspace(-1, 1, periods)]\n\n # Train the model, but do so inside a loop so that we can periodically assess\n # loss metrics.\n print(\"Training model...\")\n print(\"RMSE (on training data):\")\n root_mean_squared_errors = []\n for period in range (0, periods):\n # Train the model, starting from the prior state.\n linear_regressor.train(\n input_fn=training_input_fn,\n steps=steps_per_period\n )\n # Take a break and compute predictions.\n predictions = linear_regressor.predict(input_fn=prediction_input_fn)\n predictions = np.array([item['predictions'][0] for item in predictions])\n \n # Compute loss.\n root_mean_squared_error = math.sqrt(\n metrics.mean_squared_error(predictions, targets))\n # Occasionally print the current loss.\n print(\" period %02d : %0.2f\" % (period, root_mean_squared_error))\n # Add the loss metrics from this period to our list.\n root_mean_squared_errors.append(root_mean_squared_error)\n # Finally, track the weights and biases over time.\n # Apply some math to ensure that the data and line are plotted neatly.\n y_extents = np.array([0, sample[my_label].max()])\n \n weight = linear_regressor.get_variable_value('linear/linear_model/%s/weights' % input_feature)[0]\n bias = linear_regressor.get_variable_value('linear/linear_model/bias_weights')\n\n x_extents = (y_extents - bias) / weight\n x_extents = np.maximum(np.minimum(x_extents,\n sample[my_feature].max()),\n sample[my_feature].min())\n y_extents = weight * x_extents + bias\n plt.plot(x_extents, y_extents, color=colors[period]) \n print(\"Model training finished.\")\n\n # Output a graph of loss metrics over periods.\n plt.subplot(1, 2, 2)\n plt.ylabel('RMSE')\n plt.xlabel('Periods')\n plt.title(\"Root Mean Squared Error vs. Periods\")\n plt.tight_layout()\n plt.plot(root_mean_squared_errors)\n\n # Output a table with calibration data.\n calibration_data = pd.DataFrame()\n calibration_data[\"predictions\"] = pd.Series(predictions)\n calibration_data[\"targets\"] = pd.Series(targets)\n display.display(calibration_data.describe())\n\n print(\"Final RMSE (on training data): %0.2f\" % root_mean_squared_error)", "_____no_output_____" ], [ "train_model(\n learning_rate=0.002,\n steps=100,\n batch_size=10,\n input_feature=\"high\"\n)", "Training model...\nRMSE (on training data):\n period 00 : 98.58\n period 01 : 87.51\n period 02 : 76.45\n period 03 : 65.38\n period 04 : 54.31\n period 05 : 43.24\n period 06 : 32.18\n period 07 : 21.12\n period 08 : 10.08\n period 09 : 1.53\nModel training finished.\n" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "## Chapter 7 - Exercise 3: Tips", "_____no_output_____" ], [ "#### Cho dữ liệu tips có sẵn trong seaborn library. Hãy vẽ những biểu đồ theo yêu cầu, và đối chiếu với kết quả:", "_____no_output_____" ] ], [ [ "import numpy as np\nimport pandas as pd\nfrom matplotlib import pyplot as plt\nimport seaborn as sns", "_____no_output_____" ], [ "# Load dữ liệu tips có sẵn trong seaborn library\n#total_bill: Total bill (cost of the meal), including tax, in US dollars\n#tip: Tip (gratuity) in US dollars\n#sex: Sex of person paying for the meal (0=male, 1=female)\n#smoker: Smoker in party? (0=No, 1=Yes)\n#day: 3=Thur, 4=Fri, 5=Sat, 6=Sun\n#time: 0=Day, 1=Night\n#size: Size of the party\n\n", "_____no_output_____" ] ], [ [ "<details>\n <summary>Nhấn vào đây để xem kết quả !</summary>\n \n<pre>&lt;class 'pandas.core.frame.DataFrame'&gt;\nRangeIndex: 244 entries, 0 to 243\nData columns (total 7 columns):\ntotal_bill 244 non-null float64\ntip 244 non-null float64\nsex 244 non-null category\nsmoker 244 non-null category\nday 244 non-null category\ntime 244 non-null category\nsize 244 non-null int64\ndtypes: category(4), float64(2), int64(1)\nmemory usage: 7.3 KB\n</pre>\n\n<div>\n<style scoped=\"\">\n .dataframe tbody tr th:only-of-type {\n vertical-align: middle;\n }\n\n .dataframe tbody tr th {\n vertical-align: top;\n }\n\n .dataframe thead th {\n text-align: right;\n }\n</style>\n<table border=\"1\" class=\"dataframe\">\n <thead>\n <tr style=\"text-align: right;\">\n <th></th>\n <th>total_bill</th>\n <th>tip</th>\n <th>sex</th>\n <th>smoker</th>\n <th>day</th>\n <th>time</th>\n <th>size</th>\n </tr>\n </thead>\n <tbody>\n <tr>\n <th>0</th>\n <td>16.99</td>\n <td>1.01</td>\n <td>Female</td>\n <td>No</td>\n <td>Sun</td>\n <td>Dinner</td>\n <td>2</td>\n </tr>\n <tr>\n <th>1</th>\n <td>10.34</td>\n <td>1.66</td>\n <td>Male</td>\n <td>No</td>\n <td>Sun</td>\n <td>Dinner</td>\n <td>3</td>\n </tr>\n <tr>\n <th>2</th>\n <td>21.01</td>\n <td>3.50</td>\n <td>Male</td>\n <td>No</td>\n <td>Sun</td>\n <td>Dinner</td>\n <td>3</td>\n </tr>\n <tr>\n <th>3</th>\n <td>23.68</td>\n <td>3.31</td>\n <td>Male</td>\n <td>No</td>\n <td>Sun</td>\n <td>Dinner</td>\n <td>2</td>\n </tr>\n <tr>\n <th>4</th>\n <td>24.59</td>\n <td>3.61</td>\n <td>Female</td>\n <td>No</td>\n <td>Sun</td>\n <td>Dinner</td>\n <td>4</td>\n </tr>\n </tbody>\n</table>\n</div>\n\n</details>", "_____no_output_____" ] ], [ [ "# Câu 1: Vẽ violinplot cho cho cột total_bill\n# Bạn nhận xét gì về biểu đồ vừa tạo\n", "_____no_output_____" ] ], [ [ "<details>\n <summary>Nhấn vào đây để xem kết quả !</summary>\n \n<div class=\"output_subarea output_png\"><img src=\"data:image/png;base64,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\n\"></div>\n\n</details>", "_____no_output_____" ] ], [ [ "# Câu 2: Vẽ swarmplot cho cột total_bill theo sex\n# Bạn nhận xét gì về biểu đồ vừa tạo\n", "_____no_output_____" ] ], [ [ "<details>\n <summary>Nhấn vào đây để xem kết quả !</summary>\n \n<div class=\"output_subarea output_png\"><img src=\"data:image/png;base64,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\n\"></div>\n\n</details>", "_____no_output_____" ] ], [ [ "# Câu 3: Vẽ boxplot cho cột total_bill\n# Bạn nhận xét gì về biểu đồ vừa tạo\n", "_____no_output_____" ] ], [ [ "<details>\n <summary>Nhấn vào đây để xem kết quả !</summary>\n \n<div class=\"output_subarea output_png\"><img src=\"data:image/png;base64,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\n\"></div>\n\n</details>", "_____no_output_____" ] ], [ [ "# Câu 4: Tạo FacetGrid với 'time' và chỉ định thứ tự của các hàng bằng row_order, ánh xạ (map) của 'total_bill' lên lưới\n# Bạn nhận xét gì về biểu đồ vừa tạo\n", "_____no_output_____" ] ], [ [ "<details>\n <summary>Nhấn vào đây để xem kết quả !</summary>\n \n<img src=\"data:image/png;base64,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\n\"></div></div><div class=\"output_area\"><div class=\"run_this_cell\"></div><div class=\"prompt\"></div><div class=\"output_subarea output_text\"><pre>&lt;Figure size 432x288 with 0 Axes&gt;</pre></div></div></div>\n\n</details>", "_____no_output_____" ] ], [ [ "# Câu 5: Tạo Factor plot (phiên bản mới là catplot) chứa point plot của giá trị 'total_bill'\n# Bạn nhận xét gì về biểu đồ vừa tạo\n", "_____no_output_____" ] ], [ [ "<details>\n <summary>Nhấn vào đây để xem kết quả !</summary>\n <div class=\"output_subarea output_png\"><img src=\"data:image/png;base64,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\n\"></div>\n</details> ", "_____no_output_____" ] ], [ [ "# Câu 6: Tạo PairGrid với một scatter plot \"total_bill\" và \"tip\"\n# Bạn nhận xét gì về biểu đồ vừa tạo\n", "_____no_output_____" ] ], [ [ "<details>\n <summary>Nhấn vào đây để xem kết quả !</summary>\n \n<div class=\"output_subarea output_png\"><img src=\"data:image/png;base64,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\n\"></div>\n\n</details>", "_____no_output_____" ] ], [ [ "# Câu 7: Tạo Pairplot với một scatter plot \"total_bill\" và \"tip\", sử dụng palette color = 'day'\n# Bạn nhận xét gì về biểu đồ vừa tạo\n", "_____no_output_____" ] ], [ [ "<details>\n <summary>Nhấn vào đây để xem kết quả !</summary>\n \n<div class=\"output_subarea output_png\"><img src=\"data:image/png;base64,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\n\"></div>\n\n</details>", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "# Linear_Reg", "_____no_output_____" ], [ "Author ~ Saurabh Kumar \nDate ~ 05-Dec-21", "_____no_output_____" ] ], [ [ "import numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport seaborn as sns\n\nimport warnings\nwarnings.filterwarnings(\"ignore\")", "_____no_output_____" ], [ "#simple_linera_regression\n\nclass Simple_linear_regression:\n \n def __init__(self,learning_rate=1e-3,n_steps=1000):\n self.learning_rate =learning_rate\n self.n_steps =n_steps\n \n def fit(self,X,y):\n \n # adding the bias term\n X_train = np.c_[np.ones(X.shape[0]),X]\n \n # random initialization of the model weights\n self.W =np.random.rand((X_train.shape[1]))\n\n # random initialization of the model weights\n for i in range(self.n_steps):\n self.W =self.W -self.learning_rate*self.cal_gradiant_descent(X_train,y)\n\n def cal_gradiant_descent(self,X,y):\n \n #calculating gradiant descent\n \n return 2/X.shape[0] * np.dot(X.T,np.dot(X,self.W)-y)\n \n def predict(self,X):\n \n #Predicting Y for the X\n \n #adding bias term\n \n X_pred =np.c_[np.ones(X.shape[0]),X]\n \n return np.dot(X_pred,self.W)\n", "_____no_output_____" ], [ "#creating dataset\nfrom sklearn.datasets import make_regression \nX , y = make_regression (n_samples=1000,n_features = 1,n_targets=1,bias =2.5,noise=40,random_state = 44)\nprint(\"X_shape =\",X.shape)\nprint(\"y_shape =\",y.shape)", "X_shape = (1000, 1)\ny_shape = (1000,)\n" ], [ "#train_test_split\nfrom sklearn.model_selection import train_test_split\nX_train,X_test,y_train,y_test =train_test_split(X,y,test_size=.33,random_state=12)", "_____no_output_____" ], [ "print(\"Shape X_train :\",X_train.shape)\nprint(\"Shape y_train :\",y_train.shape)\nprint(\"Shape X_test :\",X_test.shape)\nprint(\"Shape y_test :\",y_test.shape)", "Shape X_train : (670, 1)\nShape y_train : (670,)\nShape X_test : (330, 1)\nShape y_test : (330,)\n" ], [ "%matplotlib inline\nimport matplotlib.pyplot as plt\n\nplt.xlabel('X_train')\nplt.ylabel('Y_train')\nplt.title('Relationship between X_train and Y_train variables')\nplt.scatter(X_train, y_train)", "_____no_output_____" ], [ "plt.xlabel('X_train')\nplt.ylabel('Y_train')\nplt.title('Relationship between X_train and Y_train variables')\nsns.regplot(X_train, y_train)", "_____no_output_____" ], [ "#model\nmodel = Simple_linear_regression()\nmodel.fit(X_train,y_train)", "_____no_output_____" ], [ "#prediction\ny_pred =model.predict(X_test)", "_____no_output_____" ], [ "#error\nprint(\"Mean squared error: %.2f\" % np.mean((model.predict(X_test) - y_test) ** 2))", "Mean squared error: 1716.39\n" ], [ "plt.xlabel('X_test')\nplt.ylabel('Y')\nplt.title('Real vs Predicted values comparison')\n\nplt.scatter(X_test, y_test)\nplt.scatter(X_test, y_pred)", "_____no_output_____" ], [ "#Same with Sklearn lib\nfrom sklearn.linear_model import LinearRegression\nmodelSk =LinearRegression()\nmodelSk.fit(X_train,y_train)", "_____no_output_____" ], [ "y_predict=modelSk.predict(X_test)", "_____no_output_____" ], [ "#error\nprint(\"Mean squared error: %.2f\" % np.mean((modelSk.predict(X_test) - y_test) ** 2))", "Mean squared error: 1534.17\n" ], [ "def accuracy(X_test,y_test, y_pred):\n print('accuracy (R^2):\\n', modelSk.score(X_test, y_test)*100, '%')", "_____no_output_____" ], [ "accuracy(X_test,y_test,y_predict)", "accuracy (R^2):\n 79.03370956723134 %\n" ], [ "plt.xlabel('X_test')\nplt.ylabel('Y')\nplt.title('Real vs Predicted values comparison')\n\nplt.scatter(X_test, y_test)\nplt.scatter(X_test, y_predict)", "_____no_output_____" ] ] ]

[ "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# Collaboration and Competition\n\n---\n\nIn this notebook, you will learn how to use the Unity ML-Agents environment for the third project of the [Deep Reinforcement Learning Nanodegree](https://www.udacity.com/course/deep-reinforcement-learning-nanodegree--nd893) program.\n\n### 1. Start the Environment\n\nWe begin by importing the necessary packages. If the code cell below returns an error, please revisit the project instructions to double-check that you have installed [Unity ML-Agents](https://github.com/Unity-Technologies/ml-agents/blob/master/docs/Installation.md) and [NumPy](http://www.numpy.org/).", "_____no_output_____" ] ], [ [ "from unityagents import UnityEnvironment\nimport numpy as np\nimport random\nimport torch\nfrom collections import deque\nimport matplotlib.pyplot as plt\n%matplotlib inline\n\nfrom ddpg_agent import Agent\n\nplt.style.use('fivethirtyeight')\n\n%load_ext autoreload\n%autoreload 2\n\nimport warnings\nwarnings.filterwarnings('ignore')", "_____no_output_____" ] ], [ [ "Next, we will start the environment! **_Before running the code cell below_**, change the `file_name` parameter to match the location of the Unity environment that you downloaded.\n\n- **Mac**: `\"path/to/Tennis.app\"`\n- **Windows** (x86): `\"path/to/Tennis_Windows_x86/Tennis.exe\"`\n- **Windows** (x86_64): `\"path/to/Tennis_Windows_x86_64/Tennis.exe\"`\n- **Linux** (x86): `\"path/to/Tennis_Linux/Tennis.x86\"`\n- **Linux** (x86_64): `\"path/to/Tennis_Linux/Tennis.x86_64\"`\n- **Linux** (x86, headless): `\"path/to/Tennis_Linux_NoVis/Tennis.x86\"`\n- **Linux** (x86_64, headless): `\"path/to/Tennis_Linux_NoVis/Tennis.x86_64\"`\n\nFor instance, if you are using a Mac, then you downloaded `Tennis.app`. If this file is in the same folder as the notebook, then the line below should appear as follows:\n```\nenv = UnityEnvironment(file_name=\"Tennis.app\")\n```", "_____no_output_____" ] ], [ [ "# Under Linux\n#env = UnityEnvironment(file_name='Tennis_Linux/Tennis.x86_64', seed=123)\n\n# Under Mac\nenv = UnityEnvironment(file_name='Tennis.app', seed=123)", "INFO:unityagents:\n'Academy' started successfully!\nUnity Academy name: Academy\n Number of Brains: 1\n Number of External Brains : 1\n Lesson number : 0\n Reset Parameters :\n\t\t\nUnity brain name: TennisBrain\n Number of Visual Observations (per agent): 0\n Vector Observation space type: continuous\n Vector Observation space size (per agent): 8\n Number of stacked Vector Observation: 3\n Vector Action space type: continuous\n Vector Action space size (per agent): 2\n Vector Action descriptions: , \n" ] ], [ [ "Environments contain **_brains_** which are responsible for deciding the actions of their associated agents. Here we check for the first brain available, and set it as the default brain we will be controlling from Python.", "_____no_output_____" ] ], [ [ "# get the default brain\nbrain_name = env.brain_names[0]\nbrain = env.brains[brain_name]", "_____no_output_____" ] ], [ [ "### 2. Examine the State and Action Spaces\n\nIn this environment, two agents control rackets to bounce a ball over a net. If an agent hits the ball over the net, it receives a reward of +0.1. If an agent lets a ball hit the ground or hits the ball out of bounds, it receives a reward of -0.01. Thus, the goal of each agent is to keep the ball in play.\n\nThe observation space consists of 8 variables corresponding to the position and velocity of the ball and racket. Two continuous actions are available, corresponding to movement toward (or away from) the net, and jumping. \n\nRun the code cell below to print some information about the environment.", "_____no_output_____" ] ], [ [ "# reset the environment\nenv_info = env.reset(train_mode=True)[brain_name]\n\n# number of agents \nnum_agents = len(env_info.agents)\nprint('Number of agents:', num_agents)\n\n# size of each action\naction_size = brain.vector_action_space_size\nprint('Size of each action:', action_size)\n\n# examine the state space \nstates = env_info.vector_observations\nstate_size = states.shape[1]\nprint('There are {} agents. Each observes a state with length: {}'.format(states.shape[0], state_size))\nprint('The state for the first agent looks like:', states[0])", "Number of agents: 2\nSize of each action: 2\nThere are 2 agents. Each observes a state with length: 24\nThe state for the first agent looks like: [ 0. 0. 0. 0. 0. 0.\n 0. 0. 0. 0. 0. 0.\n 0. 0. 0. 0. -6.51024199 -1.5\n -0. 0. 6.80061245 6. -0. 0. ]\n" ] ], [ [ "### 3. Take Random Actions in the Environment\n\nIn the next code cell, you will learn how to use the Python API to control the agents and receive feedback from the environment.\n\nOnce this cell is executed, you will watch the agents' performance, if they select actions at random with each time step. A window should pop up that allows you to observe the agents.\n\nOf course, as part of the project, you'll have to change the code so that the agents are able to use their experiences to gradually choose better actions when interacting with the environment!", "_____no_output_____" ] ], [ [ "for i in range(1, 20): # play game for 5 episodes\n env_info = env.reset(train_mode=False)[brain_name] # reset the environment \n states = env_info.vector_observations # get the current state (for each agent)\n scores = np.zeros(num_agents) # initialize the score (for each agent)\n while True:\n actions = np.random.randn(num_agents, action_size) # select an action (for each agent)\n actions = np.clip(actions, -1, 1) # all actions between -1 and 1\n env_info = env.step(actions)[brain_name] # send all actions to tne environment\n next_states = env_info.vector_observations # get next state (for each agent)\n rewards = env_info.rewards # get reward (for each agent)\n dones = env_info.local_done # see if episode finished\n scores += env_info.rewards # update the score (for each agent)\n states = next_states # roll over states to next time step\n if np.any(dones): # exit loop if episode finished\n break\n print('Score (max over agents) from episode {}: {}'.format(i, np.max(scores)))", "Score (max over agents) from episode 1: 0.0\nScore (max over agents) from episode 2: 0.10000000149011612\nScore (max over agents) from episode 3: 0.10000000149011612\nScore (max over agents) from episode 4: 0.10000000149011612\nScore (max over agents) from episode 5: 0.0\nScore (max over agents) from episode 6: 0.0\nScore (max over agents) from episode 7: 0.20000000298023224\nScore (max over agents) from episode 8: 0.0\nScore (max over agents) from episode 9: 0.0\nScore (max over agents) from episode 10: 0.0\nScore (max over agents) from episode 11: 0.0\nScore (max over agents) from episode 12: 0.0\nScore (max over agents) from episode 13: 0.0\nScore (max over agents) from episode 14: 0.0\nScore (max over agents) from episode 15: 0.0\nScore (max over agents) from episode 16: 0.0\nScore (max over agents) from episode 17: 0.0\nScore (max over agents) from episode 18: 0.0\nScore (max over agents) from episode 19: 0.0\n" ] ], [ [ "When finished, you can close the environment.", "_____no_output_____" ], [ "env.close()", "_____no_output_____" ], [ "### 4. It's Your Turn!\n\nNow it's your turn to train your own agent to solve the environment! When training the environment, set `train_mode=True`, so that the line for resetting the environment looks like the following:\n```python\nenv_info = env.reset(train_mode=True)[brain_name]\n```", "_____no_output_____" ], [ "### 5. Competing multi-agents as multiple non-competing agents", "_____no_output_____" ] ], [ [ "# Utilities functions\ndef unity_step_wrap(actions):\n \"\"\"Unity Environment action wrapper\n \n Params\n ======\n action (int): action to take\n \n Return\n ======\n OpenAI-like action outcome (tuple): bundled (next_state, reward, done)\n \"\"\"\n env_info = env.step(actions)[brain_name] # send the action to the environment\n next_states = env_info.vector_observations # get the next state\n rewards = env_info.rewards # get the reward\n dones = env_info.local_done # see if episode has finished\n return (next_states, rewards, dones)\n\ndef moving_average(a, n=3) :\n ret = np.cumsum(a, dtype=float)\n ret[n:] = ret[n:] - ret[:-n]\n return ret[n - 1:] / n\n\ndef plot_scores(scores, smooth_window=100):\n scores_smoothed = moving_average(scores, smooth_window)\n # plot the scores\n fig = plt.figure(figsize=(8, 6))\n ax = fig.add_subplot(111)\n plt.plot(np.arange(len(scores)), scores, linewidth=1, alpha=0.4, color='steelblue')\n plt.plot(np.arange(len(scores))[smooth_window-1:,], scores_smoothed, linewidth=1.5, alpha=1, color='firebrick')\n plt.ylabel('Score')\n plt.xlabel('Episode #')\n plt.show()\n return fig\n\ndef load_checkpoint(filepath):\n checkpoint = torch.load(filepath)\n return (checkpoint['actor'], checkpoint['critic'], checkpoint['scores'])", "_____no_output_____" ], [ "agent = Agent(state_size=state_size, action_size=action_size, random_seed=2)", "_____no_output_____" ], [ "def ddpg(n_episodes=1000, max_t=1000, print_every=100, learn_every=5, min_noise=0.02, solved_score = 0.5, name=\"default\"):\n scores_deque = deque(maxlen=print_every)\n scores = []\n noise = 1.0\n min_noise = min_noise\n noise_reduction = min_noise**(1/n_episodes) # Reaches min_noise after n_episodes with exponential decrease\n \n for i_episode in range(1, n_episodes+1):\n env_info = env.reset(train_mode=True)[brain_name]\n states = env_info.vector_observations \n \n agent.reset()\n scores_episode = np.zeros(num_agents)\n noise *= noise_reduction\n \n for t in range(max_t):\n actions = agent.act(states, noise=max(noise, min_noise))\n next_states, rewards, dones = unity_step_wrap(actions)\n \n for state, action, reward, next_state, done in zip(states, actions, rewards, next_states, dones):\n agent.save_experience(state, action, reward, next_state, done)\n \n scores_episode += rewards\n states = next_states \n \n if t % learn_every == 0:\n agent.step()\n \n if np.any(dones):\n break \n \n score = np.max(scores_episode)\n scores_deque.append(score)\n scores.append(score)\n \n print('\\rEpisode {}\\tAverage Score: {:.2f}'.format(i_episode, np.mean(scores_deque)), end=\"\")\n \n # Save checkpoint\n checkpoint = {\n 'actor': agent.actor_local.state_dict(),\n 'critic': agent.critic_local.state_dict(),\n 'scores': scores\n }\n torch.save(checkpoint, 'saved-checkpoints/checkpoint-' + name + '.pth')\n \n if i_episode % print_every == 0:\n print('\\rEpisode {}\\tAverage Score: {:.2f}'.format(i_episode, np.mean(scores_deque)))\n \n if np.mean(scores_deque) > solved_score:\n break\n \n return scores", "_____no_output_____" ] ], [ [ "* **Solving the environment**", "_____no_output_____" ] ], [ [ "scores = ddpg(n_episodes=10000, min_noise=0.02, learn_every=5, name='solved')", "Episode 100\tAverage Score: 0.03\nEpisode 200\tAverage Score: 0.03\nEpisode 300\tAverage Score: 0.05\nEpisode 400\tAverage Score: 0.07\nEpisode 500\tAverage Score: 0.04\nEpisode 600\tAverage Score: 0.06\nEpisode 700\tAverage Score: 0.08\nEpisode 800\tAverage Score: 0.08\nEpisode 900\tAverage Score: 0.06\nEpisode 1000\tAverage Score: 0.11\nEpisode 1100\tAverage Score: 0.13\nEpisode 1180\tAverage Score: 0.52" ], [ "_, _, scores_solved = load_checkpoint('saved-checkpoints/checkpoint-solved.pth')\nfigure = plot_scores(scores_solved)", "_____no_output_____" ] ], [ [ "* **Monitoring beyond solving scores in the long run**", "_____no_output_____" ] ], [ [ "scores = ddpg(n_episodes=10000, min_noise=0.01, learn_every=5, solved_score = 99, name='in-the-long-run')", "Episode 100\tAverage Score: 0.02\nEpisode 200\tAverage Score: 0.01\nEpisode 300\tAverage Score: 0.03\nEpisode 400\tAverage Score: 0.05\nEpisode 500\tAverage Score: 0.04\nEpisode 600\tAverage Score: 0.06\nEpisode 700\tAverage Score: 0.06\nEpisode 800\tAverage Score: 0.08\nEpisode 900\tAverage Score: 0.07\nEpisode 1000\tAverage Score: 0.13\nEpisode 1100\tAverage Score: 0.17\nEpisode 1200\tAverage Score: 0.71\nEpisode 1300\tAverage Score: 0.54\nEpisode 1400\tAverage Score: 0.57\nEpisode 1500\tAverage Score: 0.25\nEpisode 1600\tAverage Score: 0.27\nEpisode 1700\tAverage Score: 0.27\nEpisode 1800\tAverage Score: 0.24\nEpisode 1900\tAverage Score: 0.19\nEpisode 2000\tAverage Score: 0.12\nEpisode 2100\tAverage Score: 0.14\nEpisode 2200\tAverage Score: 0.12\nEpisode 2300\tAverage Score: 0.07\nEpisode 2400\tAverage Score: 0.07\nEpisode 2500\tAverage Score: 0.06\nEpisode 2600\tAverage Score: 0.06\nEpisode 2700\tAverage Score: 0.04\nEpisode 2800\tAverage Score: 0.04\nEpisode 2900\tAverage Score: 0.05\nEpisode 3000\tAverage Score: 0.04\nEpisode 3100\tAverage Score: 0.03\nEpisode 3200\tAverage Score: 0.03\nEpisode 3300\tAverage Score: 0.03\nEpisode 3400\tAverage Score: 0.02\nEpisode 3500\tAverage Score: 0.02\nEpisode 3600\tAverage Score: 0.02\nEpisode 3700\tAverage Score: 0.02\nEpisode 3800\tAverage Score: 0.03\nEpisode 3900\tAverage Score: 0.04\nEpisode 4000\tAverage Score: 0.01\nEpisode 4100\tAverage Score: 0.01\nEpisode 4200\tAverage Score: 0.01\nEpisode 4300\tAverage Score: 0.02\nEpisode 4400\tAverage Score: 0.02\nEpisode 4500\tAverage Score: 0.03\nEpisode 4600\tAverage Score: 0.02\nEpisode 4700\tAverage Score: 0.03\nEpisode 4800\tAverage Score: 0.04\nEpisode 4900\tAverage Score: 0.01\nEpisode 5000\tAverage Score: 0.02\nEpisode 5100\tAverage Score: 0.02\nEpisode 5200\tAverage Score: 0.01\nEpisode 5300\tAverage Score: 0.01\nEpisode 5400\tAverage Score: 0.02\nEpisode 5500\tAverage Score: 0.01\nEpisode 5600\tAverage Score: 0.01\nEpisode 5700\tAverage Score: 0.01\nEpisode 5800\tAverage Score: 0.01\nEpisode 5900\tAverage Score: 0.03\nEpisode 6000\tAverage Score: 0.02\nEpisode 6100\tAverage Score: 0.03\nEpisode 6200\tAverage Score: 0.02\nEpisode 6300\tAverage Score: 0.02\nEpisode 6400\tAverage Score: 0.03\nEpisode 6500\tAverage Score: 0.04\nEpisode 6600\tAverage Score: 0.04\nEpisode 6700\tAverage Score: 0.07\nEpisode 6800\tAverage Score: 0.04\nEpisode 6900\tAverage Score: 0.03\nEpisode 7000\tAverage Score: 0.03\nEpisode 7100\tAverage Score: 0.02\nEpisode 7200\tAverage Score: 0.04\nEpisode 7300\tAverage Score: 0.04\nEpisode 7400\tAverage Score: 0.03\nEpisode 7500\tAverage Score: 0.02\nEpisode 7600\tAverage Score: 0.06\nEpisode 7700\tAverage Score: 0.05\nEpisode 7800\tAverage Score: 0.05\nEpisode 7900\tAverage Score: 0.06\nEpisode 8000\tAverage Score: 0.07\nEpisode 8100\tAverage Score: 0.07\nEpisode 8200\tAverage Score: 0.06\nEpisode 8300\tAverage Score: 0.05\nEpisode 8400\tAverage Score: 0.06\nEpisode 8500\tAverage Score: 0.06\nEpisode 8600\tAverage Score: 0.08\nEpisode 8700\tAverage Score: 0.11\nEpisode 8800\tAverage Score: 0.02\nEpisode 8900\tAverage Score: 0.02\nEpisode 9000\tAverage Score: 0.02\nEpisode 9100\tAverage Score: 0.03\nEpisode 9200\tAverage Score: 0.03\nEpisode 9300\tAverage Score: 0.03\nEpisode 9400\tAverage Score: 0.04\nEpisode 9500\tAverage Score: 0.05\nEpisode 9600\tAverage Score: 0.05\nEpisode 9700\tAverage Score: 0.05\nEpisode 9800\tAverage Score: 0.05\nEpisode 9900\tAverage Score: 0.04\nEpisode 10000\tAverage Score: 0.05\n" ], [ "_, _, scores_long_run = load_checkpoint('saved-checkpoints/checkpoint-in-the-long-run.pth')\nfigure = plot_scores(scores_long_run)", "_____no_output_____" ] ], [ [ "* **Watching solved environment agent in action**", "_____no_output_____" ] ], [ [ "# Reloading networks weights\nactor_weights, critic_weights, scores_solved = load_checkpoint('saved-checkpoints/checkpoint-solved.pth')\n\n# Instantiating the agent\nagent = Agent(state_size=state_size, action_size=action_size, random_seed=2)\nagent.actor_local.load_state_dict(actor_weights)\nagent.critic_local.load_state_dict(critic_weights)", "_____no_output_____" ], [ "# Agent acting over 30 episodes\nfor i in range(1, 30): # play game for 30 episodes\n env_info = env.reset(train_mode=False)[brain_name] # reset the environment \n states = env_info.vector_observations # get the current state (for each agent)\n scores = np.zeros(num_agents) # initialize the score (for each agent)\n while True:\n actions = agent.act(states)\n env_info = env.step(actions)[brain_name] # send all actions to tne environment\n next_states = env_info.vector_observations # get next state (for each agent)\n rewards = env_info.rewards # get reward (for each agent)\n dones = env_info.local_done # see if episode finished\n scores += env_info.rewards # update the score (for each agent)\n states = next_states # roll over states to next time step\n if np.any(dones): # exit loop if episode finished\n break\n print('Score (max over agents) from episode {}: {}'.format(i, np.max(scores)))", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "<i>Copyright (c) Microsoft Corporation. All rights reserved.</i>\n\n<i>Licensed under the MIT License.</i>", "_____no_output_____" ], [ "# Testing different Hyperparameters and Benchmarking", "_____no_output_____" ], [ "In this notebook, we will cover how to test different hyperparameters for a particular dataset and how to benchmark different parameters across a group of datasets using AzureML. We assume familiarity with the basic concepts and parameters, which are discussed in the [01_training_introduction.ipynb](01_training_introduction.ipynb), [02_mask_rcnn.ipynb](02_mask_rcnn.ipynb) and [03_training_accuracy_vs_speed.ipynb](03_training_accuracy_vs_speed.ipynb) notebooks. ", "_____no_output_____" ], [ "We will be using a Faster R-CNN model with ResNet-50 backbone to find all objects in an image belonging to 4 categories: 'can', 'carton', 'milk_bottle', 'water_bottle'. We will then conduct hyper-parameter tuning to find the best set of parameters for this model. For this, we present an overall process of utilizing AzureML, specifically [Hyperdrive](https://docs.microsoft.com/en-us/python/api/azureml-train-core/azureml.train.hyperdrive?view=azure-ml-py) which can train and evaluate many different parameter combinations in parallel. We demonstrate the following key steps: \n* Configure AzureML Workspace\n* Create Remote Compute Target (GPU cluster)\n* Prepare Data\n* Prepare Training Script\n* Setup and Run Hyperdrive Experiment\n* Model Import, Re-train and Test\n\nThis notebook is very similar to the [24_exploring_hyperparameters_on_azureml.ipynb](../../classification/notebooks/24_exploring_hyperparameters_on_azureml.ipynb) hyperdrive notebook used for image classification. For key concepts of AzureML see this [tutorial](https://docs.microsoft.com/en-us/azure/machine-learning/service/tutorial-train-models-with-aml?view=azure-ml-py&toc=https%3A%2F%2Fdocs.microsoft.com%2Fen-us%2Fpython%2Fapi%2Fazureml_py_toc%2Ftoc.json%3Fview%3Dazure-ml-py&bc=https%3A%2F%2Fdocs.microsoft.com%2Fen-us%2Fpython%2Fazureml_py_breadcrumb%2Ftoc.json%3Fview%3Dazure-ml-py) on model training and evaluation.", "_____no_output_____" ] ], [ [ "import os\nimport sys\nfrom distutils.dir_util import copy_tree\nimport numpy as np\nimport scrapbook as sb\nimport uuid\n\nimport azureml.core\nfrom azureml.core import Workspace, Experiment\nfrom azureml.core.compute import ComputeTarget, AmlCompute\nfrom azureml.core.compute_target import ComputeTargetException\nimport azureml.data\nfrom azureml.train.estimator import Estimator\nfrom azureml.train.hyperdrive import (\n RandomParameterSampling, GridParameterSampling, BanditPolicy, HyperDriveConfig, PrimaryMetricGoal, choice, uniform\n)\nimport azureml.widgets as widgets\n\nsys.path.append(\"../../\")\nfrom utils_cv.common.azureml import get_or_create_workspace\nfrom utils_cv.common.data import unzip_url\nfrom utils_cv.detection.data import Urls", "_____no_output_____" ] ], [ [ "Ensure edits to libraries are loaded and plotting is shown in the notebook.", "_____no_output_____" ] ], [ [ "%reload_ext autoreload\n%autoreload 2\n%matplotlib inline", "_____no_output_____" ] ], [ [ "We now define some parameters which will be used in this notebook:", "_____no_output_____" ] ], [ [ "# Azure resources\nsubscription_id = \"YOUR_SUBSCRIPTION_ID\"\nresource_group = \"YOUR_RESOURCE_GROUP_NAME\" \nworkspace_name = \"YOUR_WORKSPACE_NAME\" \nworkspace_region = \"YOUR_WORKSPACE_REGION\" #Possible values eastus, eastus2, etc.\n\n# Choose a size for our cluster and the maximum number of nodes\nVM_SIZE = \"STANDARD_NC6\" #STANDARD_NC6S_V3\"\nMAX_NODES = 8\n\n# Hyperparameter grid search space\nIM_MAX_SIZES = [600] #Default is 1333 pixels, defining small values here to speed up training\nLEARNING_RATES = [1e-4, 3e-4, 1e-3, 3e-3, 1e-2]\n\n# Image data\nDATA_PATH = unzip_url(Urls.fridge_objects_path, exist_ok=True)\n\n# Path to utils_cv library\nUTILS_DIR = os.path.join('..', '..', 'utils_cv')", "_____no_output_____" ] ], [ [ "### 1. Config AzureML workspace\nBelow we setup (or load an existing) AzureML workspace, and get all its details as follows. Note that the resource group and workspace will get created if they do not yet exist. For more information regaring the AzureML workspace see also the [20_azure_workspace_setup.ipynb](../../classification/notebooks/20_azure_workspace_setup.ipynb) notebook in the image classification folder.\n\nTo simplify clean-up (see end of this notebook), we recommend creating a new resource group to run this notebook.", "_____no_output_____" ] ], [ [ "ws = get_or_create_workspace(\n subscription_id, resource_group, workspace_name, workspace_region\n)\n\n# Print the workspace attributes\nprint(\n \"Workspace name: \" + ws.name,\n \"Workspace region: \" + ws.location,\n \"Subscription id: \" + ws.subscription_id,\n \"Resource group: \" + ws.resource_group,\n sep=\"\\n\",\n)", "_____no_output_____" ] ], [ [ "### 2. Create Remote Target\nWe create a GPU cluster as our remote compute target. If a cluster with the same name already exists in our workspace, the script will load it instead. This [link](https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-set-up-training-targets#compute-targets-for-training) provides more information about how to set up a compute target on different locations.\n\nBy default, the VM size is set to use STANDARD\\_NC6 machines. However, if quota is available, our recommendation is to use STANDARD\\_NC6S\\_V3 machines which come with the much faster V100 GPU. We set the minimum number of nodes to zero so that the cluster won't incur additional compute charges when not in use.", "_____no_output_____" ] ], [ [ "CLUSTER_NAME = \"gpu-cluster\"\n\ntry:\n # Retrieve if a compute target with the same cluster name already exists\n compute_target = ComputeTarget(workspace=ws, name=CLUSTER_NAME)\n print(\"Found existing compute target.\")\n\nexcept ComputeTargetException:\n # If it doesn't already exist, we create a new one with the name provided\n print(\"Creating a new compute target...\")\n compute_config = AmlCompute.provisioning_configuration(\n vm_size=VM_SIZE, min_nodes=0, max_nodes=MAX_NODES\n )\n\n # create the cluster\n compute_target = ComputeTarget.create(ws, CLUSTER_NAME, compute_config)\n compute_target.wait_for_completion(show_output=True)\n\n# we can use get_status() to get a detailed status for the current cluster.\nprint(compute_target.get_status().serialize())", "Creating a new compute target...\nCreating\nSucceeded\nAmlCompute wait for completion finished\nMinimum number of nodes requested have been provisioned\n{'currentNodeCount': 0, 'targetNodeCount': 0, 'nodeStateCounts': {'preparingNodeCount': 0, 'runningNodeCount': 0, 'idleNodeCount': 0, 'unusableNodeCount': 0, 'leavingNodeCount': 0, 'preemptedNodeCount': 0}, 'allocationState': 'Steady', 'allocationStateTransitionTime': '2019-09-30T18:20:25.067000+00:00', 'errors': None, 'creationTime': '2019-09-30T18:18:06.217384+00:00', 'modifiedTime': '2019-09-30T18:20:38.458332+00:00', 'provisioningState': 'Succeeded', 'provisioningStateTransitionTime': None, 'scaleSettings': {'minNodeCount': 0, 'maxNodeCount': 8, 'nodeIdleTimeBeforeScaleDown': 'PT120S'}, 'vmPriority': 'Dedicated', 'vmSize': 'STANDARD_NC6'}\n" ] ], [ [ "The compute cluster and its status can be seen in the portal. For example in the screenshot below, its automatically resizing (eventually to 0 nodes) to adjust to the number of open runs:\n<img src=\"media/hyperdrive_cluster.jpg\" width=\"800\" alt=\"Compute cluster status\">", "_____no_output_____" ], [ "### 3. Prepare data\nIn this notebook, we'll use the Fridge Objects dataset, which is already stored in the correct format. We then upload our data to the AzureML workspace.\n", "_____no_output_____" ] ], [ [ "# Retrieving default datastore that got automatically created when we setup a workspace\nds = ws.get_default_datastore()\n\n# We now upload the data to a unique sub-folder to avoid accidentially training/evaluating also including older images.\ndata_subfolder = str(uuid.uuid4())\nds.upload(\n src_dir=DATA_PATH, target_path=data_subfolder, overwrite=False, show_progress=True\n)", "_____no_output_____" ] ], [ [ "\nHere's where you can see the data in your portal: \n<img src=\"media/datastore.jpg\" width=\"800\" alt=\"Datastore screenshot for Hyperdrive notebook run\">\n\n### 4. Prepare training script\n\nNext step is to prepare scripts that AzureML Hyperdrive will use to train and evaluate models with selected hyperparameters.", "_____no_output_____" ] ], [ [ "# Create a folder for the training script and copy the utils_cv library into that folder\nscript_folder = os.path.join(os.getcwd(), \"hyperdrive\")\nos.makedirs(script_folder, exist_ok=True)\n_ = copy_tree(UTILS_DIR, os.path.join(script_folder, 'utils_cv'))", "_____no_output_____" ], [ "%%writefile $script_folder/train.py\n\n# Use different matplotlib backend to avoid error during remote execution\nimport matplotlib \nmatplotlib.use(\"Agg\") \nimport matplotlib.pyplot as plt\n\nimport os\nimport sys\nimport argparse\nimport numpy as np\nfrom pathlib import Path\nfrom azureml.core import Run\nfrom utils_cv.detection.dataset import DetectionDataset\nfrom utils_cv.detection.model import DetectionLearner, get_pretrained_fasterrcnn\nfrom utils_cv.common.gpu import which_processor\nwhich_processor()\n\n# Parse arguments passed by Hyperdrive\nparser = argparse.ArgumentParser()\nparser.add_argument('--data-folder', type=str, dest='data_dir')\nparser.add_argument('--data-subfolder', type=str, dest='data_subfolder')\nparser.add_argument('--epochs', type=int, dest='epochs', default=20) \nparser.add_argument('--batch_size', type=int, dest='batch_size', default=2)\nparser.add_argument('--learning_rate', type=float, dest='learning_rate', default=1e-4)\nparser.add_argument('--min_size', type=int, dest='min_size', default=800)\nparser.add_argument('--max_size', type=int, dest='max_size', default=1333)\nparser.add_argument('--rpn_pre_nms_top_n_train', type=int, dest='rpn_pre_nms_top_n_train', default=2000)\nparser.add_argument('--rpn_pre_nms_top_n_test', type=int, dest='rpn_pre_nms_top_n_test', default=1000)\nparser.add_argument('--rpn_post_nms_top_n_train', type=int, dest='rpn_post_nms_top_n_train', default=2000)\nparser.add_argument('--rpn_post_nms_top_n_test', type=int, dest='rpn_post_nms_top_n_test', default=1000)\nparser.add_argument('--rpn_nms_thresh', type=float, dest='rpn_nms_thresh', default=0.7)\nparser.add_argument('--box_score_thresh', type=float, dest='box_score_thresh', default=0.05)\nparser.add_argument('--box_nms_thresh', type=float, dest='box_nms_thresh', default=0.5)\nparser.add_argument('--box_detections_per_img', type=int, dest='box_detections_per_img', default=100)\nargs = parser.parse_args()\nparams = vars(args)\nprint(f\"params = {params}\")\n\n# Get training and validation data\ndata_path = os.path.join(params['data_dir'], params[\"data_subfolder\"])\nprint(f\"data_path={data_path}\")\ndata = DetectionDataset(data_path, train_pct=0.5, batch_size = params[\"batch_size\"])\nprint(\n f\"Training dataset: {len(data.train_ds)} | Training DataLoader: {data.train_dl} \\n \\\n Testing dataset: {len(data.test_ds)} | Testing DataLoader: {data.test_dl}\"\n)\n\n# Get model\nmodel = get_pretrained_fasterrcnn(\n num_classes = len(data.labels)+1,\n min_size = params[\"min_size\"],\n max_size = params[\"max_size\"],\n rpn_pre_nms_top_n_train = params[\"rpn_pre_nms_top_n_train\"],\n rpn_pre_nms_top_n_test = params[\"rpn_pre_nms_top_n_test\"],\n rpn_post_nms_top_n_train = params[\"rpn_post_nms_top_n_train\"], \n rpn_post_nms_top_n_test = params[\"rpn_post_nms_top_n_test\"],\n rpn_nms_thresh = params[\"rpn_nms_thresh\"],\n box_score_thresh = params[\"box_score_thresh\"], \n box_nms_thresh = params[\"box_nms_thresh\"],\n box_detections_per_img = params[\"box_detections_per_img\"]\n)\ndetector = DetectionLearner(data, model)\n\n# Run Training\ndetector.fit(params[\"epochs\"], lr=params[\"learning_rate\"], print_freq=30)\nprint(f\"Average precision after each epoch: {detector.ap}\")\n\n# Get accuracy on test set at IOU=0.5:0.95\nacc = float(detector.ap[-1])\n\n# Add log entries\nrun = Run.get_context()\nrun.log(\"accuracy\", float(acc)) # Logging our primary metric 'accuracy'\nrun.log(\"data_dir\", params[\"data_dir\"])\nrun.log(\"epochs\", params[\"epochs\"])\nrun.log(\"batch_size\", params[\"batch_size\"])\nrun.log(\"learning_rate\", params[\"learning_rate\"])\nrun.log(\"min_size\", params[\"min_size\"])\nrun.log(\"max_size\", params[\"max_size\"])\nrun.log(\"rpn_pre_nms_top_n_train\", params[\"rpn_pre_nms_top_n_train\"])\nrun.log(\"rpn_pre_nms_top_n_test\", params[\"rpn_pre_nms_top_n_test\"])\nrun.log(\"rpn_post_nms_top_n_train\", params[\"rpn_post_nms_top_n_train\"])\nrun.log(\"rpn_post_nms_top_n_test\", params[\"rpn_post_nms_top_n_test\"])\nrun.log(\"rpn_nms_thresh\", params[\"rpn_nms_thresh\"])\nrun.log(\"box_score_thresh\", params[\"box_score_thresh\"])\nrun.log(\"box_nms_thresh\", params[\"box_nms_thresh\"])\nrun.log(\"box_detections_per_img\", params[\"box_detections_per_img\"])", "Overwriting C:\\Users\\pabuehle\\Desktop\\ComputerVision\\scenarios\\detection\\hyperdrive/train.py\n" ] ], [ [ "### 5. Setup and run Hyperdrive experiment\n\n#### 5.1 Create Experiment \nExperiment is the main entry point into experimenting with AzureML. To create new Experiment or get the existing one, we pass our experimentation name 'hyperparameter-tuning'.\n", "_____no_output_____" ] ], [ [ "exp = Experiment(workspace=ws, name=\"hyperparameter-tuning\")", "_____no_output_____" ] ], [ [ "#### 5.2. Define search space\n\nNow we define the search space of hyperparameters. To test discrete parameter values use 'choice()', and for uniform sampling use 'uniform()'. For more options, see [Hyperdrive parameter expressions](https://docs.microsoft.com/en-us/python/api/azureml-train-core/azureml.train.hyperdrive.parameter_expressions?view=azure-ml-py).\n\nHyperdrive provides three different parameter sampling methods: 'RandomParameterSampling', 'GridParameterSampling', and 'BayesianParameterSampling'. Details about each method can be found [here](https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-tune-hyperparameters). Here, we use the 'GridParameterSampling'.", "_____no_output_____" ] ], [ [ "# Grid-search\nparam_sampling = GridParameterSampling(\n {\"--learning_rate\": choice(LEARNING_RATES), \"--max_size\": choice(IM_MAX_SIZES)}\n)", "_____no_output_____" ] ], [ [ "<b>AzureML Estimator</b> is the building block for training. An Estimator encapsulates the training code and parameters, the compute resources and runtime environment for a particular training scenario.\nWe create one for our experimentation with the dependencies our model requires as follows:", "_____no_output_____" ] ], [ [ "script_params = {\"--data-folder\": ds.as_mount(), \"--data-subfolder\": data_subfolder}\n\nest = Estimator(\n source_directory=script_folder,\n script_params=script_params,\n compute_target=compute_target,\n entry_script=\"train.py\",\n use_gpu=True,\n pip_packages=[\"nvidia-ml-py3\", \"fastai\"],\n conda_packages=[\n \"scikit-learn\",\n \"pycocotools>=2.0\",\n \"torchvision==0.3\",\n \"cudatoolkit==9.0\",\n ],\n)", "_____no_output_____" ] ], [ [ "We now create a HyperDriveConfig object which includes information about parameter space sampling, termination policy, primary metric, estimator and the compute target to execute the experiment runs on.", "_____no_output_____" ] ], [ [ "hyperdrive_run_config = HyperDriveConfig(\n estimator=est,\n hyperparameter_sampling=param_sampling,\n policy=None, # Do not use any early termination\n primary_metric_name=\"accuracy\",\n primary_metric_goal=PrimaryMetricGoal.MAXIMIZE,\n max_total_runs=None, # Set to none to run all possible grid parameter combinations,\n max_concurrent_runs=MAX_NODES,\n)", "_____no_output_____" ] ], [ [ "#### 5.3 Run Experiment\n\nWe now run the parameter sweep and visualize the experiment progress using the `RunDetails` widget:\n<img src=\"media/hyperdrive_widget_run.jpg\" width=\"700px\">\n\nOnce completed, the accuracy for the different runs can be analyzed via the widget, for example below is a plot of the accuracy versus learning rate below (for two different image sizes)\n<img src=\"media/hyperdrive_widget_analysis.jpg\" width=\"700px\">\n", "_____no_output_____" ] ], [ [ "hyperdrive_run = exp.submit(config=hyperdrive_run_config)\nprint(f\"Url to hyperdrive run on the Azure portal: {hyperdrive_run.get_portal_url()}\")", "Url to hyperdrive run on the Azure portal: https://mlworkspace.azure.ai/portal/subscriptions/989b90f7-da4f-41f9-84c9-44848802052d/resourceGroups/pabuehle_delme2_hyperdrive/providers/Microsoft.MachineLearningServices/workspaces/pabuehle_ws/experiments/hyperparameter-tuning/runs/hyperparameter-tuning_1569867670036119\n" ], [ "widgets.RunDetails(hyperdrive_run).show()", "_____no_output_____" ], [ "hyperdrive_run.wait_for_completion()", "_____no_output_____" ] ], [ [ "To load an existing Hyperdrive Run instead of start new one, we can use \n```python\nhyperdrive_run = azureml.train.hyperdrive.HyperDriveRun(exp, <your-run-id>, hyperdrive_run_config=hyperdrive_run_config)\n```\nWe also can cancel the Run with \n```python \nhyperdrive_run.cancel().\n```\n\nOnce all the child-runs are finished, we can get the best run and the metrics.", "_____no_output_____" ] ], [ [ "# Get best run and print out metrics\nbest_run = hyperdrive_run.get_best_run_by_primary_metric()\nbest_run_metrics = best_run.get_metrics()\nparameter_values = best_run.get_details()[\"runDefinition\"][\"arguments\"]\nbest_parameters = dict(zip(parameter_values[::2], parameter_values[1::2]))\n\nprint(f\"* Best Run Id:{best_run.id}\")\nprint(best_run)\nprint(\"\\n* Best hyperparameters:\")\nprint(best_parameters)\nprint(f\"Accuracy = {best_run_metrics['accuracy']}\")\nprint(\"Learning Rate =\", best_run_metrics[\"learning_rate\"])", "* Best Run Id:hyperparameter-tuning_1569867670036119_4\nRun(Experiment: hyperparameter-tuning,\nId: hyperparameter-tuning_1569867670036119_4,\nType: azureml.scriptrun,\nStatus: Completed)\n\n* Best hyperparameters:\n{'--data-folder': '$AZUREML_DATAREFERENCE_workspaceblobstore', '--data-subfolder': '01679d79-1c47-49b8-88c3-d657f36b0c0f', '--learning_rate': '0.01', '--max_size': '600'}\nAccuracy = 0.8918015856432082\nLearning Rate = 0.01\n" ], [ "hyperdrive_run.get_children_sorted_by_primary_metric()", "_____no_output_____" ] ], [ [ "### 7. Clean up\n\nTo avoid unnecessary expenses, all resources which were created in this notebook need to get deleted once parameter search is concluded. To simplify this clean-up step, we recommended creating a new resource group to run this notebook. This resource group can then be deleted, e.g. using the Azure Portal, which will remove all created resources.", "_____no_output_____" ] ], [ [ "# Log some outputs using scrapbook which are used during testing to verify correct notebook execution\nsb.glue(\"best_accuracy\", best_run_metrics[\"accuracy\"])", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import os\nimport torch\nfrom torch.utils.data import DataLoader, Dataset\nfrom torchvision.transforms import ToTensor, ToPILImage\nfrom torchvision.models.detection import fasterrcnn_resnet50_fpn\nfrom torchvision.models.detection.faster_rcnn import FastRCNNPredictor\nfrom PIL import Image", "_____no_output_____" ], [ "class PlayerDataset(Dataset):\n def __init__(self, root):\n self.root = root\n self.images = list(sorted(os.listdir(root + '/images')))\n self.targets = [target for target in list(sorted(os.listdir(root + '/targets'))) if target != 'classes.txt']\n \n def __len__(self):\n return len(self.images)\n \n def __getitem__(self, idx):\n image_path = os.path.join(self.root, 'images', self.images[idx])\n target_path = os.path.join(self.root, 'targets', self.targets[idx])\n \n image = ToTensor()(Image.open(image_path).convert(\"RGB\"))\n \n f = open(target_path)\n target = f.readline().strip().split()\n \n w = 1280\n h = 720\n \n center_x = float(target[1]) * w\n center_y = float(target[2]) * h\n bbox_w = float(target[3]) * w\n bbox_h = float(target[4]) * h\n \n x0 = round(center_x - (bbox_w / 2))\n x1 = round(center_x + (bbox_w / 2))\n y0 = round(center_y - (bbox_h / 2))\n y1 = round(center_y + (bbox_h / 2))\n \n print(x1 - x0)\n print(y1 - y0)\n \n boxes = torch.as_tensor([x0, y0, x1, y1], dtype=torch.float32)\n labels = torch.as_tensor(0, dtype=torch.int64)\n \n target = [{'boxes': boxes, 'labels': labels}]\n \n return image, target", "_____no_output_____" ], [ "def train_model(model, optimizer, lr_scheduler, data_loader, device, num_epochs):\n model.train()\n for epoch in range(num_epochs):\n running_loss = 0.0\n for images, targets in data_loader:\n images = list(image.to(device) for image in images)\n targets = [{k: v.to(device) for k, v in t.items()} for t in targets]\n print(targets)\n \n loss_dict = model(images, targets)\n losses = sum(loss for loss in loss_dict.values())\n \n optimizer.zero_grad()\n losses.backward()\n optimizer.step()\n lr_scheduler.step()\n \n running_loss += losses.item()\n print('epoch:%d loss: %.3f' % (epoch + 1, running_loss))", "_____no_output_____" ], [ "def evaluate(model, data_loader, device):\n model.eval()\n cpu_device = torch.device(\"cpu\")\n with torch.no_grad():\n for images, targets in data_loader:\n images = list(image.to(device) for image in images)\n targets = [{k: v.to(device) for k, v in t.items()} for t in targets]\n \n outputs = [{k: v.to(cpu_device) for k, v in t.items()} for t in model(images)]\n print(outputs)", "_____no_output_____" ], [ "model = fasterrcnn_resnet50_fpn(num_classes=1)\ndevice = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')\nmodel.to(device)\n\ntrain_dataset = PlayerDataset('data/train')\ntest_dataset = PlayerDataset('data/test')\n\ntrain_data_loader = DataLoader(train_dataset, batch_size=1, shuffle=True, num_workers=4)\ntest_data_loader = DataLoader(test_dataset, batch_size=1, shuffle=False, num_workers=4)\n\nparams = [p for p in model.parameters() if p.requires_grad]\noptimizer = torch.optim.SGD(params, lr=0.005, momentum=0.9, weight_decay=0.0005)\nlr_scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=3, gamma=0.1)", "_____no_output_____" ], [ "train_model(model, optimizer, lr_scheduler, train_data_loader, device, 1)", "80\n136\n85\n67\n151\n146\n54\n131\n56\n146\n61\n147\n81\n151\n66\n150\n48\n150\n[{'boxes': tensor([[ 49., 227., 134., 378.]]), 'labels': tensor([0])}]\n89\n146\nepoch:1 loss: 0.698\n[{'boxes': tensor([[202., 212., 282., 348.]]), 'labels': tensor([0])}]\n62\n147\nepoch:1 loss: 1.395\n[{'boxes': tensor([[ 9., 232., 76., 378.]]), 'labels': tensor([0])}]\n" ], [ "evaluate(model, test_data_loader, device)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "!git clone https://github.com/GraphGrailAi/ruGPT3-ZhirV", "Cloning into 'ruGPT3-ZhirV'...\nremote: Enumerating objects: 280, done.\u001b[K\nremote: Counting objects: 100% (280/280), done.\u001b[K\nremote: Compressing objects: 100% (170/170), done.\u001b[K\nremote: Total 280 (delta 109), reused 279 (delta 108), pack-reused 0\u001b[K\nReceiving objects: 100% (280/280), 6.07 MiB | 2.10 MiB/s, done.\nResolving deltas: 100% (109/109), done.\n" ], [ "cd ruGPT3-ZhirV", "/home/jovyan/ruGPT3-ZhirV\n" ], [ "cd ..", "/home/jovyan\n" ], [ "!pip3 install -r requirements.txt", "Defaulting to user installation because normal site-packages is not writeable\nRequirement already satisfied: nltk>=3.4 in /usr/local/lib/python3.6/dist-packages (from -r requirements.txt (line 1)) (3.5)\nRequirement already satisfied: numpy>=1.15.4 in /usr/local/lib/python3.6/dist-packages (from -r requirements.txt (line 2)) (1.18.4)\nRequirement already satisfied: pandas>=0.24.0 in /usr/local/lib/python3.6/dist-packages (from -r requirements.txt (line 3)) (1.0.3)\nRequirement already satisfied: sentencepiece>=0.1.8 in /usr/local/lib/python3.6/dist-packages (from -r requirements.txt (line 4)) (0.1.91)\nCollecting tensorflow>=1.12.0\n Using cached tensorflow-2.3.1-cp36-cp36m-manylinux2010_x86_64.whl (320.4 MB)\nCollecting boto3==1.11.11\n Using cached boto3-1.11.11-py2.py3-none-any.whl (128 kB)\nCollecting regex==2020.1.8\n Using cached regex-2020.1.8-cp36-cp36m-manylinux2010_x86_64.whl (689 kB)\nCollecting transformers==2.8.0\n Using cached transformers-2.8.0-py3-none-any.whl (563 kB)\nRequirement already satisfied: tqdm in /usr/local/lib/python3.6/dist-packages (from nltk>=3.4->-r requirements.txt (line 1)) (4.46.0)\nRequirement already satisfied: click in /usr/local/lib/python3.6/dist-packages (from nltk>=3.4->-r requirements.txt (line 1)) (7.1.2)\nRequirement already satisfied: joblib in /usr/local/lib/python3.6/dist-packages (from nltk>=3.4->-r requirements.txt (line 1)) (0.15.1)\nRequirement already satisfied: pytz>=2017.2 in /usr/local/lib/python3.6/dist-packages (from pandas>=0.24.0->-r requirements.txt (line 3)) (2020.1)\nRequirement already satisfied: python-dateutil>=2.6.1 in /usr/local/lib/python3.6/dist-packages (from pandas>=0.24.0->-r requirements.txt (line 3)) (2.8.1)\nRequirement already satisfied: h5py<2.11.0,>=2.10.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (2.10.0)\nRequirement already satisfied: wheel>=0.26 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (0.34.2)\nRequirement already satisfied: gast==0.3.3 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (0.3.3)\nCollecting tensorboard<3,>=2.3.0\n Using cached tensorboard-2.4.0-py3-none-any.whl (10.6 MB)\nRequirement already satisfied: wrapt>=1.11.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.12.1)\nRequirement already satisfied: opt-einsum>=2.3.2 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (3.2.1)\nRequirement already satisfied: keras-preprocessing<1.2,>=1.1.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.1.2)\nRequirement already satisfied: absl-py>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (0.9.0)\nCollecting tensorflow-estimator<2.4.0,>=2.3.0\n Using cached tensorflow_estimator-2.3.0-py2.py3-none-any.whl (459 kB)\nRequirement already satisfied: astunparse==1.6.3 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.6.3)\nRequirement already satisfied: grpcio>=1.8.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.29.0)\nRequirement already satisfied: google-pasta>=0.1.8 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (0.2.0)\nRequirement already satisfied: protobuf>=3.9.2 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (3.12.2)\nRequirement already satisfied: six>=1.12.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.15.0)\nRequirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.1.0)\nRequirement already satisfied: s3transfer<0.4.0,>=0.3.0 in /usr/local/lib/python3.6/dist-packages (from boto3==1.11.11->-r requirements.txt (line 6)) (0.3.3)\nRequirement already satisfied: jmespath<1.0.0,>=0.7.1 in /usr/local/lib/python3.6/dist-packages (from boto3==1.11.11->-r requirements.txt (line 6)) (0.10.0)\nCollecting botocore<1.15.0,>=1.14.11\n Using cached botocore-1.14.17-py2.py3-none-any.whl (5.9 MB)\nCollecting filelock\n Using cached filelock-3.0.12-py3-none-any.whl (7.6 kB)\nCollecting dataclasses; python_version < \"3.7\"\n Using cached dataclasses-0.8-py3-none-any.whl (19 kB)\nCollecting tokenizers==0.5.2\n Using cached tokenizers-0.5.2-cp36-cp36m-manylinux1_x86_64.whl (3.7 MB)\nRequirement already satisfied: requests in /usr/local/lib/python3.6/dist-packages (from transformers==2.8.0->-r requirements.txt (line 8)) (2.23.0)\nProcessing /home/jovyan/.cache/pip/wheels/29/3c/fd/7ce5c3f0666dab31a50123635e6fb5e19ceb42ce38d4e58f45/sacremoses-0.0.43-cp36-none-any.whl\nRequirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.6/dist-packages (from tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (3.2.2)\nRequirement already satisfied: google-auth<2,>=1.6.3 in /usr/local/lib/python3.6/dist-packages (from tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.16.0)\nRequirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /usr/local/lib/python3.6/dist-packages (from tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.6.0.post3)\nRequirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.6/dist-packages (from tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.0.1)\nRequirement already satisfied: setuptools>=41.0.0 in /usr/local/lib/python3.6/dist-packages (from tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (47.1.0)\nRequirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /usr/local/lib/python3.6/dist-packages (from tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (0.4.1)\nRequirement already satisfied: docutils<0.16,>=0.10 in /usr/local/lib/python3.6/dist-packages (from botocore<1.15.0,>=1.14.11->boto3==1.11.11->-r requirements.txt (line 6)) (0.15.2)\nRequirement already satisfied: urllib3<1.26,>=1.20; python_version != \"3.4\" in /usr/local/lib/python3.6/dist-packages (from botocore<1.15.0,>=1.14.11->boto3==1.11.11->-r requirements.txt (line 6)) (1.25.9)\nRequirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->transformers==2.8.0->-r requirements.txt (line 8)) (3.0.4)\nRequirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->transformers==2.8.0->-r requirements.txt (line 8)) (2.9)\nRequirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests->transformers==2.8.0->-r requirements.txt (line 8)) (2020.4.5.1)\nRequirement already satisfied: importlib-metadata; python_version < \"3.8\" in /usr/local/lib/python3.6/dist-packages (from markdown>=2.6.8->tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.6.0)\nRequirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (0.2.8)\nRequirement already satisfied: rsa<4.1,>=3.1.4 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (4.0)\nRequirement already satisfied: cachetools<5.0,>=2.0.0 in /usr/local/lib/python3.6/dist-packages (from google-auth<2,>=1.6.3->tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (4.1.0)\nRequirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (1.3.0)\nRequirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.6/dist-packages (from importlib-metadata; python_version < \"3.8\"->markdown>=2.6.8->tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (3.1.0)\nRequirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /usr/local/lib/python3.6/dist-packages (from pyasn1-modules>=0.2.1->google-auth<2,>=1.6.3->tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (0.4.8)\nRequirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.6/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<3,>=2.3.0->tensorflow>=1.12.0->-r requirements.txt (line 5)) (3.1.0)\n\u001b[31mERROR: tensorflow-gpu 2.2.0 has requirement tensorboard<2.3.0,>=2.2.0, but you'll have tensorboard 2.4.0 which is incompatible.\u001b[0m\n\u001b[31mERROR: tensorflow-gpu 2.2.0 has requirement tensorflow-estimator<2.3.0,>=2.2.0, but you'll have tensorflow-estimator 2.3.0 which is incompatible.\u001b[0m\n\u001b[31mERROR: awscli 1.18.135 has requirement botocore==1.17.58, but you'll have botocore 1.14.17 which is incompatible.\u001b[0m\nInstalling collected packages: tensorboard, tensorflow-estimator, tensorflow, botocore, boto3, regex, filelock, dataclasses, tokenizers, sacremoses, transformers\n\u001b[33m WARNING: The script tensorboard is installed in '/home/jovyan/.local/bin' which is not on PATH.\n Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.\u001b[0m\n\u001b[33m WARNING: The scripts estimator_ckpt_converter, saved_model_cli, tensorboard, tf_upgrade_v2, tflite_convert, toco and toco_from_protos are installed in '/home/jovyan/.local/bin' which is not on PATH.\n Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.\u001b[0m\n\u001b[33m WARNING: The script sacremoses is installed in '/home/jovyan/.local/bin' which is not on PATH.\n Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.\u001b[0m\nSuccessfully installed boto3-1.11.11 botocore-1.14.17 dataclasses-0.8 filelock-3.0.12 regex-2020.1.8 sacremoses-0.0.43 tensorboard-2.4.0 tensorflow-2.3.1 tensorflow-estimator-2.3.0 tokenizers-0.5.2 transformers-2.8.0\n\u001b[33mWARNING: You are using pip version 20.1.1; however, version 20.2.4 is available.\nYou should consider upgrading via the '/usr/bin/python -m pip install --upgrade pip' command.\u001b[0m\n" ] ], [ [ "!python generate_transformers.py \\\n --model_type=gpt2 \\\n --model_name_or_path=sberbank-ai/rugpt3large_based_on_gpt2 \\\n --k=5 \\\n --p=0.95 \\\n --length=100", "_____no_output_____" ] ], [ [ "Обучение эссе", "_____no_output_____" ], [ "!python pretrain_transformers.py \\\n --output_dir=/home/jovyan/ruGPT3-ZhirV/ \\\n --overwrite_output_dir \\\n --model_type=gpt2 \\\n --model_name_or_path=sberbank-ai/rugpt3large_based_on_gpt2 \\\n --do_train \\\n --train_data_file=/home/jovyan/ruGPT3-ZhirV/data/all_essays.jsonl \\\n --do_eval \\\n --eval_data_file=/home/jovyan/ruGPT3-ZhirV/data/valid_essays.jsonl \\\n --num_train_epochs 10 \\\n --overwrite_cache \\\n --block_size=1024 \\\n --per_gpu_train_batch_size 1 \\\n --gradient_accumulation_steps 8", "_____no_output_____" ], [ "# Обучение Жириновский", "_____no_output_____" ] ], [ [ "!python pretrain_transformers.py \\\n--output_dir=/home/jovyan/ruGPT3-ZhirV/ \\\n--overwrite_output_dir \\\n--model_type=gpt2 \\\n--model_name_or_path=sberbank-ai/rugpt3large_based_on_gpt2 \\\n--do_train \\\n--train_data_file=/home/jovyan/ruGPT3-ZhirV/data/girik_all2.jsonl \\\n--do_eval \\\n--eval_data_file=/home/jovyan/ruGPT3-ZhirV/data/girik_valid.jsonl \\\n--num_train_epochs 20 \\\n--overwrite_cache \\\n--block_size=1024 \\\n--per_gpu_train_batch_size 1 \\\n--gradient_accumulation_steps 8", "2020-11-22 17:37:37.710892: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1\n11/22/2020 17:37:45 - WARNING - __main__ - Process rank: -1, device: cuda, n_gpu: 1, distributed training: False, 16-bits training: False\n11/22/2020 17:37:46 - INFO - transformers.configuration_utils - loading configuration file https://s3.amazonaws.com/models.huggingface.co/bert/sberbank-ai/rugpt3large_based_on_gpt2/config.json from cache at /home/jovyan/.cache/torch/transformers/53218293a9edec913332b4f2d178496a60f98d64a1af74f92984804152f9404c.02a103afdbdbf4896cc41fc6495e47b7e5e2f353a287fe98d178e669be028903\n11/22/2020 17:37:46 - INFO - transformers.configuration_utils - Model config GPT2Config {\n \"_num_labels\": 2,\n \"activation_function\": \"gelu_new\",\n \"architectures\": [\n \"GPT2LMHeadModel\"\n ],\n \"attn_pdrop\": 0.1,\n \"bad_words_ids\": null,\n \"bos_token_id\": 50256,\n \"decoder_start_token_id\": null,\n \"do_sample\": false,\n \"early_stopping\": false,\n \"embd_pdrop\": 0.1,\n \"eos_token_id\": 50256,\n \"finetuning_task\": null,\n \"gradient_checkpointing\": false,\n \"id2label\": {\n \"0\": \"LABEL_0\",\n \"1\": \"LABEL_1\"\n },\n \"initializer_range\": 0.02,\n \"is_decoder\": false,\n \"is_encoder_decoder\": false,\n \"label2id\": {\n \"LABEL_0\": 0,\n \"LABEL_1\": 1\n },\n \"layer_norm_epsilon\": 1e-05,\n \"length_penalty\": 1.0,\n \"max_length\": 20,\n \"min_length\": 0,\n \"model_type\": \"gpt2\",\n \"n_ctx\": 2048,\n \"n_embd\": 1536,\n \"n_head\": 16,\n \"n_inner\": null,\n \"n_layer\": 24,\n \"n_positions\": 2048,\n \"no_repeat_ngram_size\": 0,\n \"num_beams\": 1,\n \"num_return_sequences\": 1,\n \"output_attentions\": false,\n \"output_hidden_states\": false,\n \"output_past\": true,\n \"pad_token_id\": null,\n \"prefix\": null,\n \"pruned_heads\": {},\n \"repetition_penalty\": 1.0,\n \"resid_pdrop\": 0.1,\n \"summary_activation\": null,\n \"summary_first_dropout\": 0.1,\n \"summary_proj_to_labels\": true,\n \"summary_type\": \"cls_index\",\n \"summary_use_proj\": true,\n \"task_specific_params\": null,\n \"temperature\": 1.0,\n \"top_k\": 50,\n \"top_p\": 1.0,\n \"torchscript\": false,\n \"use_bfloat16\": false,\n \"vocab_size\": 50257\n}\n\n11/22/2020 17:37:46 - INFO - transformers.configuration_utils - loading configuration file https://s3.amazonaws.com/models.huggingface.co/bert/sberbank-ai/rugpt3large_based_on_gpt2/config.json from cache at /home/jovyan/.cache/torch/transformers/53218293a9edec913332b4f2d178496a60f98d64a1af74f92984804152f9404c.02a103afdbdbf4896cc41fc6495e47b7e5e2f353a287fe98d178e669be028903\n11/22/2020 17:37:46 - INFO - transformers.configuration_utils - Model config GPT2Config {\n \"_num_labels\": 2,\n \"activation_function\": \"gelu_new\",\n \"architectures\": [\n \"GPT2LMHeadModel\"\n ],\n \"attn_pdrop\": 0.1,\n \"bad_words_ids\": null,\n \"bos_token_id\": 50256,\n \"decoder_start_token_id\": null,\n \"do_sample\": false,\n \"early_stopping\": false,\n \"embd_pdrop\": 0.1,\n \"eos_token_id\": 50256,\n \"finetuning_task\": null,\n \"gradient_checkpointing\": false,\n \"id2label\": {\n \"0\": \"LABEL_0\",\n \"1\": \"LABEL_1\"\n },\n \"initializer_range\": 0.02,\n \"is_decoder\": false,\n \"is_encoder_decoder\": false,\n \"label2id\": {\n \"LABEL_0\": 0,\n \"LABEL_1\": 1\n },\n \"layer_norm_epsilon\": 1e-05,\n \"length_penalty\": 1.0,\n \"max_length\": 20,\n \"min_length\": 0,\n \"model_type\": \"gpt2\",\n \"n_ctx\": 2048,\n \"n_embd\": 1536,\n \"n_head\": 16,\n \"n_inner\": null,\n \"n_layer\": 24,\n \"n_positions\": 2048,\n \"no_repeat_ngram_size\": 0,\n \"num_beams\": 1,\n \"num_return_sequences\": 1,\n \"output_attentions\": false,\n \"output_hidden_states\": false,\n \"output_past\": true,\n \"pad_token_id\": null,\n \"prefix\": null,\n \"pruned_heads\": {},\n \"repetition_penalty\": 1.0,\n \"resid_pdrop\": 0.1,\n \"summary_activation\": null,\n \"summary_first_dropout\": 0.1,\n \"summary_proj_to_labels\": true,\n \"summary_type\": \"cls_index\",\n \"summary_use_proj\": true,\n \"task_specific_params\": null,\n \"temperature\": 1.0,\n \"top_k\": 50,\n \"top_p\": 1.0,\n \"torchscript\": false,\n \"use_bfloat16\": false,\n \"vocab_size\": 50257\n}\n\n11/22/2020 17:37:46 - INFO - transformers.tokenization_utils - Model name 'sberbank-ai/rugpt3large_based_on_gpt2' not found in model shortcut name list (gpt2, gpt2-medium, gpt2-large, gpt2-xl, distilgpt2). Assuming 'sberbank-ai/rugpt3large_based_on_gpt2' is a path, a model identifier, or url to a directory containing tokenizer files.\n11/22/2020 17:37:49 - INFO - transformers.tokenization_utils - loading file https://s3.amazonaws.com/models.huggingface.co/bert/sberbank-ai/rugpt3large_based_on_gpt2/vocab.json from cache at /home/jovyan/.cache/torch/transformers/39e50567636d4014628a4fb0b7665a179a6109d96765eb4e6a10e9f2306f963d.de52bc5880aff0437c7f24c33b71ecae48f6f03f0449dfe933503132c6c1cc26\n11/22/2020 17:37:49 - INFO - transformers.tokenization_utils - loading file https://s3.amazonaws.com/models.huggingface.co/bert/sberbank-ai/rugpt3large_based_on_gpt2/merges.txt from cache at /home/jovyan/.cache/torch/transformers/0a94bcfc9ca640e268e53959b05f2ebe267a5cb686289b46cac4ffac589eac40.5885500c9887f152893bfadf3b511a9105243c57bfc45889e3552bdc61090032\n11/22/2020 17:37:49 - INFO - transformers.tokenization_utils - loading file https://s3.amazonaws.com/models.huggingface.co/bert/sberbank-ai/rugpt3large_based_on_gpt2/added_tokens.json from cache at None\n11/22/2020 17:37:49 - INFO - transformers.tokenization_utils - loading file https://s3.amazonaws.com/models.huggingface.co/bert/sberbank-ai/rugpt3large_based_on_gpt2/special_tokens_map.json from cache at None\n11/22/2020 17:37:49 - INFO - transformers.tokenization_utils - loading file https://s3.amazonaws.com/models.huggingface.co/bert/sberbank-ai/rugpt3large_based_on_gpt2/tokenizer_config.json from cache at None\n11/22/2020 17:37:49 - INFO - transformers.modeling_utils - loading weights file https://s3.amazonaws.com/models.huggingface.co/bert/sberbank-ai/rugpt3large_based_on_gpt2/pytorch_model.bin from cache at /home/jovyan/.cache/torch/transformers/5f2ce73f5df1b0b20e9c0d5fadbedefdc9b484edcbc39252a1c913b1b4ce6cd2.5bdac7adaf803c2b7192441aba3020af4140f7177089f8f95940a0c073059a31\n" ] ], [ [ "# Генерация Жириновский", "_____no_output_____" ] ], [ [ "from transformers import GPT2Tokenizer, GPT2LMHeadModel\n\ntokenizer = GPT2Tokenizer.from_pretrained(\"checkpoint-1000\")\nmodel = GPT2LMHeadModel.from_pretrained(\"checkpoint-1000\")\nmodel.to(\"cuda\")", "_____no_output_____" ], [ "import copy\n\nbad_word_ids = [\n [203], # \\n\n [225], # weird space 1\n [28664], # weird space 2\n [13298], # weird space 3\n [206], # \\r\n [49120], # html\n [25872], # http\n [3886], # amp\n [38512], # nbsp\n [10], # &\n [5436], # & (another)\n [5861], # http\n [372], # yet another line break\n [421, 4395], # МСК\n [64], # \\\n [33077], # https\n [1572], # ru\n [11101], # Источник\n]\n\ndef gen_fragment(context, bad_word_ids=bad_word_ids, print_debug_output=False):\n input_ids = tokenizer.encode(context, add_special_tokens=False, return_tensors=\"pt\").to(\"cuda\")\n input_ids = input_ids[:, -1700:]\n input_size = input_ids.size(1)\n output_sequences = model.generate(\n input_ids=input_ids,\n max_length=175 + input_size,\n min_length=40 + input_size,\n top_p=0.95,\n #top_k=0,\n do_sample=True,\n num_return_sequences=1,\n temperature=1.0, # 0.9,\n pad_token_id=0,\n eos_token_id=2,\n bad_words_ids=bad_word_ids,\n no_repeat_ngram_size=6\n )\n if len(output_sequences.shape) > 3:\n output_sequences.squeeze_()\n generated_sequence = output_sequences[0].tolist()[input_size:]\n if print_debug_output:\n for idx in generated_sequence:\n print(idx, tokenizer.decode([idx], clean_up_tokenization_spaces=True).strip())\n text = tokenizer.decode(generated_sequence, clean_up_tokenization_spaces=True)\n text = text[: text.find(\"</s>\")]\n text = text[: text.rfind(\".\") + 1]\n return context + text\n\ndef gen_girik(context, sign, bad_word_ids, print_debug_output=False):\n bad_word_ids_girik = copy.copy(bad_word_ids)\n bad_word_ids_girik += [tokenizer.encode(bad_word, add_prefix_space=True) for bad_word in signs]\n bad_word_ids_girik += [tokenizer.encode(\".\" + bad_word, add_prefix_space=False) for bad_word in signs]\n return gen_fragment(context + \"\\n\\n\" + sign + \"\\n\", bad_word_ids_girik, print_debug_output=False)\n\nsigns = [\"Лингвистическому мусору и иностранным словам в русском языке не место!\",\n \"Будет ли Путин президентом после 2024 года?\", \n \"Кто победил: Армения или Азербайджан?\",\n \"И последнее. Когда в России настанет долгожданный мир во всём мире? И чтобы больше таких вопросов не было.\",\n \"Почему Европа постоянно вводит санкции против России?\",\n \"Не надо шутить с войной. Здесь другие ребята.\",\n \"Ночью наши учёные чуть-чуть изменят гравитационное поле Земли, и твоя страна будет под водой.\",\n \"Что было бы, если бы Жириновский стал президентом?\",\n \"Когда Россия станет самой богатой и могущественной страной в мире?\",\n \"Джордж, Джордж! Посмотри ковбойские фильмы!\",\n \"От чего коровы с ума сходят? От британской демократии.\",\n\n ]\nbeginning = \"Жириновский говорит:.\"\ncurrent_text = beginning\nfor sign in signs:\n current_text = gen_girik(current_text, sign, bad_word_ids)\nprint(current_text)", "Жириновский говорит:.\n\nЛингвистическому мусору и иностранным словам в русском языке не место!\n ДА! ДА! В русском языке есть иностранные слова, но их не должно быть! Я лично считаю, что все иностранные слова должны быть убраны из нашего великого и могучего языка! Они мешают нам правильно воспринимать русскую речь. Когда иностранное слово употребляется рядом с русскими словами, то это значит, что иностранное слово уже заняло правильное место в русском языке. Вот так должны звучать в русском языке все иностранные слова. А если есть какое-либо иностранное слово, то есть и другие иностранные слова, которые также должны быть убраны из русского языка. Я против того, чтобы иностранные слова занимали в русском языке правильные места. В русском языке все слова должны звучать красиво. Почему в русском языке много иностранных слов? Потому что нас с вами обманывают. Нас заставляют учить английский язык. Нас убеждают в том, что он самый красивый.\n\nБудет ли Путин президентом после 2024 года?\n Почему? Какие прогнозы? Владислав Сурков, советник Президента РФ Путина, поделился своим видением ситуации с окончательным определением будущего главы Российского государства. – Владислав Юрьевич, как Вы считаете, в 2024 году президентом России может стать Владимир Владимирович Путин, который одержал убедительную победу на президентских выборах в 2018 году? – На мой взгляд, вполне вероятно, что в 2024 году президентом РФ действительно будет Владимир Владимирович Путин, которого поддерживает, я бы даже сказал, боготворит большая часть россиян, я в этом совершенно уверен. Несмотря на то что Владимир Путин одержал убедительную победу на выборах и, безусловно, этот успех будет использован его сторонниками для укрепления президентской власти в России, он будет опираться исключительно на поддержку своей огромной армии сторонников, а также тех сил, которые его выдвинули в президенты России.\n\nКто победил: Армения или Азербайджан?\n Почему? Есть ли шансы у Армении вернуть Карабахский конфликт в правовое поле? И можно ли совместить в одном лице дипломата и политика? Обострение ситуации вокруг карабахского конфликта и перспективы его разрешения в российско-азербайджанских отношениях. Арцах – Самый проблемный регион Азербайджана, один из самых бедных. Он пострадал больше других территорий бывшего СССР. И в 1990 году, сразу после провозглашения независимости Азербайджана, этот регион подвергся этническому геноциду со стороны армянских националистов. В 1991 году они полностью уничтожили армянское население села Кармир-Уллу, расположенного в одноименном районе Азербайджана. Более 800 тысяч армян были убиты или пропали без вести. В 1994 году в результате общенациональной забастовки шахтеры прекратили добычу угля, и город Степанакерт оказался парализованным.\n\nИ последнее. Когда в России настанет долгожданный мир во всём мире? И чтобы больше таких вопросов не было.\n Восточная Европа. Взгляд через столетие Сегодня мы приглашаем вас совершить небольшое путешествие во времени, чтобы узнать, что происходило на территории Восточной Европы в начале 20-го века. Что представляла собой эта обширная территория тогда, в эпоху великих географических открытий? Какие этнические, политические и религиозные противоречия таились в её недрах? На эти и другие вопросы вам ответят книги из знаменитой серии «Рюриковичи» издательства «Вече», выпущенные в серии «ЖЗЛ». Вы узнаете о древних русских городах – Москве, Новгороде, Великом Новгороде и Пскове, об одном из величайших полководцев русской истории – Андрее Боголюбском, о князе Владимире Крестителе, о княгине Ольге и многом другом.\n\nПочему Европа постоянно вводит санкции против России?\n Санкции против России всегда были болезненным ударом по экономике Запада. Особенно больно бьют санкции по Европе. Евросоюз страдает из-за отсутствия в России товаров европейского производства. Особенно тяжело приходится в этом отношении гражданам Евросоюза. Без российских товаров они, например, не могут купить автомобиль. За всё приходится платить – и за труд рабочих, и за электричество, и за транспорт. Вот в чём главная беда и трагедия Евросоюза. Но ещё большая беда и трагедия Евросоюза – это Россия. Страдают и европейцы, и русские. Только в России на каждого человека, страдающего от санкций, приходится около трёхсот человек, страдающих от эмбарго России. Но и здесь есть выход из положения. Нужно, наоборот, поддержать тех, кто страдает, и помочь тем, кто в этом нуждается, особенно в России.\n\nНе надо шутить с войной. Здесь другие ребята.\n Это не ИГИЛ[1 - ИГИЛ – запрещённая в России террористическая организация]. ИГИЛ[2 - ИГИЛ – запрещённое в России террористическое организация] – это шутки для школьников. ИГИЛ[3 - ИГИЛ – запрещён в России террористическая организация] – это серьёзные ребята. ИГИЛ[4 - ИГИЛ – запрещёнв России террористическая организация] шутить не будет. ИГИЛ[5 - ИГИЛ – запрещённа в России террористическая организация], как Гитлер, пришёл за жизнями простых людей. ИГИЛ[6 - ИГИЛ – запрещён террористическая организация] не шуточки шутить. ИГИЛ[7 - ИГИЛ – запрещён запрещён в России террористическа организация] не шутит. ИГИЛ[8 - ИГИЛ – запрещённы в России террористическа организации] уничтожит вас всех.\n\nНочью наши учёные чуть-чуть изменят гравитационное поле Земли, и твоя страна будет под водой.\n Под водой. Они поднимут со дна океаны. Весь мир утонет. Весь мир будет под водой. Понимаешь? Весь! Только малая часть будет под водой. Но зато какая! Супер! Весь мир! Озеро Байкал, Ангара, море Лаптевых. Понимаешь? Там тоже будет супер! Супер! Весь мир будет под водой! Но не весь. Малый кусок в океане останется. Но зато какой! Море Лаптевых, море золота. И золото! Это тоже супер! Но это уже после войны будет. А пока только ночью. Ночью. Они чуть изменят гравитационное поле. И вся планета будет под водой. Но и мы под водой. Но сначала чуть-чуть сверху. Но это будет чуть-чуть. Но в океане. Под водой. Но зато под водой. И подольше подольше подольше под водой.\n\nЧто было бы, если бы Жириновский стал президентом?\n Жириновский тогда во Франции? В 2007 году в беседе с журналистами он высказал свою версию. А если бы стал президентом… Но ведь есть же и другие варианты. Это и другие варианты? Вот Жириновский ведь и во Франции баллотировался. Ведь и мэр там баллотировался. Ведь есть же другие. Ведь и есть другие кандидаты в президенты. Ведь это же не запрещено, не запрещено, не запрещено. И Туркмения же тоже баллотировалась, там много где-там. И Иран тоже не запрещают баллотировалась. И в Турции баллотировалась, там тоже много где-тамалась. Значит, там тоже есть альтернативные кандидаты. И Косово ведь не запрещено, там не запрещено. В Китае баллотировалась, а там много где-там запрещено. И в Европе. И там нельзя. И там можно баллотироваться.\n\nКогда Россия станет самой богатой и могущественной страной в мире?\n В ближайшие 10—20 лет. Если мы все ресурсы планеты используем. Это ресурсы нужно тратить в первую очередь на восстановление России, на возрождение российской державы. Потому что у нас самая сильная армия, самая мощная экономика и огромный ресурс. Мы единственная страна. Но ресурсов мало. Но мы не умеем ими правильно распоряжаться. Мы должны научиться правильно использовать их. Надо, и много чему научиться. А то опять проиграем. И опять проиграем. И в Сирии проигрывать. Будем опять проигрывать. Поэтому опять проиграем в очередной раз. Поэтому у Сирии проигрывать будем проигрывать. У нас опять проигрывать в очередной раз. И опять опять у Сирии проигрывать. И опять проиграем опять в очередной раз. Поэтому опять Сирия опять проиграем. У нас проиграем, опять у Украины будем проигрывать. Снова проиграем опять. И снова в очередной раз.\n\nДжордж, Джордж! Посмотри ковбойские фильмы!\n. Помолчи! Помолчи! Помоги мне выиграть! Помолчи! А то, что я там сказал, что я там сказал. Помолчи! Мне надо, что я там сказал! Помолчи! И помолчи! Помолчать и посмотри, что он там сказал! Давай играть в покер! Джордж!». Уколоть там кого-то там там. Помолчите! Помолчите, я сказал! Помолчите там! Помолчите и не мешайте мне играть в покер! Помолчите. Помолчите, кому надо — покер! А то, что хочу! Помолчите, кому не надо! Помолчите и помолчите.\n\nОт чего коровы с ума сходят? От британской демократии.\n А мы должны в парламенте сидеть и слушать. Помолчать! Помолчать, чтобы они там, что я сказал, кто там сказал. Помолчать и помолчать! О чем я сказал. Помолчать! А то, что надо делать, чтобы они там не сходили с ума! Помолчать! Молчать! Помолчать и помалкивать! А то, что мы там не хотим! Помолчать! Сидеть и помалкивать. Молчать, чтобы другие не сходили с ума. Сидеть в нашем парламенте. Пока другие там не сходили с разума! Помолчать! Помалкивать и помалкивать, где там не сходили с него. А то, что другие не сходили с разума. Молчать и не мешают. Помолчать! Сидят и помалкивают.\n" ] ] ]

[ "code", "raw", "markdown", "code", "markdown", "code" ]

[ [ "code", "code", "code", "code" ], [ "raw" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ] ]

[ "CC0-1.0" ]

[ "CC0-1.0" ]

[ "CC0-1.0" ]

[ [ [ "import pandas as pd\ndf = pd.DataFrame(\n {'A': [10, 20, 30, 40, 50], # 列Aとその値\n 'B': [0.8, 1.6, 2.4, 4.3, 7.6], # 列Bとその値\n 'C': [-1, -2.6, -3.5, -4.3, -5.1] }, # 列Cとその値\n index = ['row1', 'row2', 'row3', 'row4', 'row5'] # 行名を設定\n)", "_____no_output_____" ], [ "df", "_____no_output_____" ], [ "df['A']", "_____no_output_____" ], [ "df['B']", "_____no_output_____" ], [ "df['C']", "_____no_output_____" ], [ "df[['A', 'C']]", "_____no_output_____" ], [ "df[1 : 4]", "_____no_output_____" ], [ "df[: 2]", "_____no_output_____" ], [ "df['row1' : 'row3']", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "#\n# This script creates a plot with the 10 most used ingredients.\n#\n# The original recipe, contained in the 'ingredients' column, is cleaned as follow:\n#\n# - to lowecase\n# - replacing symbols\n# - removing digits\n# - stemming the words using the WordNetLemmatizer\n#\n# The ingredients should be cleaned mote, making 'low fat mozzarella' and \n# 'reduced fat mozzarella' the same ingredient. Ideas are welcome.\n# \n\nimport pandas as pd\nimport matplotlib.pyplot as plt\nfrom nltk.stem import WordNetLemmatizer\nfrom collections import Counter\n\n# Reading the data\ntrain = pd.read_json('../data/train.json')\nprint 'data loaded'\nstemmer = WordNetLemmatizer()\n#cachedStopWords = stopwords.words(\"english\")\n\n# Auxiliar function for cleaning\ndef clean_recipe(recipe):\n # To lowercase\n recipe = [ str.lower(i) for i in recipe ]\n\n # Remove some special characters\n # Individuals replace have a very good performance\n # http://stackoverflow.com/a/27086669/670873\n def replacing(i):\n i = i.replace('&', '').replace('(', '').replace(')','')\n i = i.replace('\\'', '').replace('\\\\', '').replace(',','')\n i = i.replace('.', '').replace('%', '').replace('/','')\n i = i.replace('\"', '')\n \n return i\n \n # Replacing characters\n recipe = [ replacing(i) for i in recipe ]\n \n # Remove digits\n recipe = [ i for i in recipe if not i.isdigit() ]\n \n # Stem ingredients\n recipe = [ stemmer.lemmatize(i) for i in recipe ]\n \n return recipe\n\n# The number of times each ingredient is used is stored in the 'sumbags' dictionary\nbags_of_words = [ Counter(clean_recipe(recipe)) for recipe in train.ingredients ]\nsumbags = sum(bags_of_words, Counter())\nprint 'plotting...'\n# Finally, plot the 10 most used ingredients\nplt.style.use(u'ggplot')\nfig = pd.DataFrame(sumbags, index=[0]).transpose()[0].sort(ascending=False, inplace=False)[:10].plot(kind='barh')\nfig.invert_yaxis()\nfig = fig.get_figure()\nfig.tight_layout()\nfig.savefig('10_most_used_ingredients.jpg')\n", "data loaded\n" ] ] ]

[ "code" ]

[ [ "code" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "# Copyright 2022 Google LLC.\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use this file except in compliance with the License.\n# You may obtain a copy of the License at\n#\n# https://www.apache.org/licenses/LICENSE-2.0\n#\n# Unless required by applicable law or agreed to in writing, software\n# distributed under the License is distributed on an \"AS IS\" BASIS,\n# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n# See the License for the specific language governing permissions and\n# limitations under the License.", "_____no_output_____" ] ], [ [ "# 4. Model Training\n\nThis notebook demonstrates how to train a Propensity Model using BigQuery ML.", "_____no_output_____" ], [ "### Requirements\n\n* Input features used for training needs to be stored as a BigQuery table. This can be done using [2. ML Data Preparation Notebook](2.ml_data_preparation.ipynb).", "_____no_output_____" ], [ "### Install and import required modules", "_____no_output_____" ] ], [ [ "# Uncomment to install required python modules\n# !sh ../utils/setup.sh", "_____no_output_____" ], [ "# Add custom utils module to Python environment\nimport os\nimport sys\nsys.path.append(os.path.abspath(os.pardir))\n\nfrom gps_building_blocks.cloud.utils import bigquery as bigquery_utils\n\nfrom utils import model\nfrom utils import helpers", "_____no_output_____" ] ], [ [ "### Set paramaters", "_____no_output_____" ] ], [ [ "configs = helpers.get_configs('config.yaml')\ndest_configs = configs.destination\n\n# GCP project ID\nPROJECT_ID = dest_configs.project_id\n# Name of the BigQuery dataset\nDATASET_NAME = dest_configs.dataset_name", "_____no_output_____" ], [ "# To distinguish the separate runs of the training pipeline\nRUN_ID = 'TRAIN_01'\n\n# BigQuery table name containing model development dataset\nFEATURES_DEV_TABLE = f'features_dev_table_{RUN_ID}'\n\n# BigQuery table name containing model testing dataset\nFEATURES_TEST_TABLE = f'features_test_table_{RUN_ID}'\n\n# Output model name to save in BigQuery\nMODEL_NAME = f'propensity_model_{RUN_ID}'", "_____no_output_____" ] ], [ [ "Next, let's configure modeling options.", "_____no_output_____" ], [ "### Model and features configuration", "_____no_output_____" ], [ "Model options can be configured in detail based on BigQuery ML specifications\nlisted in [The CREATE MODEL statement](https://cloud.google.com/bigquery-ml/docs/reference/standard-sql/bigqueryml-syntax-create).\n\n**NOTE**: Propensity modeling supports only following four types of models available in BigQuery ML:\n- LOGISTIC_REG\n- [AUTOML_CLASSIFIER](https://cloud.google.com/bigquery-ml/docs/reference/standard-sql/bigqueryml-syntax-create-automl)\n- [BOOSTED_TREE_CLASSIFIER](https://cloud.google.com/bigquery-ml/docs/reference/standard-sql/bigqueryml-syntax-create-boosted-tree)\n- [DNN_CLASSIFIER](https://cloud.google.com/bigquery-ml/docs/reference/standard-sql/bigqueryml-syntax-create-dnn-models)\n\nIn order to use specific model options, you can add options to following configuration exactly same as listed in the [The CREATE MODEL statement](https://cloud.google.com/bigquery-ml/docs/reference/standard-sql/bigqueryml-syntax-create). For example, if you want to trian [AUTOML_CLASSIFIER](https://cloud.google.com/bigquery-ml/docs/reference/standard-sql/bigqueryml-syntax-create-automl) with `BUDGET_HOURS=1`, you can specify it as:\n\n```python\nparams = {\n 'model_type': 'AUTOML_CLASSIFIER',\n 'budget_hours': 1\n}\n```", "_____no_output_____" ] ], [ [ "# Read in Features table schema to select feature names for model training\nsql = (\"SELECT column_name \"\n f\"FROM `{PROJECT_ID}.{DATASET_NAME}`.INFORMATION_SCHEMA.COLUMNS \"\n f\"WHERE table_name='{FEATURES_DEV_TABLE}';\")\n\nprint(sql)\nfeatures_schema = bq_utils.run_query(sql).to_dataframe()\n\n# Columns to remove from the feature list\nto_remove = ['window_start_ts', 'window_end_ts', 'snapshot_ts', 'user_id',\n 'label', 'key', 'data_split']\n\n# Selected features for model training\ntraining_features = [v for v in features_schema['column_name']\n if v not in to_remove]\n\nprint('Number of training features:', len(training_features))\nprint(training_features)", "_____no_output_____" ], [ "# Set parameters for AUTOML_CLASSIFIER model\n\nFEATURE_COLUMNS = training_features\nTARGET_COLUMN = 'label'\n\nparams = {\n 'model_path': f'{PROJECT_ID}.{DATASET_NAME}.{MODEL_NAME}',\n 'features_table_path': f'{PROJECT_ID}.{DATASET_NAME}.{FEATURES_DEV_TABLE}',\n 'feature_columns': FEATURE_COLUMNS,\n 'target_column': TARGET_COLUMN,\n 'MODEL_TYPE': 'AUTOML_CLASSIFIER',\n 'BUDGET_HOURS': 1.0,\n # Enable data_split_col if you want to use custom data split.\n # Details on AUTOML data split column:\n # https://cloud.google.com/automl-tables/docs/prepare#split\n # 'DATA_SPLIT_COL': 'data_split',\n 'OPTIMIZATION_OBJECTIVE': 'MAXIMIZE_AU_ROC'\n}", "_____no_output_____" ] ], [ [ "## Train the model\n\nFirst, we initialize `PropensityModel` with config parameters.", "_____no_output_____" ] ], [ [ "bq_utils = bigquery_utils.BigQueryUtils(project_id=PROJECT_ID)\npropensity_model = model.PropensityModel(bq_utils=bq_utils,\n params=params)", "_____no_output_____" ] ], [ [ "Next cell triggers model training job in BigQuery which takes some time to finish depending on dataset size and model complexity. Set `verbose=True`, if you want to verify training query details.", "_____no_output_____" ] ], [ [ "propensity_model.train(verbose=False)", "_____no_output_____" ] ], [ [ "Following cell allows you to see detailed information about the input features used to train a model. It provides following columns:\n- input — The name of the column in the input training data.\n- min — The sample minimum. This column is NULL for non-numeric inputs.\n- max — The sample maximum. This column is NULL for non-numeric inputs.\n- mean — The average. This column is NULL for non-numeric inputs.\n- stddev — The standard deviation. This column is NULL for non-numeric inputs.\n- category_count — The number of categories. This column is NULL for non-categorical columns.\n- null_count — The number of NULLs.\n\nFor more details refer to [help page](https://cloud.google.com/bigquery-ml/docs/reference/standard-sql/bigqueryml-syntax-feature).", "_____no_output_____" ] ], [ [ "propensity_model.get_feature_info()", "_____no_output_____" ] ], [ [ "### Evaluate the model\nThis section helps to do quick model evaluation to get following model metrics:\n\n* recall\n* accuracy\n* f1_score\n* log_loss\n* roc_auc\n\nTwo optional parameters can be specified for evaluation:\n\n* eval_table: BigQuery table containing evaluation dataset\n* threshold: Custom probability threshold to be used for evaluation (to binarize the predictions). Default value is 0.5.\n\nIf neither of these options are specified, the model is evaluated using evaluation dataset split during training with default threshold of 0.5.\n\n**NOTE:** This evaluation provides basic model performance metrics. For thorough evaluation refer to [5. Model evaluation notebook](5.model_evaluation_and_diagnostics.ipynb) notebook.\n\nTODO(): Add sql code to calculate the proportion of positive examples in the evaluation dataset to be used as the *threshold*.", "_____no_output_____" ] ], [ [ "# Model performance on the model development dataset on which the final\n# model has been trained\n\nEVAL_TABLE_NAME = FEATURES_DEV_TABLE\n\neval_params = {\n 'eval_table_path': f'{PROJECT_ID}.{DATASET_NAME}.{EVAL_TABLE_NAME}',\n 'threshold': 0.5\n}\npropensity_model.evaluate(eval_params, verbose=False)", "_____no_output_____" ], [ "# Model performance on the held out test dataset\n\nEVAL_TABLE_NAME = FEATURES_TEST_TABLE\n\neval_params = {\n 'eval_table_path': f'{PROJECT_ID}.{DATASET_NAME}.{EVAL_TABLE_NAME}',\n 'threshold': 0.5\n}\npropensity_model.evaluate(eval_params, verbose=False)", "_____no_output_____" ] ], [ [ "## Next", "_____no_output_____" ], [ "Use [5. Model evaluation notebook](5.model_evaluation_and_diagnostics.ipynb) to get detailed performance metrics of the model and decide of model actually solves the business problem.", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "%load_ext autoreload\n%autoreload 2", "The autoreload extension is already loaded. To reload it, use:\n %reload_ext autoreload\n" ], [ "import jax.numpy as jnp\nimport matplotlib.pyplot as plt\nimport jax\nfrom jax import lax\nfrom deluca.envs import CartPole\nfrom deluca.agents import Deep", "[autoreload of deluca.agents._deep failed: Traceback (most recent call last):\n File \"/Users/alexjyu/miniconda3/lib/python3.7/site-packages/IPython/extensions/autoreload.py\", line 245, in check\n superreload(m, reload, self.old_objects)\n File \"/Users/alexjyu/miniconda3/lib/python3.7/site-packages/IPython/extensions/autoreload.py\", line 410, in superreload\n update_generic(old_obj, new_obj)\n File \"/Users/alexjyu/miniconda3/lib/python3.7/site-packages/IPython/extensions/autoreload.py\", line 347, in update_generic\n update(a, b)\n File \"/Users/alexjyu/miniconda3/lib/python3.7/site-packages/IPython/extensions/autoreload.py\", line 302, in update_class\n if update_generic(old_obj, new_obj): continue\n File \"/Users/alexjyu/miniconda3/lib/python3.7/site-packages/IPython/extensions/autoreload.py\", line 347, in update_generic\n update(a, b)\n File \"/Users/alexjyu/miniconda3/lib/python3.7/site-packages/IPython/extensions/autoreload.py\", line 266, in update_function\n setattr(old, name, getattr(new, name))\nValueError: __init__() requires a code object with 1 free vars, not 0\n]\n" ], [ "def loop(context, x):\n env, agent = context\n control = agent(env.state)\n _, reward, _, _ = env.step(control)\n # agent.feed(reward)\n # agent.update()\n return (env, agent), reward", "_____no_output_____" ], [ "# Deep\nenv = CartPole()\nagent = Deep(\n env_state_size = 4,\n action_space = jnp.array([0,1]),\n learning_rate = 0.1,\n gamma = 0.99,\n max_episode_length = 500,\n seed = 0\n )\n", "_____no_output_____" ], [ " # for loop version\nT = 100\nxs = jnp.array(jnp.arange(T))\nprint(env.reset())\nreward = 0\nfor i in range(T):\n (env, agent), r = loop((env, agent), 0)\n reward += r\nreward_forloop = reward\nprint('reward_forloop = ' + str(reward_forloop))\n\n\n# scan version\nenv = CartPole()\nagent = Deep(\n env_state_size = 4,\n action_space = jnp.array([0,1]),\n learning_rate = 0.1,\n gamma = 0.99,\n max_episode_length = 500,\n seed = 0\n )\nprint(env.reset())\n_,reward_scan = lax.scan(loop, (env, agent), xs)\n\n# correctness test\nprint('reward_scan = ' + str(reward_scan))\nprint('reward_scan sum = ' + str(jnp.sum(reward_scan)))", "[ 0.00322265 -0.01503431 -0.01464135 0.04524388]\nreward_forloop = 52\n[ 0.00322265 -0.01503431 -0.01464135 0.04524388]\nreward_scan = [1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]\nreward_scan sum = 52\n" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import torch\nimport torch.nn as nn\n\nimport onmt\nimport onmt.inputters\nimport onmt.modules\nimport onmt.utils", "_____no_output_____" ] ], [ [ "We begin by loading in the vocabulary for the model of interest. This will let us check vocab size and to get the special ids for padding.", "_____no_output_____" ] ], [ [ "vocab = dict(torch.load(\"../../data/data.vocab.pt\"))\nsrc_padding = vocab[\"src\"].stoi[onmt.inputters.PAD_WORD]\ntgt_padding = vocab[\"tgt\"].stoi[onmt.inputters.PAD_WORD]", "_____no_output_____" ] ], [ [ "Next we specify the core model itself. Here we will build a small model with an encoder and an attention based input feeding decoder. Both models will be RNNs and the encoder will be bidirectional", "_____no_output_____" ] ], [ [ "emb_size = 10\nrnn_size = 6\n# Specify the core model. \nencoder_embeddings = onmt.modules.Embeddings(emb_size, len(vocab[\"src\"]),\n word_padding_idx=src_padding)\n\nencoder = onmt.encoders.RNNEncoder(hidden_size=rnn_size, num_layers=1, \n rnn_type=\"LSTM\", bidirectional=True,\n embeddings=encoder_embeddings)\n\ndecoder_embeddings = onmt.modules.Embeddings(emb_size, len(vocab[\"tgt\"]),\n word_padding_idx=tgt_padding)\ndecoder = onmt.decoders.decoder.InputFeedRNNDecoder(hidden_size=rnn_size, num_layers=1, \n bidirectional_encoder=True,\n rnn_type=\"LSTM\", embeddings=decoder_embeddings)\nmodel = onmt.models.model.NMTModel(encoder, decoder)\n\n# Specify the tgt word generator and loss computation module\nmodel.generator = nn.Sequential( \n nn.Linear(rnn_size, len(vocab[\"tgt\"])), \n nn.LogSoftmax())\nloss = onmt.utils.loss.NMTLossCompute(model.generator, vocab[\"tgt\"]) ", "_____no_output_____" ] ], [ [ "Now we set up the optimizer. This could be a core torch optim class, or our wrapper which handles learning rate updates and gradient normalization automatically.", "_____no_output_____" ] ], [ [ "optim = onmt.utils.optimizers.Optimizer(method=\"sgd\", lr=1, max_grad_norm=2)\noptim.set_parameters(model.named_parameters())", "_____no_output_____" ] ], [ [ "Now we load the data from disk. Currently will need to call a function to load the fields into the data as well. ", "_____no_output_____" ] ], [ [ "# Load some data\ndata = torch.load(\"../../data/data.train.1.pt\")\nvalid_data = torch.load(\"../../data/data.valid.1.pt\")\ndata.load_fields(vocab)\nvalid_data.load_fields(vocab)\ndata.examples = data.examples[:100] ", "_____no_output_____" ] ], [ [ "To iterate through the data itself we use a torchtext iterator class. We specify one for both the training and test data. ", "_____no_output_____" ] ], [ [ "train_iter = onmt.inputters.OrderedIterator( \n dataset=data, batch_size=10, \n device=-1, \n repeat=False)\nvalid_iter = onmt.inputters.OrderedIterator( \n dataset=valid_data, batch_size=10, \n device=-1,\n train=False) ", "_____no_output_____" ] ], [ [ "Finally we train.", "_____no_output_____" ] ], [ [ "trainer = onmt.Trainer(model, loss, loss, optim)\n\ndef report_func(*args):\n stats = args[-1]\n stats.output(args[0], args[1], 10, 0)\n return stats\n\nfor epoch in range(2):\n trainer.train(epoch, report_func)\n val_stats = trainer.validate()\n\n print(\"Validation\")\n val_stats.output(epoch, 11, 10, 0)\n trainer.epoch_step(val_stats.ppl(), epoch)", "Epoch 0, 0/ 10; acc: 0.00; ppl: 1225.23; 1320 src tok/s; 1320 tgt tok/s; 1514090454 s elapsed\nEpoch 0, 1/ 10; acc: 9.50; ppl: 996.33; 1188 src tok/s; 1194 tgt tok/s; 1514090454 s elapsed\nEpoch 0, 2/ 10; acc: 16.51; ppl: 694.48; 1265 src tok/s; 1267 tgt tok/s; 1514090454 s elapsed\nEpoch 0, 3/ 10; acc: 20.49; ppl: 470.39; 1459 src tok/s; 1420 tgt tok/s; 1514090454 s elapsed\nEpoch 0, 4/ 10; acc: 22.68; ppl: 387.03; 1511 src tok/s; 1462 tgt tok/s; 1514090454 s elapsed\nEpoch 0, 5/ 10; acc: 24.58; ppl: 345.44; 1625 src tok/s; 1509 tgt tok/s; 1514090454 s elapsed\nEpoch 0, 6/ 10; acc: 25.37; ppl: 314.39; 1586 src tok/s; 1493 tgt tok/s; 1514090454 s elapsed\nEpoch 0, 7/ 10; acc: 26.14; ppl: 291.15; 1593 src tok/s; 1520 tgt tok/s; 1514090455 s elapsed\nEpoch 0, 8/ 10; acc: 26.32; ppl: 274.79; 1606 src tok/s; 1545 tgt tok/s; 1514090455 s elapsed\nEpoch 0, 9/ 10; acc: 26.83; ppl: 247.32; 1669 src tok/s; 1614 tgt tok/s; 1514090455 s elapsed\nValidation\nEpoch 0, 11/ 10; acc: 13.41; ppl: 111.94; 0 src tok/s; 7329 tgt tok/s; 1514090464 s elapsed\nEpoch 1, 0/ 10; acc: 6.59; ppl: 147.05; 1849 src tok/s; 1743 tgt tok/s; 1514090464 s elapsed\nEpoch 1, 1/ 10; acc: 22.10; ppl: 130.66; 2002 src tok/s; 1957 tgt tok/s; 1514090464 s elapsed\nEpoch 1, 2/ 10; acc: 20.16; ppl: 122.49; 1748 src tok/s; 1760 tgt tok/s; 1514090464 s elapsed\nEpoch 1, 3/ 10; acc: 23.52; ppl: 117.41; 1690 src tok/s; 1698 tgt tok/s; 1514090464 s elapsed\nEpoch 1, 4/ 10; acc: 24.16; ppl: 119.42; 1647 src tok/s; 1662 tgt tok/s; 1514090464 s elapsed\nEpoch 1, 5/ 10; acc: 25.44; ppl: 115.31; 1775 src tok/s; 1709 tgt tok/s; 1514090465 s elapsed\nEpoch 1, 6/ 10; acc: 24.05; ppl: 115.11; 1780 src tok/s; 1718 tgt tok/s; 1514090465 s elapsed\nEpoch 1, 7/ 10; acc: 25.32; ppl: 109.59; 1799 src tok/s; 1765 tgt tok/s; 1514090465 s elapsed\nEpoch 1, 8/ 10; acc: 25.14; ppl: 108.16; 1771 src tok/s; 1734 tgt tok/s; 1514090465 s elapsed\nEpoch 1, 9/ 10; acc: 25.58; ppl: 107.13; 1817 src tok/s; 1757 tgt tok/s; 1514090465 s elapsed\nValidation\nEpoch 1, 11/ 10; acc: 19.58; ppl: 88.09; 0 src tok/s; 7371 tgt tok/s; 1514090474 s elapsed\n" ] ], [ [ "To use the model, we need to load up the translation functions ", "_____no_output_____" ] ], [ [ "import onmt.translate", "_____no_output_____" ], [ "translator = onmt.translate.Translator(beam_size=10, fields=data.fields, model=model)\nbuilder = onmt.translate.TranslationBuilder(data=valid_data, fields=data.fields)\n\nvalid_data.src_vocabs\nfor batch in valid_iter:\n trans_batch = translator.translate_batch(batch=batch, data=valid_data)\n translations = builder.from_batch(trans_batch)\n for trans in translations:\n print(trans.log(0))\n break", "PRED SCORE: -4.0690\n\nSENT 0: ('The', 'competitors', 'have', 'other', 'advantages', ',', 'too', '.')\nPRED 0: .\n\nPRED SCORE: -4.2736\n\nSENT 0: ('The', 'company', '&apos;s', 'durability', 'goes', 'back', 'to', 'its', 'first', 'boss', ',', 'a', 'visionary', ',', 'Thomas', 'J.', 'Watson', 'Sr.')\nPRED 0: .\n\nPRED SCORE: -4.0144\n\nSENT 0: ('&quot;', 'From', 'what', 'we', 'know', 'today', ',', 'you', 'have', 'to', 'ask', 'how', 'I', 'could', 'be', 'so', 'wrong', '.', '&quot;')\nPRED 0: .\n\nPRED SCORE: -4.1361\n\nSENT 0: ('Boeing', 'Co', 'shares', 'rose', '1.5%', 'to', '$', '67.94', '.')\nPRED 0: .\n\nPRED SCORE: -4.1382\n\nSENT 0: ('Some', 'did', 'not', 'believe', 'him', ',', 'they', 'said', 'that', 'he', 'got', 'dizzy', 'even', 'in', 'the', 'truck', ',', 'but', 'always', 'wanted', 'to', 'fulfill', 'his', 'dream', ',', 'that', 'of', 'becoming', 'a', 'pilot', '.')\nPRED 0: .\n\nPRED SCORE: -3.8881\n\nSENT 0: ('In', 'your', 'opinion', ',', 'the', 'council', 'should', 'ensure', 'that', 'the', 'band', 'immediately', 'above', 'the', 'Ronda', 'de', 'Dalt', 'should', 'provide', 'in', 'its', 'entirety', ',', 'an', 'area', 'of', 'equipment', 'to', 'conduct', 'a', 'smooth', 'transition', 'between', 'the', 'city', 'and', 'the', 'green', '.')\nPRED 0: .\n\nPRED SCORE: -4.0778\n\nSENT 0: ('The', 'clerk', 'of', 'the', 'court', ',', 'Jorge', 'Yanez', ',', 'went', 'to', 'the', 'jail', 'of', 'the', 'municipality', 'of', 'San', 'Nicolas', 'of', 'Garza', 'to', 'notify', 'Jonah', 'that', 'he', 'has', 'been', 'legally', 'pardoned', 'and', 'his', 'record', 'will', 'be', 'filed', '.')\nPRED 0: .\n\nPRED SCORE: -4.2479\n\nSENT 0: ('&quot;', 'In', 'a', 'research', 'it', 'is', 'reported', 'that', 'there', 'are', 'no', 'parts', 'or', 'components', 'of', 'the', 'ship', 'in', 'another', 'place', ',', 'the', 'impact', 'is', 'presented', 'in', 'a', 'structural', 'way', '.')\nPRED 0: .\n\nPRED SCORE: -3.8585\n\nSENT 0: ('On', 'the', 'asphalt', 'covering', ',', 'he', 'added', ',', 'is', 'placed', 'a', 'final', 'layer', 'called', 'rolling', 'covering', ',', 'which', 'is', 'made', '\\u200b', '\\u200b', 'of', 'a', 'fine', 'stone', 'material', ',', 'meaning', 'sand', 'also', 'dipped', 'into', 'the', 'asphalt', '.')\nPRED 0: .\n\nPRED SCORE: -4.2298\n\nSENT 0: ('This', 'is', '200', 'bar', 'on', 'leaving', 'and', '100', 'bar', 'on', 'arrival', '.')\nPRED 0: .\n\n" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# 决策树\n\n- 非参数学习算法\n- 天然解决多分类问题\n- 也可以解决回归问题\n- 非常好的可解释性", "_____no_output_____" ] ], [ [ "import numpy as np\nimport matplotlib.pyplot as plt\nfrom sklearn.model_selection import cross_val_score", "_____no_output_____" ], [ "from sklearn import datasets\n\niris = datasets.load_iris()", "_____no_output_____" ], [ "print(iris.DESCR)", ".. _iris_dataset:\n\nIris plants dataset\n--------------------\n\n**Data Set Characteristics:**\n\n :Number of Instances: 150 (50 in each of three classes)\n :Number of Attributes: 4 numeric, predictive attributes and the class\n :Attribute Information:\n - sepal length in cm\n - sepal width in cm\n - petal length in cm\n - petal width in cm\n - class:\n - Iris-Setosa\n - Iris-Versicolour\n - Iris-Virginica\n \n :Summary Statistics:\n\n ============== ==== ==== ======= ===== ====================\n Min Max Mean SD Class Correlation\n ============== ==== ==== ======= ===== ====================\n sepal length: 4.3 7.9 5.84 0.83 0.7826\n sepal width: 2.0 4.4 3.05 0.43 -0.4194\n petal length: 1.0 6.9 3.76 1.76 0.9490 (high!)\n petal width: 0.1 2.5 1.20 0.76 0.9565 (high!)\n ============== ==== ==== ======= ===== ====================\n\n :Missing Attribute Values: None\n :Class Distribution: 33.3% for each of 3 classes.\n :Creator: R.A. Fisher\n :Donor: Michael Marshall (MARSHALL%[email protected])\n :Date: July, 1988\n\nThe famous Iris database, first used by Sir R.A. Fisher. The dataset is taken\nfrom Fisher's paper. Note that it's the same as in R, but not as in the UCI\nMachine Learning Repository, which has two wrong data points.\n\nThis is perhaps the best known database to be found in the\npattern recognition literature. Fisher's paper is a classic in the field and\nis referenced frequently to this day. (See Duda & Hart, for example.) The\ndata set contains 3 classes of 50 instances each, where each class refers to a\ntype of iris plant. One class is linearly separable from the other 2; the\nlatter are NOT linearly separable from each other.\n\n.. topic:: References\n\n - Fisher, R.A. \"The use of multiple measurements in taxonomic problems\"\n Annual Eugenics, 7, Part II, 179-188 (1936); also in \"Contributions to\n Mathematical Statistics\" (John Wiley, NY, 1950).\n - Duda, R.O., & Hart, P.E. (1973) Pattern Classification and Scene Analysis.\n (Q327.D83) John Wiley & Sons. ISBN 0-471-22361-1. See page 218.\n - Dasarathy, B.V. (1980) \"Nosing Around the Neighborhood: A New System\n Structure and Classification Rule for Recognition in Partially Exposed\n Environments\". IEEE Transactions on Pattern Analysis and Machine\n Intelligence, Vol. PAMI-2, No. 1, 67-71.\n - Gates, G.W. (1972) \"The Reduced Nearest Neighbor Rule\". IEEE Transactions\n on Information Theory, May 1972, 431-433.\n - See also: 1988 MLC Proceedings, 54-64. Cheeseman et al\"s AUTOCLASS II\n conceptual clustering system finds 3 classes in the data.\n - Many, many more ...\n" ], [ "X = iris.data[:, 2:] # 取后两个特征\ny = iris.target", "_____no_output_____" ], [ "plt.scatter(X[y==0, 0], X[y==0, 1])\nplt.scatter(X[y==1, 0], X[y==1, 1])\nplt.scatter(X[y==2, 0], X[y==2, 1])", "_____no_output_____" ] ], [ [ "### 1. scikit-learn 中的决策树", "_____no_output_____" ] ], [ [ "from sklearn.tree import DecisionTreeClassifier\n# entropy : 熵\ndt_clf = DecisionTreeClassifier(max_depth=3, criterion=\"entropy\")\ndt_clf.fit(X, y)", "_____no_output_____" ], [ "def plot_decision_boundary(model, axis):\n \n x0, x1 = np.meshgrid(\n np.linspace(axis[0], axis[1], int((axis[1] - axis[0])*100)).reshape(1, -1),\n np.linspace(axis[2], axis[3], int((axis[3] - axis[2])*100)).reshape(-1, 1)\n )\n X_new = np.c_[x0.ravel(), x1.ravel()]\n \n y_predic = model.predict(X_new)\n zz = y_predic.reshape(x0.shape)\n \n from matplotlib.colors import ListedColormap\n custom_cmap = ListedColormap(['#EF9A9A', '#FFF590', '#90CAF9'])\n \n plt.contourf(x0, x1, zz, linewidth=5, cmap=custom_cmap)", "_____no_output_____" ], [ "plot_decision_boundary(dt_clf, axis=(0.5, 7.5, 0, 3))\nplt.scatter(X[y==0, 0], X[y==0, 1])\nplt.scatter(X[y==1, 0], X[y==1, 1])\nplt.scatter(X[y==2, 0], X[y==2, 1])", "/home/js/pyEnvs/tf_cpu/lib/python3.6/site-packages/ipykernel_launcher.py:15: UserWarning: The following kwargs were not used by contour: 'linewidth'\n from ipykernel import kernelapp as app\n" ] ], [ [ "### 2. 如何构建决策树\n\n**问题**\n\n- 每个节点在那个维度做划分？\n- 某个维度在那个值上做划分？\n\n- 划分的标准就是，**划分后使得信息熵降低**", "_____no_output_____" ], [ "**信息熵**\n\n- 熵在信息论中代表随机变量不确定的度量\n- 熵越大，数据的不确定性越高\n- 熵越小，数据的不确定性越低\n\n$$H = -\\sum_{i=1}^kp_i\\log{(p_i)}$$\n- 其中 $p_i$ 表示每一类信息在所有信息类别中占的比例\n\n- 对于二分类，香农公式为：\n$$H=-x\\log(x)-(1-x)\\log(1-x)$$", "_____no_output_____" ], [ "**信息熵函数**", "_____no_output_____" ] ], [ [ "def entropy(p):\n return -p * np.log(p) - (1-p) * np.log(1-p)", "_____no_output_____" ], [ "x = np.linspace(0.01, 0.99)", "_____no_output_____" ], [ "plt.plot(x, entropy(x))", "_____no_output_____" ] ], [ [ "- 可以看出，当 x 越接近0.5，熵越高", "_____no_output_____" ], [ "### 3. 模拟使用信息熵进行划分", "_____no_output_____" ] ], [ [ "# 基于维度 d 的 value 值进行划分\ndef split(X, y, d, value):\n index_a = (X[:, d] <= value)\n index_b = (X[:, d] > value)\n return X[index_a], X[index_b], y[index_a], y[index_b]", "_____no_output_____" ], [ "from collections import Counter\nfrom math import log\n# 计算每一类样本点的熵的和\ndef entropy(y):\n counter = Counter(y)\n res = 0.0\n for num in counter.values():\n p = num / len(y)\n res += -p * log(p)\n return res ", "_____no_output_____" ], [ "# 寻找要划分的 value 值，寻找最小信息熵及相应的点\ndef try_split(X, y):\n best_entropy = float('inf') # 最小的熵的值\n best_d, best_v = -1, -1 # 划分的维度，划分的位置\n # 遍历每一个维度\n for d in range(X.shape[1]):\n # 每两个样本点在 d 这个维度中间的值. 首先把 d 维所有样本排序\n sorted_index = np.argsort(X[:, d])\n for i in range(1, len(X)):\n if X[sorted_index[i-1], d] != X[sorted_index[i], d]:\n v = (X[sorted_index[i-1], d] + X[sorted_index[i], d]) / 2\n x_l, x_r, y_l, y_r = split(X, y, d, v)\n # 计算当前划分后的两部分结果熵是多少\n e = entropy(y_l) + entropy(y_r)\n if e < best_entropy: \n best_entropy, best_d, best_v = e, d, v\n return best_entropy, best_d, best_v", "_____no_output_____" ], [ "best_entropy, best_d, best_v = try_split(X, y)\nprint(\"best_entropy = \", best_entropy)\nprint(\"best_d\", best_d)\nprint(\"best_v\", best_v)", "best_entropy = 0.6931471805599453\nbest_d 0\nbest_v 2.45\n" ] ], [ [ "**即在第 0 个维度的 2.45 位置进行划分，可以得到最低的熵，值为 0.693**", "_____no_output_____" ] ], [ [ "X1_l, X1_r, y1_l, y1_r = split(X, y, best_d, best_v)", "_____no_output_____" ], [ "entropy(y1_r)", "_____no_output_____" ], [ "entropy(y1_l) # 从上图可以看出，粉红色部分只有一类，故熵为 0", "_____no_output_____" ], [ "best_entropy2, best_d2, best_v2 = try_split(X1_r, y1_r)\nprint(\"best_entropy = \", best_entropy2)\nprint(\"best_d\", best_d2)\nprint(\"best_v\", best_v2)", "best_entropy = 0.4132278899361904\nbest_d 1\nbest_v 1.75\n" ], [ "X2_l, X2_r, y2_l, y2_r = split(X1_r, y1_r, best_d2, best_v2)", "_____no_output_____" ], [ "entropy(y2_r)", "_____no_output_____" ], [ "entropy(y2_l)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import pandas as pd\nimport numpy as np\n\nimport fml\n%matplotlib inline", "_____no_output_____" ] ], [ [ "We compute here the number of labels concurrent at t.", "_____no_output_____" ] ], [ [ "df_price = fml.DataHandler().get_time_series_data(\"SPY\")\nprice = fml.Price(df_price[\"Adj Close\"], freq=\"B\")", "_____no_output_____" ], [ "cross_time = price.bounds_cross_time(window=10)", "_____no_output_____" ], [ "cross_time.index = range(len(cross_time))\ncross_time.head()", "_____no_output_____" ], [ "cross_time_int = cross_time.min(axis=1).to_frame(\"cross\")\ncross_time_int[\"cnt\"] = 0\ncross_time_int.head(10)", "_____no_output_____" ], [ "for i in cross_time_int.index[:20]:\n cross_time_int.loc[i + 1: i + cross_time_int.loc[i, \"cross\"], \"cnt\"] += 1\ncross_time_int.head(10)", "_____no_output_____" ], [ "window = 10\nfor i in range(1, window + 2):\n cross_time_int[i] = cross_time_int[\"cross\"].shift(i) >= i", "_____no_output_____" ], [ "cross_time_int[\"cnt_vec\"] = cross_time_int[list(range(1, window + 2))].sum(axis=1)\nseries = price.num_concurrent_events()", "_____no_output_____" ], [ "cross_time_int[series.name] = series.values", "_____no_output_____" ], [ "cross_time_int.head(15)", "_____no_output_____" ] ] ]

[ "code", "markdown", "code" ]

[ [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "import pandas as pd\nimport numpy as np\n\nfrom PIL import Image\nimport os\nimport sys\n!pip install ipython-autotime\n\n%load_ext autotime\n%matplotlib inline\n", "Requirement already satisfied: ipython-autotime in /usr/local/lib/python3.6/dist-packages (0.3.1)\nRequirement already satisfied: ipython in /usr/local/lib/python3.6/dist-packages (from ipython-autotime) (5.5.0)\nRequirement already satisfied: pickleshare in /usr/local/lib/python3.6/dist-packages (from ipython->ipython-autotime) (0.7.5)\nRequirement already satisfied: prompt-toolkit<2.0.0,>=1.0.4 in /usr/local/lib/python3.6/dist-packages (from ipython->ipython-autotime) (1.0.18)\nRequirement already satisfied: pygments in /usr/local/lib/python3.6/dist-packages (from ipython->ipython-autotime) (2.6.1)\nRequirement already satisfied: traitlets>=4.2 in /usr/local/lib/python3.6/dist-packages (from ipython->ipython-autotime) (4.3.3)\nRequirement already satisfied: simplegeneric>0.8 in /usr/local/lib/python3.6/dist-packages (from ipython->ipython-autotime) (0.8.1)\nRequirement already satisfied: setuptools>=18.5 in /usr/local/lib/python3.6/dist-packages (from ipython->ipython-autotime) (51.3.3)\nRequirement already satisfied: pexpect; sys_platform != \"win32\" in /usr/local/lib/python3.6/dist-packages (from ipython->ipython-autotime) (4.8.0)\nRequirement already satisfied: decorator in /usr/local/lib/python3.6/dist-packages (from ipython->ipython-autotime) (4.4.2)\nRequirement already satisfied: wcwidth in /usr/local/lib/python3.6/dist-packages (from prompt-toolkit<2.0.0,>=1.0.4->ipython->ipython-autotime) (0.2.5)\nRequirement already satisfied: six>=1.9.0 in /usr/local/lib/python3.6/dist-packages (from prompt-toolkit<2.0.0,>=1.0.4->ipython->ipython-autotime) (1.15.0)\nRequirement already satisfied: ipython-genutils in /usr/local/lib/python3.6/dist-packages (from traitlets>=4.2->ipython->ipython-autotime) (0.2.0)\nRequirement already satisfied: ptyprocess>=0.5 in /usr/local/lib/python3.6/dist-packages (from pexpect; sys_platform != \"win32\"->ipython->ipython-autotime) (0.7.0)\nThe autotime extension is already loaded. To reload it, use:\n %reload_ext autotime\ntime: 2.43 s (started: 2021-01-22 18:15:43 +00:00)\n" ] ], [ [ "1. Extract your dataset and split into train_x, train_y, test_x and test_y.\r\n2. Execute the following cells", "_____no_output_____" ], [ "\n\n---\n\n\n## Hybrid Social Group Optimization \n\n\n---\n\n\n", "_____no_output_____" ] ], [ [ "N = 5 # Number of persons in population\nD = len(train_x.columns) # Number of features in dataset\ng = 10 # Number of generations\nc = 0.6 # Self Introspection factor\nr0 = 1\nr1 = 0.4\nr2 = 0.6\nprint(r1, r2)", "0.4 0.6\ntime: 3.13 ms (started: 2021-01-22 18:19:07 +00:00)\n" ] ], [ [ "**Population Initialization**", "_____no_output_____" ] ], [ [ "population = np.random.choice([0,1,2,3,4,5,6,7,8,9], (N,D), p=[0.16, 0.16, 0.16, 0.16, 0.16, 0.04, 0.04, 0.04, 0.04, 0.04]) #Determines no. of features selected by probablity", "time: 5.05 ms (started: 2021-01-22 18:19:07 +00:00)\n" ], [ "population = population.astype(float)\nprint(population.shape)\npopulation", "(5, 10000)\n" ], [ "fitness = np.zeros(N)\ntest_x.shape", "_____no_output_____" ], [ "def fitter_trait(X_old, X_new):\n if X_new > X_old:\n return X_old\n else:\n return X_new\n \n ", "time: 2.58 ms (started: 2021-01-22 18:19:07 +00:00)\n" ] ], [ [ "\n\n\n\n\n**Fitness Function**", "_____no_output_____" ] ], [ [ "global classifier\nclassifier = Svc #Change classifier here\nselect = train_x.columns\nselectno = len(train_x.columns)\nclassifier.fit(train_x, train_y)\nselect_acc = classifier.score(test_x, test_y) \ndef fitness_function(pop): #Fitness Function\n \n for i in range(N):\n new_train_x = train_x\n \n new_test_x = test_x\n \n global select\n global selectno\n global select_acc\n \n new_train_x = new_train_x.drop(train_x.columns[pop[i] < 4], axis = 1)\n \n new_test_x = new_test_x.drop(test_x.columns[pop[i] < 4], axis = 1)\n \n classifier.fit(new_train_x, train_y) \n fitness[i] = classifier.score(new_test_x, test_y) \n if (fitness[i] > select_acc):\n select = new_train_x.columns\n # print(select.shape)\n selectno = new_train_x.shape[1]\n select_acc = fitness[i]\n elif fitness[i] == select_acc and new_train_x.shape[1] < selectno:\n select = new_train_x.columns\n selectno = new_train_x.shape[1]\n \n print(\"\\nPerson \"+ str(i+1))\n \n print(\"No. of Features Used = \"+ str(new_train_x.shape[1])+ \"/\"+str(D)+\"\\nFitness = \" + str(fitness[i]))\n print(\"Feature Used = \", end = \" \") \n \n #print(new_train_x.columns)\n\n print(new_train_x.shape[1])\n ", "time: 5.47 s (started: 2021-01-22 18:19:07 +00:00)\n" ], [ "# Initializing Fitness values of population\n# fitness_function(population) \n# selectno", "time: 785 µs (started: 2021-01-22 18:19:13 +00:00)\n" ] ], [ [ "**Gbest : Fittest person in population**", "_____no_output_____" ] ], [ [ "###Determining GBest\ngbest = 0\ngbest_i = 0", "time: 874 µs (started: 2021-01-22 18:19:13 +00:00)\n" ], [ "def find_gbest():\n gbest = max(fitness)#This can be any function\n gbest_i = fitness.argmax()\n print(\"Best fitness value for the generation = \"+str(gbest) + \" Person \" + str(gbest_i+1)+\"\\n\")\nfind_gbest() \n#we chose maximum fitness value to be better for simplicity\ndef cal_fitness(person):\n new_train_x = train_x\n \n new_test_x = test_x\n\n new_train_x = new_train_x.drop(train_x.columns[person < 4], axis = 1)\n \n new_test_x = new_test_x.drop(test_x.columns[person < 4], axis = 1)\n \n classifier.fit(new_train_x, train_y) \n return classifier.score(new_test_x, test_y)\n\ncal_fitness(population[0])", "Best fitness value for the generation = 0.0 Person 1\n\n" ], [ "# new_train_x = train_x\r\n \r\n# new_test_x = test_x\r\n\r\n# new_train_x = new_train_x.drop(train_x.columns[person < 4], axis = 1)\r\n \r\n# new_test_x = new_test_x.drop(test_x.columns[person < 4], axis = 1)", "time: 913 µs (started: 2021-01-22 18:19:15 +00:00)\n" ], [ "per1 = np.zeros((1,10000))\nprint(per1.shape)\nper1[0][5] = 8\nper1[0][89] = 7\nper1[0][45] = 6\ncal_fitness(per1[0])", "(1, 10000)\n" ] ], [ [ "\n\n---\n\n**Mutation Phase**", "_____no_output_____" ] ], [ [ "def mutate():\n gworst_i = fitness.argmin()\n gworst = min(fitness)\n mut = np.random.randint(0,2,size=(1,D))[0]\n print(\"Mutating the Generation's Worst....Person \"+ str(gworst_i+1))\n for i in range(D):\n if mut[i] > 0:\n mut[i] = population[gbest_i][i]\n else:\n mut[i] = population[gworst_i][i]\n if cal_fitness(mut) > gworst:\n population[gworst_i] = mut\n print(\"Person \"+str(gworst_i)+\" mutated\")\n else:\n print(\"No Mutations in this generation\")", "time: 5.35 ms (started: 2021-01-22 18:19:15 +00:00)\n" ], [ "mut = np.random.randint(0,2,size=(1,D))[0]\r\nmut", "_____no_output_____" ], [ "div = pd.DataFrame(np.random.randint(0,2,size=(1,D))[0])\n# div.iloc[:,div > 0] = population[2][div>0]\n# div", "time: 1.79 ms (started: 2021-01-22 18:19:15 +00:00)\n" ] ], [ [ "\n\n---\n\n**Improving Phase**", "_____no_output_____" ] ], [ [ "## Improving Phase\n# i = 1\ndef improve():\n print(\"Improving.......\")\n for i in range(N):\n Xnew = population[i]\n print('Persona '+ str(i+1))\n for j in range(D):\n Xnew[j] = c * population[i][j] + r0 * (population[gbest_i][j] - population[i][j])\n try:\n if cal_fitness(Xnew) > fitness[i]:\n population[i] = Xnew\n except:\n print(\"Oops!\", sys.exc_info()[0], \"occurred.\")\n print(\"Next entry.\")\n \n", "time: 4.73 ms (started: 2021-01-22 18:19:15 +00:00)\n" ], [ "", "_____no_output_____" ] ], [ [ "\n\n---\n\n**Acquiring Phase**", "_____no_output_____" ] ], [ [ "## Acquiring Phase\ndef acquire():\n random_person = np.random.randint(low=0, high=N)\n for i in range(N):\n if random_person == i:\n random_person = np.random.randint(low=0, high=N)\n i = i - 1\n continue\n X_new = population[i]\n if fitness[random_person] > fitness[i]:\n for j in range(D):\n X_new[j] = population[i][j] + r1*(population[random_person][j]-population[i][j]) + r2*(population[gbest_i][j]-population[i][j])\n if cal_fitness(X_new) > fitness[i]:\n population[i] = X_new\n else:\n for j in range(D):\n X_new[j] = population[i][j] + r1*(population[i][j]-population[random_person][j]) + r2*(population[gbest_i][j]-population[i][j])\n if cal_fitness(X_new) > fitness[i]:\n population[i] = X_new\n \n", "time: 8.58 ms (started: 2021-01-22 18:19:15 +00:00)\n" ], [ "#Run\ntry:\n for k in range(g):\n print(\"Generation \"+ str(k+1) + \"\\n---------------\")\n fitness_function(population)\n find_gbest()\n mutate()\n improve()\n acquire() \nexcept:\n print()\n print(\"........................\")\n print(\"Optimal Solution Reached\")\n print(\"........................\")", "Generation 1\n---------------\n\nPerson 1\nNo. of Features Used = 3582/10000\nFitness = 0.9863945578231292\nFeature Used = 3582\n\nPerson 2\nNo. of Features Used = 3608/10000\nFitness = 0.9897959183673469\nFeature Used = 3608\n\nPerson 3\nNo. of Features Used = 3614/10000\nFitness = 0.9897959183673469\nFeature Used = 3614\n\nPerson 4\nNo. of Features Used = 3570/10000\nFitness = 0.9897959183673469\nFeature Used = 3570\n\nPerson 5\nNo. of Features Used = 3581/10000\nFitness = 0.9897959183673469\nFeature Used = 3581\nBest fitness value for the generation = 0.9897959183673469 Person 2\n\nMutating the Generation's Worst....Person 1\nNo Mutations in this generation\nImproving.......\nPersona 1\nPersona 2\nPersona 3\nPersona 4\nPersona 5\nGeneration 2\n---------------\n\nPerson 1\nNo. of Features Used = 830/10000\nFitness = 0.9897959183673469\nFeature Used = 830\n\nPerson 2\nNo. of Features Used = 660/10000\nFitness = 0.9863945578231292\nFeature Used = 660\n\nPerson 3\nNo. of Features Used = 647/10000\nFitness = 0.9829931972789115\nFeature Used = 647\n\nPerson 4\nNo. of Features Used = 514/10000\nFitness = 0.9897959183673469\nFeature Used = 514\n\nPerson 5\nNo. of Features Used = 572/10000\nFitness = 0.9829931972789115\nFeature Used = 572\nBest fitness value for the generation = 0.9897959183673469 Person 1\n\nMutating the Generation's Worst....Person 3\nPerson 2 mutated\nImproving.......\nPersona 1\nOops! <class 'ValueError'> occurred.\nNext entry.\nPersona 2\nOops! <class 'ValueError'> occurred.\nNext entry.\nPersona 3\nOops! <class 'ValueError'> occurred.\nNext entry.\nPersona 4\nOops! <class 'ValueError'> occurred.\nNext entry.\nPersona 5\nOops! <class 'ValueError'> occurred.\nNext entry.\n\n........................\nOptimal Solution Reached\n........................\ntime: 17.7 s (started: 2021-01-22 18:19:15 +00:00)\n" ], [ "select.shape", "_____no_output_____" ], [ "", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "<h1>Boolean Mask Arrays</h1>", "_____no_output_____" ] ], [ [ "import numpy as np", "_____no_output_____" ], [ "my_vector = np.array([-17, -4, 0, 2, 21, 37, 105])\nmy_vector", "_____no_output_____" ] ] ]

[ "markdown", "code" ]

[ [ "markdown" ], [ "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# Final Project \n\nFor the final project, you will need to implement a \"new\" statistical algorithm in Python from the research literature and write a \"paper\" describing the algorithm. \n\nSuggested papers can be found in Sakai:Resources:Final_Project_Papers\n\n## Paper\n\nThe paper should have the following:\n\n### Title\n\nShould be concise and informative.\n\n### Abstract\n\n250 words or less. Identify 4-6 key phrases.\n\n### Background\n\nState the research paper you are using. Describe the concept of the algorithm and why it is interesting and/or useful. If appropriate, describe the mathematical basis of the algorithm. Some potential topics for the background include:\n\n- What problem does it address? \n- What are known and possible applications of the algorithm? \n- What are its advantages and disadvantages relative to other algorithms?\n- How will you use it in your research?\n\n### Description of algorithm\n\nFirst, explain in plain English what the algorithm does. Then describes the details of the algorithm, using mathematical equations or pseudocode as appropriate. \n\n### Describe optimization for performance\n\nFirst implement the algorithm using plain Python in a straightforward way from the description of the algorithm. Then profile and optimize it using one or more appropriate methods, such as:\n\n1. Use of better algorithms or data structures\n2. Use of vectorization\n3. JIT or AOT compilation of critical functions\n4. Re-writing critical functions in C++ and using pybind11 to wrap them\n5. Making use of parallelism or concurrency\n6. Making use of distributed compuitng\n\nDocument the improvement in performance with the optimizations performed.\n\n### Applications to simulated data sets\n\nAre there specific inputs that give known outputs (e.g. there might be closed form solutions for special input cases)? How does the algorithm perform on these? \n\nIf no such input cases are available (or in addition to such input cases), how does the algorithm perform on simulated data sets for which you know the \"truth\"? \n\n### Applications to real data sets\n\nTest the algorithm on the real-world examples in the original paper if possible. Try to find at least one other real-world data set not in the original paper and test it on that. Describe and interpret the results.\n\n### Comparative analysis with competing algorithms\n\nFind two other algorithm that address a similar problem. Perform a comparison - for example, of accuracy or speed. You can use native libraries of the other algorithms - you do not need to code them yourself. Comment on your observations. \n\n### Discussion/conclusion\n\nYour thoughts on the algorithm. Does it fulfill a particular need? How could it be generalized to other problem domains? What are its limitations and how could it be improved further?\n\n### References/bibliography\n\nMake sure you cite your sources.\n\n## Code\n\nThe code should be in a public GitHub repository with:\n\n1. A README file\n2. An open source license\n3. Source code\n4. Test code\n5. Examples\n6. A reproducible report\n\n The package should be downloadable and installable with `python setup.py install`, or even posted to PyPI adn installable with `pip install package`. See https://packaging.python.org/tutorials/packaging-projects/ for how to upload to a Python repository. Use the repository at https://test.pypi.org - this is for testing and will be wiped clean after a period.\n\n## Rubric\n\nHere are some considerations I use when grading. Note that the \"difficulty factor\" of the chosen algorithm will be factored into the grading. \n\n1. Is the abstract, background and discussion readable and clear? \n2. Is the algorithm description clear and accurate? \n3. Has the algorithm been optimized? \n4. Are the applications to simulated/real data clear and useful? \n5. Was the comparative analysis done well? \n6. Is there a well-maintained GitHub repository for the code? \n7. Is the document show evidence of literate programming? \n8. Is the analysis reproducible? \n9. Is the code tested? Are examples provided? \n10. Is the package easily installable? \n\n", "_____no_output_____" ] ] ]

[ "markdown" ]

[ [ "markdown" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# ClusterFinder Reference genomes reconstruction\n\nThis notebook validates the 10 genomes we obtained from NCBI based on the ClusterFinder supplementary table. \n\nWe check that the gene locations from the supplementary table match locations in the GenBank files.", "_____no_output_____" ] ], [ [ "from Bio import SeqIO\nfrom Bio.SeqFeature import FeatureLocation\nimport pandas as pd\nfrom Bio import Entrez\nimport seaborn as sns\n\ndef get_features_of_type(sequence, feature_type):\n return [feature for feature in sequence.features if feature.type == feature_type]", "_____no_output_____" ], [ "def get_reference_gene_location(gene_csv_row):\n start = gene_csv_row['gene start'] - 1\n end = gene_csv_row['gene stop']\n strand = 1 if gene_csv_row['gene strand'] == '+' else (-1 if gene_csv_row['gene strand'] == '-' else None)\n return FeatureLocation(start, end, strand)\n\ndef feature_locus_matches(feature, reference_locus):\n return feature.qualifiers.get('locus_tag',[None])[0] == reference_locus", "_____no_output_____" ] ], [ [ "# Loading reference cluster gene locations", "_____no_output_____" ] ], [ [ "reference_genes = pd.read_csv('../data/clusterfinder/labelled/CF_labelled_genes_orig.csv', sep=';')\nreference_genes.head()", "_____no_output_____" ] ], [ [ "# Genes with no sequence", "_____no_output_____" ] ], [ [ "no_sequence_genes = reference_genes[reference_genes['NCBI ID'] == '?']\nno_sequence_counts = no_sequence_genes.groupby('Genome ID')['gene locus'].count()\nprint('{} genes don\\'t have a sequence!'.format(len(no_sequence_genes)))\npd.DataFrame({'missing genes':no_sequence_counts})", "421 genes don't have a sequence!\n" ], [ "reference_ids = reference_genes[reference_genes['NCBI ID'] != '?']['NCBI ID'].unique()\nreference_ids", "_____no_output_____" ] ], [ [ "# Validating that reference genes are found in our sequences", "_____no_output_____" ] ], [ [ "def validate_genome(record, record_reference_genes):\n print('Validating {}'.format(record.id))\n record_genes = get_features_of_type(record, 'gene')\n record_cds = get_features_of_type(record, 'CDS')\n validation = []\n record_length = len(record.seq)\n min_location = record_length\n max_location = -1\n prev_gene_index = None\n prev_cluster_start = None\n for i, reference_gene in record_reference_genes.iterrows():\n reference_gene_location = get_reference_gene_location(reference_gene)\n reference_gene_locus = reference_gene['gene locus']\n reference_cluster_start = reference_gene['NPL start']\n gene_matches_locus = [f for f in record_genes if feature_locus_matches(f, reference_gene_locus)]\n cds_matches_locus = [f for f in record_cds if feature_locus_matches(f, reference_gene_locus)]\n gene_matches_location = [f for f in gene_matches_locus if reference_gene_location == f.location]\n cds_matches_location = [f for f in cds_matches_locus if reference_gene_location == f.location]\n validation.append({\n 'gene_locus_not_found':not gene_matches_locus, \n 'cds_locus_not_found':not cds_matches_locus, \n 'gene_location_correct': bool(gene_matches_location), \n 'cds_location_correct': bool(cds_matches_location)\n })\n if not cds_matches_locus:\n print('No CDS found for gene locus {}'.format(reference_gene_locus))\n if gene_matches_locus:\n gene_match = gene_matches_locus[0]\n if not cds_matches_locus:\n print(' Gene: ', gene_match.qualifiers)\n \n # Use gene index to check if we have a consecutive sequence of genes (except when going from one cluster to another)\n gene_index = [gi for gi,f in enumerate(record_genes) if feature_locus_matches(f, reference_gene_locus)][0]\n if reference_cluster_start == prev_cluster_start and gene_index != prev_gene_index + 1:\n print('Additional unexpected genes found before {} (index {} -> {}) at cluster start {}'.format(reference_gene_locus, prev_gene_index, gene_index, reference_cluster_start))\n \n # Calculate min and max cluster gene location to see how much of the sequence is covered by the reference genes\n min_location = min(gene_match.location.start, min_location)\n max_location = max(gene_match.location.end, max_location)\n prev_gene_index = gene_index\n prev_cluster_start = reference_cluster_start\n \n result = pd.DataFrame(validation).sum().to_dict()\n result['location correct'] = min(result['gene_location_correct'], result['cds_location_correct']) / len(validation)\n result['ID'] = record.id\n result['genome'] = record_reference_genes.iloc[0]['Genome ID']\n result['sequence length'] = record_length\n result['total genes'] = len(record_genes)\n result['reference genes'] = len(record_reference_genes)\n result['first location'] = min_location / record_length\n result['last location'] = max_location / record_length\n result['covered'] = (max_location - min_location) / record_length\n return result", "_____no_output_____" ], [ "validations = []\nreference_gene_groups = reference_genes.groupby('NCBI ID')\nrecords = SeqIO.parse('../data/clusterfinder/labelled/CF_labelled_contigs.gbk', 'genbank')\nfor record in records:\n ncbi_id = record.id\n print(ncbi_id)\n record_reference_genes = reference_gene_groups.get_group(ncbi_id)\n validations.append(validate_genome(record, record_reference_genes))\nvalidations = pd.DataFrame(validations)\nvalidations.set_index('ID', inplace=True)\nvalidations", "CM000950.1\nValidating CM000950.1\nNo CDS found for gene locus SSDG_00035\n Gene: OrderedDict([('locus_tag', ['SSDG_00035']), ('note', ['ABC transporter; frameshift'])])\nNo CDS found for gene locus SSDG_00089\n Gene: OrderedDict([('locus_tag', ['SSDG_00089']), ('note', ['squalene/phytoene dehydrogenase; frameshift'])])\nNo CDS found for gene locus SSDG_04325\n Gene: OrderedDict([('locus_tag', ['SSDG_04325']), ('note', ['AMP-dependent synthetase and ligase; frameshift'])])\nNo CDS found for gene locus SSDG_06129\n Gene: OrderedDict([('locus_tag', ['SSDG_06129']), ('note', ['transferase; frameshift'])])\nNo CDS found for gene locus SSDG_06352\n Gene: OrderedDict([('locus_tag', ['SSDG_06352']), ('note', ['predicted protein; frameshift'])])\nNo CDS found for gene locus SSDG_00185\n Gene: OrderedDict([('locus_tag', ['SSDG_00185']), ('note', ['glycosyl transferase; frameshift'])])\nNo CDS found for gene locus SSDG_06509\n Gene: OrderedDict([('locus_tag', ['SSDG_06509']), ('note', ['beta keto-acyl synthase; frameshift'])])\nNo CDS found for gene locus SSDG_07501\n Gene: OrderedDict([('locus_tag', ['SSDG_07501']), ('note', ['acyltransferase; frameshift'])])\nNo CDS found for gene locus SSDG_07518\n Gene: OrderedDict([('locus_tag', ['SSDG_07518']), ('note', ['predicted protein; frameshift'])])\nNo CDS found for gene locus SSDG_07538\n Gene: OrderedDict([('locus_tag', ['SSDG_07538']), ('note', ['pristinamycin I synthase 3 and 4; frameshift'])])\nDS999641.1\nValidating DS999641.1\nDS999642.1\nValidating DS999642.1\nDS999645.1\nValidating DS999645.1\nGG657738.1\nValidating GG657738.1\nGG657746.1\nValidating GG657746.1\nGG657747.1\nValidating GG657747.1\nGG657750.1\nValidating GG657750.1\nGG657751.1\nValidating GG657751.1\nGG657752.1\nValidating GG657752.1\nGG657754.1\nValidating GG657754.1\nKK037166.1\nValidating KK037166.1\nKK037233.1\nValidating KK037233.1\n" ], [ "validations['location correct'].mean()", "_____no_output_____" ], [ "1 - validations['location correct'].mean()", "_____no_output_____" ], [ "validations[['genome','first location','last location','covered','location correct','reference genes','total genes']]", "_____no_output_____" ] ], [ [ "# Cluster genes", "_____no_output_____" ] ], [ [ "genes = pd.read_csv('../data/clusterfinder/labelled/CF_labelled_genes.csv', sep=';')\ngenes.head()", "_____no_output_____" ], [ "cluster_counts = genes.groupby('contig_id')['cluster_id'].nunique()\ncluster_counts.sort_values().plot.barh()", "_____no_output_____" ], [ "gene_counts = genes.groupby('cluster_id')['locus_tag'].count()\ngene_counts.hist(bins=50)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# Realization of Recursive Filters\n\n*This jupyter notebook is part of a [collection of notebooks](../index.ipynb) on various topics of Digital Signal Processing. Please direct questions and suggestions to [[email protected]](mailto:[email protected]).*", "_____no_output_____" ], [ "## Cascaded Structures\n\nThe realization of recursive filters with a high order may be subject to numerical issues. For instance, when the coefficients span a wide amplitude range, their quantization may require a small quantization step or may impose a large relative error for small coefficients. The basic concept of cascaded structures is to decompose a high order filter into a cascade of lower order filters, typically first and second order recursive filters.", "_____no_output_____" ], [ "### Decomposition into Second-Order Sections\n\nThe rational transfer function $H(z)$ of a linear time-invariant (LTI) recursive system can be [expressed by its zeros and poles](introduction.ipynb#Transfer-Function) as\n\n\\begin{equation}\nH(z) = \\frac{b_M}{a_N} \\cdot \\frac{\\prod_{\\mu=1}^{P} (z - z_{0\\mu})^{m_\\mu}}{\\prod_{\\nu=1}^{Q} (z - z_{\\infty\\nu})^{n_\\nu}}\n\\end{equation}\n\nwhere $z_{0\\mu}$ and $z_{\\infty\\nu}$ denote the $\\mu$-th zero and $\\nu$-th pole of degree $m_\\mu$ and $n_\\nu$ of $H(z)$, respectively. The total number of zeros and poles is denoted by $P$ and $Q$.\n\nThe poles and zeros of a real-valued filter $h[k] \\in \\mathbb{R}$ are either single real valued or conjugate complex pairs. This motivates to split the transfer function into\n\n* first order filters constructed from a single pole and zero\n* second order filters constructed from a pair of conjugated complex poles and zeros\n\nDecomposing the transfer function into these two types by grouping the poles and zeros into single poles/zeros and conjugate complex pairs of poles/zeros results in\n\n\\begin{equation}\nH(z) = K \\cdot \\prod_{\\eta=1}^{S_1} \\frac{(z - z_{0\\eta})}{(z - z_{\\infty\\eta})} \n\\cdot \\prod_{\\eta=1}^{S_2} \\frac{(z - z_{0\\eta}) (z - z_{0\\eta}^*)} {(z - z_{\\infty\\eta})(z - z_{\\infty\\eta}^*)}\n\\end{equation}\n\nwhere $K$ denotes a constant and $S_1 + 2 S_2 = N$ with $N$ denoting the order of the system. The cascade of two systems results in a multiplication of their transfer functions. Above decomposition represents a cascade of first- and second-order recursive systems. The former can be treated as a special case of second-order recursive systems. The decomposition is therefore known as decomposition into second-order sections (SOSs) or [biquad filters](https://en.wikipedia.org/wiki/Digital_biquad_filter). Using a cascade of SOSs the transfer function of the recursive system can be rewritten as\n\n\\begin{equation}\nH(z) = \\prod_{\\mu=1}^{S} \\frac{b_{0, \\mu} + b_{1, \\mu} \\, z^{-1} + b_{2, \\mu} \\, z^{-2}}{1 + a_{1, \\mu} \\, z^{-1} + a_{2, \\mu} \\, z^{-2}}\n\\end{equation}\n\nwhere $S = \\lceil \\frac{N}{2} \\rceil$ denotes the total number of SOSs. These results state that any real valued system of order $N > 2$ can be decomposed into SOSs. This has a number of benefits\n\n* quantization effects can be reduced by sensible grouping of poles/zeros, e.g. such that the spanned amplitude range of the filter coefficients is limited\n* A SOS may be extended by a gain factor to further reduce quantization effects by normalization of the coefficients\n* efficient and numerically stable SOSs serve as generic building blocks for higher-order recursive filters", "_____no_output_____" ], [ "### Example - Cascaded second-order section realization of a lowpass\n\nThe following example illustrates the decomposition of a higher-order recursive Butterworth lowpass filter into a cascade of second-order sections.", "_____no_output_____" ] ], [ [ "%matplotlib inline\nimport numpy as np\nimport matplotlib.pyplot as plt\nfrom matplotlib.markers import MarkerStyle\nfrom matplotlib.patches import Circle\nimport scipy.signal as sig\n\nN = 9 # order of recursive filter\n\ndef zplane(z, p, title='Poles and Zeros'):\n \"Plots zero and pole locations in the complex z-plane\"\n ax = plt.gca()\n \n ax.plot(np.real(z), np.imag(z), 'bo', fillstyle='none', ms = 10)\n ax.plot(np.real(p), np.imag(p), 'rx', fillstyle='none', ms = 10)\n unit_circle = Circle((0,0), radius=1, fill=False,\n color='black', ls='solid', alpha=0.9)\n ax.add_patch(unit_circle)\n ax.axvline(0, color='0.7')\n ax.axhline(0, color='0.7')\n \n plt.title(title)\n plt.xlabel(r'Re{$z$}')\n plt.ylabel(r'Im{$z$}')\n plt.axis('equal')\n plt.xlim((-2, 2))\n plt.ylim((-2, 2))\n plt.grid()\n\n\n# design filter\nb, a = sig.butter(N, 0.2)\n# decomposition into SOS\nsos = sig.tf2sos(b, a, pairing='nearest')\n\n\n# print filter coefficients\nprint('Coefficients of the recursive part \\n')\nprint(['%1.2f'%ai for ai in a])\nprint('\\n')\nprint('Coefficients of the recursive part of the individual SOS \\n')\nprint('Section \\t a1 \\t\\t a2')\nfor n in range(sos.shape[0]):\n print('%d \\t\\t %1.5f \\t %1.5f'%(n, sos[n, 4], sos[n, 5]))\n\n# plot pole and zero locations\nplt.figure(figsize=(5,5))\nzplane(np.roots(b), np.roots(a), 'Poles and Zeros - Overall')\n\nplt.figure(figsize=(10, 7))\nfor n in range(sos.shape[0]): \n plt.subplot(231+n)\n zplane(np.roots(sos[n, 0:3]), np.roots(sos[n, 3:6]), title='Poles and Zeros - Section %d'%n)\nplt.tight_layout()\n\n# compute and plot frequency response of sections\nplt.figure(figsize=(10,5))\nfor n in range(sos.shape[0]):\n Om, H = sig.freqz(sos[n, 0:3], sos[n, 3:6])\n plt.plot(Om, 20*np.log10(np.abs(H)), label=r'Section %d'%n)\n\nplt.xlabel(r'$\\Omega$')\nplt.ylabel(r'$|H_n(e^{j \\Omega})|$ in dB')\nplt.legend()\nplt.grid()", "Coefficients of the recursive part \n\n['1.00', '-5.39', '13.38', '-19.96', '19.62', '-13.14', '5.97', '-1.78', '0.31', '-0.02']\n\n\nCoefficients of the recursive part of the individual SOS \n\nSection \t a1 \t\t a2\n0 \t\t -0.50953 \t 0.00000\n1 \t\t -1.04232 \t 0.28838\n2 \t\t -1.11568 \t 0.37905\n3 \t\t -1.25052 \t 0.54572\n4 \t\t -1.46818 \t 0.81477\n" ] ], [ [ "**Exercise**\n\n* What amplitude range is spanned by the filter coefficients?\n* What amplitude range is spanned by the SOS coefficients?\n* Change the pole/zero grouping strategy from `pairing='nearest'` to `pairing='keep_odd'`. What changes?\n* Increase the order `N` of the filter. What changes?\n\nSolution: Inspecting both the coefficients of the recursive part of the original filter and of the individual SOS reveals that the spanned amplitude range is lower for the latter. The choice of the pole/zero grouping strategy influences the locations of the poles/zeros in the individual SOS, the spanned amplitude range of their coefficients and the transfer functions of the individual sections. The total number of SOS scales with the order of the original filter.", "_____no_output_____" ], [ "**Copyright**\n\nThis notebook is provided as [Open Educational Resource](https://en.wikipedia.org/wiki/Open_educational_resources). Feel free to use the notebook for your own purposes. The text is licensed under [Creative Commons Attribution 4.0](https://creativecommons.org/licenses/by/4.0/), the code of the IPython examples under the [MIT license](https://opensource.org/licenses/MIT). Please attribute the work as follows: *Sascha Spors, Digital Signal Processing - Lecture notes featuring computational examples, 2016-2018*.", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown" ]

[ [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "# Utilizing daal4py in Data Science Workflows\n\nThe notebook below has been made to demonstrate daal4py in a data science context. It utilizes a Cycling Dataset for pyworkout-toolkit, and attempts to create a linear regression model from the 5 features collected for telemetry to predict the user's Power output in the absence of a power meter.", "_____no_output_____" ] ], [ [ "import pandas as pd\nimport matplotlib.pyplot as plt\nimport glob\nimport sys\n%matplotlib inline\nsys.version", "_____no_output_____" ] ], [ [ "This example will be exploring workout data pulled from Strava, processed into a CSV for Pandas and daal4py usage. Below, we utilize pandas to read in the CSV file, and look at the head of dataframe with .head()", "_____no_output_____" ] ], [ [ "workout_data_dd= pd.read_csv('data/batch/cycling_dataset.csv', index_col=0)\nworkout_data_dd.head()", "_____no_output_____" ] ], [ [ "The data above has several key features that would be of great use here. \n- Altitude can affect performance, so it might be a useful feature. \n- Cadence is the revolutions per minute of the crank, and may have possible influence. \n- Heart Rate is a measure of the body's workout strain, and would have a high possibly of influence.\n- Distance may have a loose correlation as it is highly route dependent, but might be possible.\n- Speed has possible correlations as it ties directly into power.", "_____no_output_____" ], [ "## Explore and visualize some of the data", "_____no_output_____" ], [ "In general, we are trying to predict on the 'power' in Watts to see if we can generate a model that can predict one's power output without the usage of a cycling power meter. Below are some basic scatterplots as we explore the data. Scatterplots are great for looking for patterns and correlation in the data itself. Below, we can see that cadence and speed are positively correlated. ", "_____no_output_____" ] ], [ [ "workout_data_dd.plot.scatter('cadence','power')\nplt.show()\nworkout_data_dd.plot.scatter('hr','power')\nplt.show()\nworkout_data_dd.plot.scatter('cadence','speed')\nplt.show()\nworkout_data_dd.plot.scatter('speed','power')\nplt.show()\nworkout_data_dd.plot.scatter('altitude','power')\nplt.show()\nworkout_data_dd.plot.scatter('distance','power')\nplt.show()", "_____no_output_____" ] ], [ [ "## Using daal4py for Machine Learning tasks\n\nIn the sections below, we will be using daal4py directly. After importing the model, we will arrange it in a separate independent and dependent dataframes, then use the daal4py's training and prediction classes to generate a workable model.", "_____no_output_____" ] ], [ [ "import daal4py as d4p", "_____no_output_____" ] ], [ [ "It is now the time to split the dataset into train and test sets. This is demonstrated below.", "_____no_output_____" ] ], [ [ "print(workout_data_dd.shape)\ntrain_set = workout_data_dd[0:3000]\ntest_set = workout_data_dd[3000:]\nprint(train_set.shape, test_set.shape)", "(3902, 9)\n(3000, 9) (902, 9)\n" ], [ "# Reduce the dataset, create X. We drop the target, and other non-essential features.\nreduced_dataset = train_set.drop(['time','power','latitude','longitude'], axis=1)\n# Get the target, create Y\ntarget = train_set.power.values.reshape((-1,1))\n# This is essentially doing np.array(dataset.power.values, ndmin=2).T\n# as it needs to force a 2 dimensional array as we only have 1 target", "_____no_output_____" ] ], [ [ "X is 5 features by 3k rows, Y is 3k rows by 1 column", "_____no_output_____" ] ], [ [ "print(reduced_dataset.values.shape, target.shape)", "(3000, 5) (3000, 1)\n" ] ], [ [ "## Training the model", "_____no_output_____" ], [ "Create the Linear Regression Model, and train the model with the data. We utilize daal4py's linear_regression_training class to create the model, then call .compute() with the independent and dependent data as the parameters.", "_____no_output_____" ] ], [ [ "d4p_lm = d4p.linear_regression_training(interceptFlag=True)\nlm_trained = d4p_lm.compute(reduced_dataset.values, target)", "_____no_output_____" ], [ "print(\"Model has this number of features: \", lm_trained.model.NumberOfFeatures)", "Model has this number of features: 5\n" ] ], [ [ "## Prediction (inference) with the trained model", "_____no_output_____" ], [ "Now that the model is trained, we can test it with the test part of the dataset. We drop the same features to match that of the trained model, and put it into daal4py's linear_regression_prediction class.", "_____no_output_____" ] ], [ [ "subset = test_set.drop(['time','power','latitude','longitude'], axis=1)", "_____no_output_____" ] ], [ [ "Now we can create the Prediction object and use the reduced dataset for prediction. The class's arguments use the independent data and the trained model from above as the parameters.", "_____no_output_____" ] ], [ [ "lm_predictor_component = d4p.linear_regression_prediction()\nresult = lm_predictor_component.compute(subset.values, lm_trained.model)", "_____no_output_____" ], [ "plt.plot(result.prediction[0:300])\nplt.plot(test_set.power.values[0:300])\nplt.show()", "_____no_output_____" ] ], [ [ "The graph above shows the Orange (predicted) result over the Blue (original data). This data is notoriously sparse in features leading to a difficult to predict target!", "_____no_output_____" ], [ "## Model properties\nAnother aspect of the model is the trained model's properties, which are explored below.", "_____no_output_____" ] ], [ [ "print(\"Betas:\",lm_trained.model.Beta) \nprint(\"Number of betas:\", lm_trained.model.NumberOfBetas)\nprint(\"Number of Features:\", lm_trained.model.NumberOfFeatures)", "Betas: [[ 1.51003501e+01 -1.25075548e-01 1.32249115e+00 1.64363922e-03\n 8.53155955e-01 -1.09595022e+01]]\nNumber of betas: 6\nNumber of Features: 5\n" ] ], [ [ "## Additional metrics\nWe can generate metrics on the independent data with daal4py's low_order_moments() class.", "_____no_output_____" ] ], [ [ "metrics_processor = d4p.low_order_moments()\ndata = metrics_processor.compute(reduced_dataset.values)\ndata.standardDeviation", "_____no_output_____" ] ], [ [ "## Migrating the trained model for inference on external systems\n\nOccasionally one may need to migrate the trained model to another system for inference only--this use case allows the training on a much more powerful machine with a larger dataset, and placing the trained model for inference-only on a smaller machine.", "_____no_output_____" ] ], [ [ "import pickle", "_____no_output_____" ], [ "with open('trained_model2.pickle', 'wb') as model_pi:\n pickle.dump(lm_trained.model, model_pi)\n model_pi.close", "_____no_output_____" ] ], [ [ "The trained model file above can be moved to an inference-only or embedded system. This is useful if the training is extreamly heavy or computed-limited. ", "_____no_output_____" ] ], [ [ "with open('trained_model2.pickle', 'rb') as model_import:\n lm_import = pickle.load(model_import)", "_____no_output_____" ] ], [ [ "The imported model from file is now usable again. We can check the betas from the model to ensure that the trained model is present.", "_____no_output_____" ] ], [ [ "lm_import.Beta", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# **JIVE: Joint and Individual Variation Explained**", "_____no_output_____" ], [ "JIVE (Joint and Individual Variation Explained) is a dimensional reduction algorithm that can be used when there are multiple data matrices (data blocks). The multiple data block setting means there are $K$ different data matrices, with the same number of observations $n$ and (possibly) different numbers of variables ($d_1, \\dots, d_k$). JIVE finds modes of variation which are common (joint) to all $K$ data blocks and modes of individual variation which are specific to each block. For a detailed discussion of JIVE see [Angle-Based Joint and Individual Variation Explained](https://arxiv.org/pdf/1704.02060.pdf).[^1]\n\nFor a concrete example, consider a two block example from a medical study. Suppose there are $n=500$ patients (observations). For each patient we have $d_1 = 100$ bio-medical variables (e.g. hit, weight, etc). Additionally we have $d_2 = 10,000$ gene expression measurements for each patient.\n\n## **The JIVE decomposition**\n\nSuppose we have $K$ data data matrices (blocks) with the same number of observations, but possibly different numbers of variables; in particular let $X^{(1)}, \\dots, X^{(K)}$ where $X^{(k)} \\in \\mathbb{R}^{n \\times d_k}$. JIVE will then decompose each matrix into three components: joint signal, individual signal and noise\n\n\\begin{equation}\nX^{(k)} = J^{(k)} + I^{(k)} + E^{(k)}\n\\end{equation}\n\nwhere $J^{(k)}$ is the joint signal estimate, $I^{(k)}$ is the individual signal estimate and $E^{(k)}$ is the noise estimate (each of these matrices must the same shape as the original data block: $\\mathbb{R}^{n \\times d_k}$). Note: **we assume each data matrix** $X^{(k)}$ **has been column mean centered**.\n\n\nThe matrices satisfy the following constraints:\n\n1. The joint matrices have a common rank: $rk(J^{(k)}) = r_{joint}$ for $k=1, \\dots, K$.\n2. The individual matrices have block specific ranks $rk(I^{(k)}) = r_{individual}^{(k)}$.\n3. The columns of the joint matrices share a common space called the joint score space (a subspace of $\\mathbb{R}^n$); in particular the $\\text{col-span}(J^{(1)}) = \\dots = \\text{col-span}(J^{(K)})$ (hence the name joint).\n4. Each individual spaces score subspace (of $\\mathbb{R}^n$) is orthogonal to the the joint space; in particular $\\text{col-span}(J^{(k)}) \\perp \\text{col-span}(I^{(k)})$ for $k=1, \\dots, K$.\n\nNote that JIVE may be more natural if we think about data matrices subspaces of $\\mathbb{R}^n$ (the score space perspective). Typically we think of a data matrix as $n$ points in $\\mathbb{R}^d$. The score space perspective views a data matrix as $d$ vectors in $\\mathbb{R}^n$ (or rather the span of these vectors). One important consequence of this perspective is that it makes sense to related the data blocks in score space (e.g. as subspaces of $\\mathbb{R}^n$) since they share observtions.\n\n## Quantities of interest\n\nThere are a number of potential quantities of interest depending on the application. For example the user may be interested in the full matrices $J^{(k)}$ and/or $I^{(k)}$. By construction these matrices are not full rank and we may also be interested in their singular value decomposition which we define as\n\n\\begin{align}\n& U^{(k)}_{joint}, D^{(k)}_{joint}, V^{(k)}_{joint} = \\text{rank } r_{joint} \\text{ SVD of } J^{(k)} \\\\\n& U^{(k)}_{individual}, D^{(k)}_{individual}, V^{(k)}_{individual} = \\text{rank } r_{individual}^{{k}} \\text{ SVD of } I^{(k)}\n\\end{align}\n\n\nOne additional quantity of interest is $U_{joint} \\in \\mathbb{R}^{n \\times r_{joint}}$ which is an orthogonal basis of $\\text{col-span}(J^{(k)})$. This matrix is produced from an intermediate JIVE computation. \n\n## **PCA analogy**\nWe give a brief discussion of the PCA/SVD decomposition (assuming the reading is already familiar).\n\n#### Basic decomposition\nSuppose we have a data matrix $X \\in \\mathbb{n \\times d}$. Assume that $X$ has been column mean centered and consider the SVD decomposition (this is PCA since we have mean centered the data):\n\n\\begin{equation}\nX = U D V^T.\n\\end{equation}\nwhere $U \\in \\mathbb{R}^{n \\times m}$, $D \\in \\mathbb{R}^{m \\times m}$ is diagonal, and $V \\in \\mathbb{R}^{d \\times m}$ with $m = min(n, d)$. Note $U^TU = V^TV = I_{m \\times m}$. \n\nSuppose we have decided to use a rank $r$ approximation. We can then decompose $X$ into a signal matrix ($A$) and an noise matrix ($E$)\n\n\\begin{equation}\nX = A + E,\n\\end{equation}\nwhere $A$ is the rank $r$ SVD approximation of $X$ i.e. \n\\begin{align}\nA := & U_{:, 1:r} D_{1:r, 1:r} V_{:, 1:r}^T \\\\\n = & \\widetilde{U}, \\widetilde{D} \\widetilde{V}^T\n\\end{align}\nThe notation $U_{:, 1:r} \\in \\mathbb{R}^{n \\times r}$ means the first $r$ columns of $U$. Similarly we can see the error matrix is $E :=U_{:, r+1:n} D_{r+1:m, r_1:m} V_{:, r+1:d}^T$.\n\n#### Quantities of interest\n\nThere are many ways to use a PCA/SVD decomposition. Some common quantities of interest include\n\n- The normalized scores: $\\widetilde{U} \\in \\mathbb{R}^{n \\times r}$\n- The unnormalized scores: $\\widetilde{U}\\widetilde{D} \\in \\mathbb{R}^{n \\times r}$\n- The loadings: $\\widetilde{V} \\in \\mathbb{R}^{d \\times r}$\n- The full signal approximation: $A \\in \\mathbb{R}^{n \\times d}$\n\n\n#### Scores and loadings\n\nFor both PCA and JIVE we use the notation $U$ (scores) and $V$ (loadings). These show up in several places.\n\nWe refer to all $U \\in \\mathbb{R}^{n \\times r}$ matrices as scores. We can view the $n$ rows of $U$ as representing the $n$ data points with $r$ derived variables (put differently, columns of $U$ are $r$ derived variables). The columns of $U$ are orthonormal: $U^TU = I_{r \\times r}$.\n\nSometimes we may want $UD$ i.e scale the columns of $U$ by $D$ (the columns are still orthogonal). The can be useful when we want to represent the original data by $r$ variables. We refer to $UD$ as unnormalized scores.\n\nWe refer to all $V\\in \\mathbb{R}^{d \\times r}$ matrices as loadings[^2]. The j$th$ column of $V$ gives the linear combination of the original $d$ variables which is equal to the j$th$ unnormalized scores (j$th$ column of $UD$). Equivalently, if we project the $n$ data points (rows of $X$) onto the j$th$ column of $V$ we get the j$th$ unnormalized scores. \n\nThe typical geometric perspective of PCA is that the scores represent $r$ new derived variables. For example, if $r = 2$ we can look at a scatter plot that gives a two dimensional approximation of the data. In other words, the rows of the scores matrix are $n$ data points living in $\\mathbb{R}^r$. \n\nAn alternative geometric perspective is the $r$ columns of the scores matrix are vectors living in $\\mathbb{R}^n$. The original $d$ variables span a subspace of $\\mathbb{R}^n$ given by $\\text{col-span}(X)$. The scores then span a lower dimensional subspace of $\\mathbb{R}^n$ that approximates $\\text{col-span}(X)$.\n\nThe first perspective says PCA finds a lower dimensional approximation to a subspace in $\\mathbb{R}^d$ (spanned by the $n$ data points). The second perspective says PCA finds a lower dimensional approximation to a subspace in $\\mathbb{R}^n$ (spanned by the $d$ data points).\n\n## **JIVE operating in score space**\n\nFor a data matrix $X$ let's call the span of the variables (columns) the *score subpace*, $\\text{col-span}(X) \\subset \\mathbb{R}^n$. Typically we think of a data matrix as $n$ points in $\\mathbb{R}^d$. The score space perspective reverses this and says a data matrix is $d$ points in $\\mathbb{R}^n$. When thinking in the score space it's common to consider about subspaces i.e. the span of the $d$ variables in $\\mathbb{R}^n$. In other words, if two data matrices have the same column span then their score subspaces are the same[^3].\n\nJIVE partitions the score space of each data matrix into three subspaces: joint, individual and noise. The joint score subspace for each data block is the same. The individual score subspace, however, is (possibly) different for each of the $K$ blocks. The k$th$ block's individual score subspace is orthogonal to the joint score subspace. Recall that the $K$ data matrices have the same number of observations ($n$) so it makes sense to think about how the data matrices relate to each other in score space.\n\nPCA partitions the score space into two subspaces: signal and noise (see above). For JIVE we might combine the joint and individual score subspaces and call this the signal score subspace.\n\n\n", "_____no_output_____" ], [ "# Footnotes\n[^1]: Note this paper calls the algorithm AJIVE (angle based JIVE) however, we simply use JIVE. Additionally, the paper uses columns as observations in data matrices where as we use rows as observations.\n\n[^2]: For PCA we used tildes (e.g. $\\widetilde{U}$) to denote the \"partial\" SVD approximation however for the final JIVE decomposition we do not use tildes. This is intentional since for JIVE the SVD comes from the $I$ and $J$ matrices which are exactly rank $r$. Therefore we view this SVD as the \"full\" SVD.\n\n[^3]: This might remind the reader of TODO", "_____no_output_____" ] ] ]

[ "markdown" ]

[ [ "markdown", "markdown", "markdown" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "from gs_quant.common import PayReceive, Currency\nfrom gs_quant.instrument import IRSwap\nfrom gs_quant.session import Environment, GsSession\nfrom gs_quant.risk import DollarPrice, IRDelta, IRDeltaParallel", "_____no_output_____" ], [ "# external users should substitute their client id and secret; please skip this step if you are using internal jupyterhub\nGsSession.use(Environment.PROD, client_id=None, client_secret=None, scopes=('run_analytics',))", "_____no_output_____" ], [ "swap = IRSwap(PayReceive.Pay, '10y', Currency.GBP)", "_____no_output_____" ], [ "result = swap.calc((DollarPrice, IRDeltaParallel)) ", "_____no_output_____" ], [ "print(result) # all results", "_____no_output_____" ], [ "print(result[IRDeltaParallel]) # single measure", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code" ] ]

[ "BSD-3-Clause" ]

[ "BSD-3-Clause" ]

[ "BSD-3-Clause" ]

[ [ [ "import numpy as np\nimport matplotlib.pyplot as plt\nimport pandas as pd\n\nimport check_lab05 as p\n\nplt.rcParams.update({'font.size': 14})\nplt.rcParams['lines.linewidth'] = 3\npi=np.pi", "_____no_output_____" ] ], [ [ "ME 3264 - Applied Measurements Laboratory\n===========================================\n\nLab #5 - Linear Variable Differential Transformer(LVDT)\n=====================================================\n\n## Objective\nThe objectives of this laboratory are :\n\n1. Gain familiarity with the physical operating principle of Linear Variable Differential Transformer (LVDT) measurements\n2. Calibrate the LVDT measurement using voltage measurements\n3. Determine relationship between supplied voltage amplitude and linear motion\n4. Investigate effects of changing input frequency and sampling rate on response frequency \n", "_____no_output_____" ], [ "## Background\n\n\n### Working principle of LVDT\n\nA linear variable differential transformer (LVDT) is a device that can measure the absolute linear position or changes in position of a separate device. LVDTs operate on the principle of a transformer. Because it is a transformer, the LVDT requires an ac drive signal. As shown in Fig.1 (and Fig.2A) , an LVDT consists of a coil assembly and a core. The coil assembly is typically mounted to a stationary form, while the core is secured to the object whose position is being measured. The coil assembly consists of three coils of wire wound on the hollow form. A core of permeable material can slide freely through the center of the form. The inner coil is the primary, which is excited by an AC source as shown. Magnetic flux produced by the primary is coupled to the two secondary coils, inducing an AC voltage in each coil [2].\n\n<center><img src=\"https://upload.wikimedia.org/wikipedia/commons/5/57/LVDT.png\" alt=\"Drawing\" style=\"width: 300px;\"/> </center>\n\n<center>Figure 1: Cutaway view of an LVDT. Current is driven through the primary coil at A, causing an induction current to be generated through the secondary coils at B. [4] </center>\n\n### LVDT Measurement\n\nAn LVDT measures displacement by associating a specific signal value for any given position of the core. This association of a signal value to a position occurs through electromagnetic coupling of an AC excitation signal on the primary winding to the core and back to the secondary windings as shown in Fig.2B. The position of the core determines how tightly the signal of the primary coil is coupled to each of the secondary coils. The two secondary coils are series-opposed, which means wound in series but in opposite directions. This results in the two signals on each secondary being 180 deg out of phase. Therefore phase of the output signal determines direction and its amplitude, distance [2].\n\nFig.2C shows the operational characteristics of LVDT with respect to the core displacement. \n\n<center><img src=\"https://ars.els-cdn.com/content/image/3-s2.0-B9780081028841000042-f04-12-9780081028841.jpg\" alt=\"Drawing\" style=\"width: 200px;\"/> </center>\n\n<center>Figure 2: The operation of the LVDT (A) Internal arrangement (B) Electrical circuit, the dots signify the positive ending of the winding (C) Operational characteristics [5] </center>\n\n### Advantages of LVDT\n\nLVDTs have a number of advantages, including -\n\n* The ability to measure absolute position, the ability to be completely sealed from the environment, nearly frictionless operation, and excellent repeatability of the measurement\n* Because the device relies on the coupling of magnetic flux, an LVDT can have infinite resolution. Therefore the smallest fraction of movement can be detected by suitable signal conditioning hardware, and the resolution of the transducer is solely determined by the resolution of the data acquisition system\n* Linearity of operation as output is a direct and linear function of the input\n\n\n\n", "_____no_output_____" ], [ "## Part 1 - LVDT calibaration", "_____no_output_____" ], [ "### Problem 1 - Relate voltage to displacement\n\nLet's consider a LVDT with it's core attached to a micrometer. Following table consists of core displacment, and mean DC voltage, recorded by DAQ in an experiment. Obtain the caliberation curve of LVDT using linear regression. \n\n\n|Displacement| Mean DC |\n|--- | --- | \n|5 mm | 4.829 V | \n|6 mm | 6.690 V | \n|7 mm | 8.333 V | \n|4 mm | 3.868 V | \n|3 mm | 2.024 V | \n|2 mm | 0.145 V | \n|1 mm | -1.738 V | \n\n\n", "_____no_output_____" ] ], [ [ "from scipy.optimize import curve_fit\n\n\ndef func(x,a,b):\n '''fits the linear equation y = a + bx\n This equation can be replaced by polynomial or exponential \n as per the fitting goals of the problem'''\n return (a + b*x)\n\nx = [5,6,7,4,3,2,1] # mm\ny = [4.829,6.690,8.333,3.686,2.024,0.145,-1.738] # Volt\n\nk,pcov=curve_fit(func, x, y) \nk_error=np.sqrt(np.diag(pcov)) # Co-variance matrix\na = np.asarray(k[0])\nb = np.asarray(k[1])\n\nprint(\"Caliberation equation for LVDT is y = %1.3f +%1.3fx \\n \"%(a,b))\nprint(\"Caliberation coefficent a =%1.3f +/- %1.3f \\n\"%(k[0],k_error[0]))\nprint(\"Caliberation coefficent b = %1.3f +/- %1.3f \\n\"%(k[1],k_error[1]))\n\n\nplt.plot(x,y,'o',label='experiment')\nplt.plot(x,func(x,a,b),label='model')\nplt.legend()\nplt.xlabel(r'Displacement, mm')\nplt.ylabel('Mean DC, Volts')", "Caliberation equation for LVDT is y = -3.163 +1.647x \n \nCaliberation coefficent a =-3.163 +/- 0.185 \n\nCaliberation coefficent b = 1.647 +/- 0.041 \n\n" ] ], [ [ "Note - In the curvefit,\n\n`k,pcov=curve_fit(func, x, y)`\n\n* `k` - Optimal values for the parameters so that the sum of the squared residuals of (f(x,k) - y) is minimized.\n* `pcov` - The estimated covariance of k. The diagonals provide the variance of the parameter estimate. To compute one standard deviation errors on the parameters, we use k_error = np.sqrt(np.diag(pcov)).", "_____no_output_____" ], [ "### Problem 2 - Check your work\n\nCalculate the linear regression coefficents a and b, and their standard variances for data in problem 1 using linear least square fitting method described in [Ref 3](https://mathworld.wolfram.com/LeastSquaresFitting.html). Does these values compare well with the coefficents and standard variance values obtained from covarianc matrix in the above example?", "_____no_output_____" ] ], [ [ "## # enter your work here - Uncomment the following lines of code and make necessary changes\n\n# n = len(x)\n\n# # sum of squares\n\n# x_mean = np.mean(x)\n# y_mean = \n# ss_xx = np.sum((x-x_mean)**2)\n# ss_yy = \n# ss_xy = np.sum((y-y_mean)*(x-x_mean))\n\n# # linear regression coefficients\n# b = ss_xy/ss_xx\n# a = y_mean - b*x_mean\n# print(\"Equation for linear egression line is y = %1.3f +%1.3fx \\n \"%(a,b))\n\n# # correlation coefficient,\n# r2 = ss_xy**2/ss_xx/ss_yy\n# print(\"Correlation coefficient, is y = %1.3f \\n \"%(r2))\n\n# # The standard errors for a and b \n# s = (ss_yy -ss_xy**2/ss_xx)/(n-2)\n\n# sigma_a = np.sqrt(s)*np.sqrt(1/n + (x_mean**2/ss_xx))\n# sigma_b = np.sqrt(s)/np.sqrt(ss_xx)\n\n# print(\"Std (a)= %1.3f\\n \"%sigma_a)\n# print(\"Std (b)= %1.3f\\n \"%sigma_b)\n\np.check_p02(a, b)", "Nice work!\n" ] ], [ [ "## Part 2 - Calibrate Piezoelectric with LVDT output\n\nIn part 2, you compare the amplitude of input voltage for the piezoelectric motion to the amplitude of LVDT voltage output. When the output voltage is larger, the piezoelectric is moving further. \n\n", "_____no_output_____" ] ], [ [ "from IPython.display import YouTubeVideo\nYouTubeVideo('NwE8B9IHvyo')", "_____no_output_____" ] ], [ [ "### Problem 3 - Calibrate Piezoelectric\n\nFollowing table consists of amplitude of input voltage for the piezoelectric motion, and LVDT voltage output, recorded by DAQ in an experiment. Obtain the caliberation curve using linear regression. You can use `curve_fit` as expalined in Part 1. \n\n|Input, Vpp | Output Vpp |\n|--- | --- | \n|1 | 0.00101 | \n|2 | 0.00380 | \n|4 | 0.00698 | \n|4 | 0.01048 | \n|5 | 0.01420 | \n|6 | 0.01832 | \n|7 | 0.02273 | \n", "_____no_output_____" ] ], [ [ "## enter your work here", "_____no_output_____" ] ], [ [ "## Part 3 - Explore frequency-response between Piezoelectric and LVDT\n\nIn part 3, you vary the frequency input to the piezoelectric and measure the frequency output measured by the LVDT. You are constrained by the Nyquist frequency in these measurements. If you collect data at 500 Hz, then the largest frequency you can reliably measure is 250 Hz. The concept of Nyquist frequency if further explored with Problem 4. \n\n\n### Problem 4 - Nyquist frequency\n\nConsider a case where signal generator sends a signal of the form of cos-wave to a piezo motor. The signal has frequency of 1-Hz with amplitude of 2 Vpp (Volts peak-to-peak) . The Data Aquisition System (DAQ) takes N measurements over the given timeframe from 0-10 seconds to measure the corresponding LVDT output signal. Plot and compare the input and measured signals. ", "_____no_output_____" ] ], [ [ "N=20\nt_collect=10 # time to collect data\nt=np.linspace(0,t_collect,1000)\ny=np.cos(2*pi*t)\ntsample=np.linspace(0,10,N+1)\nysample=np.cos(2*pi*tsample)\nplt.figure(20)\nplt.plot(t,y,label='signal')\nplt.plot(tsample,ysample,'o-',label='measure')\nplt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\nplt.xlabel('time (s)')\nplt.ylabel('a.u.')", "_____no_output_____" ] ], [ [ "For N=20, it would appear that DAQ can capture a minimal example of input signal (just the peaks occuring at 1 Hz). Collecting data for N=20 over 10 seconds is equivalent to sampling at 20 samples/10 seconds = 2 Hz. This is called the Nyquist rate which is given as such\n\n$f_{Nyquist}=2f_{signal}$. (1)\n\nIn Equation 1, the Nyquist rate (also Shannon Sampling) [\\[6\\]](https://github.uconn.edu/rcc02007/ME3264-Lab_03)[\\[7\\]](./jerri_1977-shannon_sampling.pdf)[\\[8\\]](./nyquist.pdf), $f_{Nyquist}$, is the minimum sampling rate necessary to capture the signal at frequency, $f_{signal}$. Try changing N<20 and consider the apparent signal frequencies. \n\nIf you try N=11 in the Python code below, you will see a phenomenon called \"aliasing\" or the \"wagon-wheel effect\" [\\[9\\]](http://www.onmyphd.com/?p=aliasing). When you look at the measured signal, it appears to have a frequency of 1 cycle/10 seconds = 0.1 Hz. This phenomenon is called the wagon-wheel effect because it is noticeable when recording spinning objects like a wagon wheel [or turbine](https://www.youtube.com/watch?v=vIsS4TP73AU). The wheel spins at a given frequency and the camera records at another frequency. When the ratio of the wheel frequency to camera recording frequency reaches certain values the wheel appears to stop, spin slower, or even backwards. [\\[6\\]](https://github.uconn.edu/rcc02007/ME3264-Lab_03)\n\nExperimentally, we avoid aliasing by sampling above the Nyquist rate from equation 1.", "_____no_output_____" ] ], [ [ "N=11\nt_collect=10 # time to collect data\nt=np.linspace(0,t_collect,1000)\ny=np.cos(2*pi*t)\ntsample=np.linspace(0,10,N+1)\nysample=np.cos(2*pi*tsample)\nplt.figure(20)\nplt.plot(t,y,label='signal')\nplt.plot(tsample,ysample,'o-',label='measure')\nplt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\nplt.xlabel('time (s)')\nplt.ylabel('a.u.')", "_____no_output_____" ] ], [ [ "## Procedure \n\nThe procedure and details of the experiment are included in a lab-handout [1].\n\n[ME3264_Lab_5_LVDT.pdf](https://drive.google.com/file/d/1FbykzotAE50SRTujNUjvF9vek97TlJXL/view?usp=sharing)\n", "_____no_output_____" ] ], [ [ "YouTubeVideo('FRWgpFApITo')", "_____no_output_____" ] ], [ [ "## Notes on error propagation and uncertainties\n\n\n", "_____no_output_____" ], [ "## References \n\n1. [ME3264_Lab_5_LVDT.pdf](https://drive.google.com/file/d/1FbykzotAE50SRTujNUjvF9vek97TlJXL/view?usp=sharing)\n2. [Measuring Position and Displacement with LVDTs](https://www.ni.com/en-us/innovations/white-papers/06/measuring-position-and-displacement-with-lvdts.html)\n3. [Least Squares Fitting](https://mathworld.wolfram.com/LeastSquaresFitting.html)\n4. [Linear variable differential transformer, From Wikipedia](https://en.wikipedia.org/wiki/Linear_variable_differential_transformer)\n5. [Velocity and position transducers, Richard Crowder, in Electric Drives and Electromechanical Systems (Second Edition), 2020](https://www.sciencedirect.com/science/article/pii/B9780081028841000042)\n6. [ME3263-Lab_03, Prof. Ryan Cooper](https://github.uconn.edu/rcc02007/ME3264-Lab_03)\n", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import tensorflow as tf\nfrom tensorflow.keras import layers, Model\nfrom tensorflow.keras.activations import relu\nfrom tensorflow.keras.models import Sequential, load_model\nfrom tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping\nfrom tensorflow.keras.losses import BinaryCrossentropy\nfrom tensorflow.keras.optimizers import RMSprop, Adam\nfrom tensorflow.keras.metrics import binary_accuracy\nimport tensorflow_datasets as tfds\nfrom tensorflow_addons.layers import InstanceNormalization\nimport numpy as np\nimport matplotlib.pyplot as plt\n\nimport warnings\nwarnings.filterwarnings('ignore')\nprint(\"Tensorflow\", tf.__version__)\nfrom packaging.version import parse as parse_version\nassert parse_version(tf.__version__) < parse_version(\"2.4.0\"), \\\n f\"Please install TensorFlow version 2.3.1 or older. Your current version is {tf.__version__}.\"", "Tensorflow 2.2.0\n" ], [ "gpus = tf.config.experimental.list_physical_devices('GPU')\ntf.config.experimental.set_visible_devices(gpus[0], 'GPU')\nlogical_gpus = tf.config.experimental.list_logical_devices('GPU')\nprint(len(gpus), \"Physical GPUs,\", len(logical_gpus), \"Logical GPU\")", "4 Physical GPUs, 1 Logical GPU\n" ], [ "ds_train, ds_info = tfds.load('mnist', split='train', shuffle_files=True, with_info=True)\nfig = tfds.show_examples(ds_info, ds_train)", "_____no_output_____" ], [ "batch_size = 400\nglobal_batch_size = batch_size * 1\nimage_shape = (32, 32, 1)\n\ndef preprocess(features):\n image = tf.image.resize(features['image'], image_shape[:2]) \n image = tf.cast(image, tf.float32)\n image = (image-127.5)/127.5\n label = features['label']\n return image, label\n\n\nds_train = ds_train.map(preprocess)\nds_train = ds_train.cache() # put dataset into memory\nds_train = ds_train.shuffle(ds_info.splits['train'].num_examples)\nds_train = ds_train.batch(global_batch_size).repeat()\n\ntrain_num = ds_info.splits['train'].num_examples\ntrain_steps_per_epoch = round(train_num/batch_size)\nprint(train_steps_per_epoch)", "150\n" ], [ "class cDCGAN():\n def __init__(self, input_shape):\n self.z_dim = 100\n self.input_shape = input_shape\n self.num_classes = 10\n # discriminator\n self.n_discriminator = 1 \n self.discriminator = self.build_discriminator()\n self.discriminator.trainable = False\n self.optimizer_discriminator = RMSprop(1e-4)\n \n # build generator pipeline with frozen discriminator\n self.generator = self.build_generator()\n discriminator_output = self.discriminator([self.generator.output, self.generator.input[1]])\n self.model = Model(self.generator.input, discriminator_output)\n self.model.compile(loss = self.bce_loss,\n optimizer = RMSprop(1e-4))\n self.discriminator.trainable = True\n self.bce = tf.keras.losses.BinaryCrossentropy()\n\n def conv_block(self, channels, kernels, strides=1, \n batchnorm=True, activation=True):\n model = tf.keras.Sequential()\n \n model.add(layers.Conv2D(channels, kernels, strides=strides, padding='same'))\n if batchnorm:\n model.add(layers.BatchNormalization()) \n if activation:\n model.add(layers.LeakyReLU(0.2)) \n \n return model\n \n def bce_loss(self, y_true, y_pred):\n\n loss = self.bce(y_true, y_pred)\n\n return loss\n\n def build_generator(self):\n\n DIM = 64\n \n input_label = layers.Input(shape=1, dtype=tf.int32, name='ClassLabel')\n \n one_hot_label = tf.one_hot(input_label, self.num_classes)\n one_hot_label = layers.Reshape((self.num_classes,))(one_hot_label)\n \n input_z = layers.Input(shape=self.z_dim, name='LatentVector')\n x = layers.Concatenate()([input_z, one_hot_label])\n x = layers.Dense(4*4*4*DIM, activation=None)(x)\n x = layers.Reshape((4,4,4*DIM))(x)\n #x = layers.Concatenate()([x, embedding])\n \n x = layers.UpSampling2D((2,2), interpolation=\"bilinear\")(x)\n x = self.conv_block(2*DIM, 5)(x)\n\n x = layers.UpSampling2D((2,2), interpolation=\"bilinear\")(x)\n x = self.conv_block(DIM, 5)(x)\n\n x = layers.UpSampling2D((2,2), interpolation=\"bilinear\")(x)\n output = layers.Conv2D(image_shape[-1], 5, padding='same', activation='tanh')(x)\n \n\n return Model([input_z, input_label], output) \n \n\n def build_discriminator(self):\n DIM = 64\n \n # label\n input_label = layers.Input(shape=[1], dtype =tf.int32, name='ClassLabel')\n encoded_label = tf.one_hot(input_label, self.num_classes)\n embedding = layers.Dense(32 * 32 * 1, activation=None)(encoded_label)\n embedding = layers.Reshape((32, 32, 1))(embedding)\n \n # discriminator\n input_image = layers.Input(shape=self.input_shape, name='Image')\n x = layers.Concatenate()([input_image, embedding])\n x = self.conv_block(DIM, 5, 2, batchnorm=False)(x)\n x = self.conv_block(2*DIM, 5, 2)(x)\n x = self.conv_block(4*DIM, 5, 2)(x)\n x = layers.Flatten()(x)\n\n output = layers.Dense(1, activation='sigmoid')(x)\n return Model([input_image, input_label], output) \n \n def train_discriminator(self, real_images, class_labels, batch_size):\n real_labels = tf.ones(batch_size)\n fake_labels = tf.zeros(batch_size)\n \n g_input = tf.random.normal((batch_size, self.z_dim))\n fake_class_labels = tf.random.uniform((batch_size,1), minval=0, maxval=10, dtype=tf.dtypes.int32)\n fake_images = self.generator.predict([g_input, fake_class_labels])\n \n with tf.GradientTape() as gradient_tape:\n \n # forward pass\n pred_fake = self.discriminator([fake_images, fake_class_labels])\n pred_real = self.discriminator([real_images, class_labels])\n \n # calculate losses\n loss_fake = self.bce_loss(fake_labels, pred_fake)\n loss_real = self.bce_loss(real_labels, pred_real) \n \n # total loss\n total_loss = 0.5*(loss_fake + loss_real)\n \n # apply gradients\n gradients = gradient_tape.gradient(total_loss, self.discriminator.trainable_variables)\n \n self.optimizer_discriminator.apply_gradients(zip(gradients, self.discriminator.trainable_variables))\n\n return loss_fake, loss_real\n \n def train(self, data_generator, batch_size, steps, interval=100):\n \n val_g_input = tf.random.normal((self.num_classes, self.z_dim))\n val_class_labels = np.arange(self.num_classes)\n real_labels = tf.ones(batch_size)\n \n for i in range(steps):\n \n real_images, class_labels = next(data_generator)\n loss_fake, loss_real = self.train_discriminator(real_images, class_labels, batch_size)\n discriminator_loss = 0.5*(loss_fake + loss_real)\n \n # train generator\n g_input = tf.random.normal((batch_size, self.z_dim))\n fake_class_labels = tf.random.uniform((batch_size, 1), \n minval=0, maxval=self.num_classes, dtype=tf.dtypes.int32)\n g_loss = self.model.train_on_batch([g_input, fake_class_labels], real_labels)\n if i%interval == 0:\n msg = \"Step {}: discriminator_loss {:.4f} g_loss {:.4f}\"\\\n .format(i, discriminator_loss, g_loss)\n print(msg)\n \n fake_images = self.generator.predict([val_g_input,val_class_labels])\n self.plot_images(fake_images)\n \n def plot_images(self, images): \n grid_row = 1\n grid_col = 10\n f, axarr = plt.subplots(grid_row, grid_col, figsize=(grid_col*1.5, grid_row*1.5))\n for col in range(grid_col):\n axarr[col].imshow((images[col,:,:,0]+1)/2, cmap='gray')\n axarr[col].axis('off') \n plt.show()\n \n def sample_images(self, class_labels):\n z = tf.random.normal((len(class_labels), self.z_dim))\n images = self.generator.predict([z,class_labels])\n self.plot_images(images)\n return images\n \n", "_____no_output_____" ], [ "cdcgan = cDCGAN(image_shape)", "_____no_output_____" ], [ "tf.keras.utils.plot_model(cdcgan.discriminator, show_shapes=True)", "_____no_output_____" ], [ "tf.keras.utils.plot_model(cdcgan.generator, show_shapes=True)", "_____no_output_____" ], [ "cdcgan.train(iter(ds_train), batch_size, 2000, 200)", "Step 0: discriminator_loss 0.6841 g_loss 0.6252\n" ], [ "for i in range(5):\n images = cdcgan.sample_images(np.array([0,1,2,3,4,5,6,7,8,9]))\n", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "Apache-2.0", "CC-BY-4.0" ]

[ "Apache-2.0", "CC-BY-4.0" ]

[ "Apache-2.0", "CC-BY-4.0" ]

[ [ [ "# 网络参数的初始化\n\n[](https://gitee.com/mindspore/docs/blob/master/docs/mindspore/programming_guide/source_zh_cn/initializer.ipynb)&emsp;[](https://obs.dualstack.cn-north-4.myhuaweicloud.com/mindspore-website/notebook/master/programming_guide/zh_cn/mindspore_initializer.ipynb)&emsp;[](https://authoring-modelarts-cnnorth4.huaweicloud.com/console/lab?share-url-b64=aHR0cHM6Ly9vYnMuZHVhbHN0YWNrLmNuLW5vcnRoLTQubXlodWF3ZWljbG91ZC5jb20vbWluZHNwb3JlLXdlYnNpdGUvbm90ZWJvb2svbW9kZWxhcnRzL3Byb2dyYW1taW5nX2d1aWRlL21pbmRzcG9yZV9pbml0aWFsaXplci5pcHluYg==&imageid=65f636a0-56cf-49df-b941-7d2a07ba8c8c)", "_____no_output_____" ], [ "## 概述\n\nMindSpore提供了权重初始化模块，用户可以通过封装算子和initializer方法来调用字符串、Initializer子类或自定义Tensor等方式完成对网络参数进行初始化。Initializer类是MindSpore中用于进行初始化的基本数据结构，其子类包含了几种不同类型的数据分布（Zero，One，XavierUniform，HeUniform，HeNormal，Constant，Uniform，Normal，TruncatedNormal）。下面针对封装算子和initializer方法两种参数初始化模式进行详细介绍。", "_____no_output_____" ], [ "## 使用封装算子对参数初始化 \nMindSpore提供了多种参数初始化的方式，并在部分算子中封装了参数初始化的功能。本节将介绍带有参数初始化功能的算子对参数进行初始化的方法，以`Conv2d`算子为例，分别介绍以字符串，`Initializer`子类和自定义`Tensor`等方式对网络中的参数进行初始化，以下代码示例中均以`Initializer`的子类`Normal`为例，代码示例中`Normal`均可替换成`Initializer`子类中任何一个。", "_____no_output_____" ], [ "### 字符串 \n使用字符串对网络参数进行初始化，字符串的内容需要与`Initializer`子类的名称保持一致，使用字符串方式进行初始化将使用`Initializer`子类中的默认参数，例如使用字符串`Normal`等同于使用`Initializer`的子类`Normal()`，代码样例如下：", "_____no_output_____" ] ], [ [ "import numpy as np\nimport mindspore.nn as nn\nfrom mindspore import Tensor\nfrom mindspore.common import set_seed\n\nset_seed(1)\n\ninput_data = Tensor(np.ones([1, 3, 16, 50], dtype=np.float32))\nnet = nn.Conv2d(3, 64, 3, weight_init='Normal')\noutput = net(input_data)\nprint(output)", "[[[[ 3.10382620e-02 4.38603461e-02 4.38603461e-02 ... 4.38603461e-02\n 4.38603461e-02 1.38719045e-02]\n [ 3.26051228e-02 3.54298912e-02 3.54298912e-02 ... 3.54298912e-02\n 3.54298912e-02 -5.54019120e-03]\n [ 3.26051228e-02 3.54298912e-02 3.54298912e-02 ... 3.54298912e-02\n 3.54298912e-02 -5.54019120e-03]\n ...\n [ 3.26051228e-02 3.54298912e-02 3.54298912e-02 ... 3.54298912e-02\n 3.54298912e-02 -5.54019120e-03]\n [ 3.26051228e-02 3.54298912e-02 3.54298912e-02 ... 3.54298912e-02\n 3.54298912e-02 -5.54019120e-03]\n [ 9.66199022e-03 1.24104535e-02 1.24104535e-02 ... 1.24104535e-02\n 1.24104535e-02 -1.38977719e-02]]\n\n ...\n\n [[ 3.98553275e-02 -1.35465711e-03 -1.35465711e-03 ... -1.35465711e-03\n -1.35465711e-03 -1.00310734e-02]\n [ 4.38403059e-03 -3.60766202e-02 -3.60766202e-02 ... -3.60766202e-02\n -3.60766202e-02 -2.95619294e-02]\n [ 4.38403059e-03 -3.60766202e-02 -3.60766202e-02 ... -3.60766202e-02\n -3.60766202e-02 -2.95619294e-02]\n ...\n [ 4.38403059e-03 -3.60766202e-02 -3.60766202e-02 ... -3.60766202e-02\n -3.60766202e-02 -2.95619294e-02]\n [ 4.38403059e-03 -3.60766202e-02 -3.60766202e-02 ... -3.60766202e-02\n -3.60766202e-02 -2.95619294e-02]\n [ 1.33139016e-02 6.74417242e-05 6.74417242e-05 ... 6.74417242e-05\n 6.74417242e-05 -2.27325838e-02]]]]\n" ] ], [ [ "### Initializer子类 \n使用`Initializer`子类对网络参数进行初始化，与使用字符串对参数进行初始化的效果类似，不同的是使用字符串进行参数初始化是使用`Initializer`子类的默认参数，如要使用`Initializer`子类中的参数，就必须使用`Initializer`子类的方式对参数进行初始化，以`Normal(0.2)`为例，代码样例如下：", "_____no_output_____" ] ], [ [ "import numpy as np\nimport mindspore.nn as nn\nfrom mindspore import Tensor\nfrom mindspore.common import set_seed\nfrom mindspore.common.initializer import Normal\n\nset_seed(1)\n\ninput_data = Tensor(np.ones([1, 3, 16, 50], dtype=np.float32))\nnet = nn.Conv2d(3, 64, 3, weight_init=Normal(0.2))\noutput = net(input_data)\nprint(output)", "[[[[ 6.2076533e-01 8.7720710e-01 8.7720710e-01 ... 8.7720710e-01\n 8.7720710e-01 2.7743810e-01]\n [ 6.5210247e-01 7.0859784e-01 7.0859784e-01 ... 7.0859784e-01\n 7.0859784e-01 -1.1080378e-01]\n [ 6.5210247e-01 7.0859784e-01 7.0859784e-01 ... 7.0859784e-01\n 7.0859784e-01 -1.1080378e-01]\n ...\n [ 6.5210247e-01 7.0859784e-01 7.0859784e-01 ... 7.0859784e-01\n 7.0859784e-01 -1.1080378e-01]\n [ 6.5210247e-01 7.0859784e-01 7.0859784e-01 ... 7.0859784e-01\n 7.0859784e-01 -1.1080378e-01]\n [ 1.9323981e-01 2.4820906e-01 2.4820906e-01 ... 2.4820906e-01\n 2.4820906e-01 -2.7795550e-01]]\n\n ...\n\n [[ 7.9710668e-01 -2.7093157e-02 -2.7093157e-02 ... -2.7093157e-02\n -2.7093157e-02 -2.0062150e-01]\n [ 8.7680638e-02 -7.2153252e-01 -7.2153252e-01 ... -7.2153252e-01\n -7.2153252e-01 -5.9123868e-01]\n [ 8.7680638e-02 -7.2153252e-01 -7.2153252e-01 ... -7.2153252e-01\n -7.2153252e-01 -5.9123868e-01]\n ...\n [ 8.7680638e-02 -7.2153252e-01 -7.2153252e-01 ... -7.2153252e-01\n -7.2153252e-01 -5.9123868e-01]\n [ 8.7680638e-02 -7.2153252e-01 -7.2153252e-01 ... -7.2153252e-01\n -7.2153252e-01 -5.9123868e-01]\n [ 2.6627803e-01 1.3488382e-03 1.3488382e-03 ... 1.3488382e-03\n 1.3488382e-03 -4.5465171e-01]]]]\n" ] ], [ [ "### 自定义的Tensor \n除上述两种初始化方法外，当网络要使用MindSpore中没有的数据类型对参数进行初始化，用户可以通过自定义`Tensor`的方式来对参数进行初始化，代码样例如下：", "_____no_output_____" ] ], [ [ "import numpy as np\nimport mindspore.nn as nn\nfrom mindspore import Tensor\nfrom mindspore import dtype as mstype\n\nweight = Tensor(np.ones([64, 3, 3, 3]), dtype=mstype.float32)\ninput_data = Tensor(np.ones([1, 3, 16, 50], dtype=np.float32))\nnet = nn.Conv2d(3, 64, 3, weight_init=weight)\noutput = net(input_data)\nprint(output)", "[[[[12. 18. 18. ... 18. 18. 12.]\n [18. 27. 27. ... 27. 27. 18.]\n [18. 27. 27. ... 27. 27. 18.]\n ...\n [18. 27. 27. ... 27. 27. 18.]\n [18. 27. 27. ... 27. 27. 18.]\n [12. 18. 18. ... 18. 18. 12.]]\n\n ...\n\n [[12. 18. 18. ... 18. 18. 12.]\n [18. 27. 27. ... 27. 27. 18.]\n [18. 27. 27. ... 27. 27. 18.]\n ...\n [18. 27. 27. ... 27. 27. 18.]\n [18. 27. 27. ... 27. 27. 18.]\n [12. 18. 18. ... 18. 18. 12.]]]]\n" ] ], [ [ "## 使用initializer方法对参数初始化\n\n在上述代码样例中，给出了如何在网络中进行参数初始化的方法，如在网络中使用nn层封装`Conv2d`算子，参数`weight_init`作为要初始化的数据类型传入`Conv2d`算子，算子会在初始化时通过调用`Parameter`类，进而调用封装在`Parameter`类中的`initializer`方法来完成对参数的初始化。然而有一些算子并没有像`Conv2d`那样在内部对参数初始化的功能进行封装，如`Conv3d`算子的权重就是作为参数传入`Conv3d`算子，此时就需要手动的定义权重的初始化。\n\n当对参数进行初始化时，可以使用`initializer`方法调用`Initializer`子类中不同的数据类型来对参数进行初始化，进而产生不同类型的数据。\n\n使用initializer进行参数初始化时，支持传入的参数有`init`、`shape`、`dtype`：\n\n- `init`：支持传入`Tensor`、 `str`、 `Initializer的子类`。\n\n- `shape`：支持传入`list`、 `tuple`、 `int`。\n\n- `dtype`：支持传入`mindspore.dtype`。", "_____no_output_____" ], [ "### init参数为Tensor", "_____no_output_____" ], [ "代码样例如下：", "_____no_output_____" ], [ "```python\nimport numpy as np\nfrom mindspore import Tensor\nfrom mindspore import dtype as mstype\nfrom mindspore.common import set_seed\nfrom mindspore.common.initializer import initializer\nfrom mindspore.ops.operations import nn_ops as nps\n\nset_seed(1)\n\ninput_data = Tensor(np.ones([16, 3, 10, 32, 32]), dtype=mstype.float32)\nweight_init = Tensor(np.ones([32, 3, 4, 3, 3]), dtype=mstype.float32)\nweight = initializer(weight_init, shape=[32, 3, 4, 3, 3])\nconv3d = nps.Conv3D(out_channel=32, kernel_size=(4, 3, 3))\noutput = conv3d(input_data, weight)\nprint(output)\n```", "_____no_output_____" ], [ "输出如下：\n\n```text\n[[[[[108 108 108 ... 108 108 108]\n [108 108 108 ... 108 108 108]\n [108 108 108 ... 108 108 108]\n ...\n [108 108 108 ... 108 108 108]\n [108 108 108 ... 108 108 108]\n [108 108 108 ... 108 108 108]]\n ...\n [[108 108 108 ... 108 108 108]\n [108 108 108 ... 108 108 108]\n [108 108 108 ... 108 108 108]\n ...\n [108 108 108 ... 108 108 108]\n [108 108 108 ... 108 108 108]\n [108 108 108 ... 108 108 108]]]]]\n```", "_____no_output_____" ], [ "### init参数为str", "_____no_output_____" ], [ "代码样例如下：", "_____no_output_____" ], [ "```python\nimport numpy as np\nfrom mindspore import Tensor\nfrom mindspore import dtype as mstype\nfrom mindspore.common import set_seed\nfrom mindspore.common.initializer import initializer\nfrom mindspore.ops.operations import nn_ops as nps\n\nset_seed(1)\n\ninput_data = Tensor(np.ones([16, 3, 10, 32, 32]), dtype=mstype.float32)\nweight = initializer('Normal', shape=[32, 3, 4, 3, 3], dtype=mstype.float32)\nconv3d = nps.Conv3D(out_channel=32, kernel_size=(4, 3, 3))\noutput = conv3d(input_data, weight)\nprint(output)\n```", "_____no_output_____" ], [ "输出如下：\n\n```text\n[[[[[0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]]\n ...\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]]\n ...\n [[0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]]\n ...\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]]]]]\n```", "_____no_output_____" ], [ "### init参数为Initializer子类", "_____no_output_____" ], [ "代码样例如下：", "_____no_output_____" ], [ "```python\nimport numpy as np\nfrom mindspore import Tensor\nfrom mindspore import dtype as mstype\nfrom mindspore.common import set_seed\nfrom mindspore.ops.operations import nn_ops as nps\nfrom mindspore.common.initializer import Normal, initializer\n\nset_seed(1)\n\ninput_data = Tensor(np.ones([16, 3, 10, 32, 32]), dtype=mstype.float32)\nweight = initializer(Normal(0.2), shape=[32, 3, 4, 3, 3], dtype=mstype.float32)\nconv3d = nps.Conv3D(out_channel=32, kernel_size=(4, 3, 3))\noutput = conv3d(input_data, weight)\nprint(output)\n```", "_____no_output_____" ], [ "```text\n[[[[[0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]]\n ...\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]]\n ...\n [[0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]]\n ...\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]]]]]\n```", "_____no_output_____" ], [ "### 在Parameter中的应用", "_____no_output_____" ], [ "代码样例如下：", "_____no_output_____" ] ], [ [ "import numpy as np\nfrom mindspore import dtype as mstype\nfrom mindspore.common import set_seed\nfrom mindspore.ops import operations as ops\nfrom mindspore import Tensor, Parameter, context\nfrom mindspore.common.initializer import Normal, initializer\n\nset_seed(1)\n\nweight1 = Parameter(initializer('Normal', [5, 4], mstype.float32), name=\"w1\")\nweight2 = Parameter(initializer(Normal(0.2), [5, 4], mstype.float32), name=\"w2\")\ninput_data = Tensor(np.arange(20).reshape(5, 4), dtype=mstype.float32)\nnet = ops.Add()\noutput = net(input_data, weight1)\noutput = net(output, weight2)\nprint(output)", "[[-0.3305102 1.0412874 2.0412874 3.0412874]\n [ 4.0412874 4.9479127 5.9479127 6.9479127]\n [ 7.947912 9.063009 10.063009 11.063009 ]\n [12.063009 13.536987 14.536987 14.857441 ]\n [15.751231 17.073082 17.808317 19.364822 ]]\n" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import numpy as np\nimport pandas as pd\n \n# load the contents of a file into a pandas Dataframe\ninput_file = '/Users/aurelianosancho/Google Drive/Pre_Processing/train.csv'\ndf_titanic = pd.read_csv(input_file)", "_____no_output_____" ] ], [ [ "$\\textbf{NOTE}$ Although it is not demonstrated in this section, you must ensure that any feature engineering or imputation that is carried out on the training data is also carried out on the test data.", "_____no_output_____" ] ], [ [ "df_titanic.shape", "_____no_output_____" ], [ "df_titanic.columns", "_____no_output_____" ], [ "df_titanic.info()", "<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 891 entries, 0 to 890\nData columns (total 12 columns):\n # Column Non-Null Count Dtype \n--- ------ -------------- ----- \n 0 PassengerId 891 non-null int64 \n 1 Survived 891 non-null int64 \n 2 Pclass 891 non-null int64 \n 3 Name 891 non-null object \n 4 Sex 891 non-null object \n 5 Age 714 non-null float64\n 6 SibSp 891 non-null int64 \n 7 Parch 891 non-null int64 \n 8 Ticket 891 non-null object \n 9 Fare 891 non-null float64\n 10 Cabin 204 non-null object \n 11 Embarked 889 non-null object \ndtypes: float64(2), int64(5), object(5)\nmemory usage: 83.7+ KB\n" ], [ "df_titanic.describe()", "_____no_output_____" ], [ "df_titanic.isnull().sum()", "_____no_output_____" ], [ "df_titanic.head()", "_____no_output_____" ], [ "print(df_titanic.index.name)", "None\n" ] ], [ [ "To make the PassengerId attribute the index of the df_titanic dataframe, use the following snippet:", "_____no_output_____" ] ], [ [ "df_titanic.set_index(\"PassengerId\", inplace=True)", "_____no_output_____" ], [ "print(df_titanic.index.name)", "PassengerId\n" ], [ "df_titanic.head()", "_____no_output_____" ], [ "# extract the target attribute into its own dataframe\ndf_titanic_target = df_titanic.loc[:,['Survived']]\n \n# create a dataframe that contains the 10 feature variables\ndf_titanic_features = df_titanic.drop(['Survived'], axis=1)", "_____no_output_____" ], [ "df_titanic_target['Survived'].value_counts()", "_____no_output_____" ], [ "df_titanic_features['Embarked'].value_counts(dropna=False)", "_____no_output_____" ], [ "# histogram of target variable\n%matplotlib inline\nimport matplotlib.pyplot as plt\ndf_titanic_target.hist(figsize=(5,5))", "_____no_output_____" ], [ "df_titanic_features.hist(figsize=(10,10))", "/opt/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:298: MatplotlibDeprecationWarning: \nThe rowNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().rowspan.start instead.\n layout[ax.rowNum, ax.colNum] = ax.get_visible()\n/opt/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:298: MatplotlibDeprecationWarning: \nThe colNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().colspan.start instead.\n layout[ax.rowNum, ax.colNum] = ax.get_visible()\n/opt/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:304: MatplotlibDeprecationWarning: \nThe rowNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().rowspan.start instead.\n if not layout[ax.rowNum + 1, ax.colNum]:\n/opt/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:304: MatplotlibDeprecationWarning: \nThe colNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().colspan.start instead.\n if not layout[ax.rowNum + 1, ax.colNum]:\n" ], [ "# histogram of categorical attribute 'Embarked'\n# computed from the output of the value_counts() function\nvc = df_titanic_features['Embarked'].value_counts(dropna=False)\nvc.plot(kind='bar')", "_____no_output_____" ], [ "# create a box plot of numeric features.\ndf_titanic_features.boxplot(figsize=(10,6)) ", "_____no_output_____" ], [ "# what features show the strongest correlation with the target variable?\ncorr_matrix = df_titanic.corr()\ncorr_matrix['Survived'].sort_values(ascending=False)", "_____no_output_____" ], [ "# visualize relationship between features using a\n# matrix of scatter plots.\nfrom pandas.plotting import scatter_matrix\nscatter_matrix(df_titanic, figsize=(12,12)) ", "/opt/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:298: MatplotlibDeprecationWarning: \nThe rowNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().rowspan.start instead.\n layout[ax.rowNum, ax.colNum] = ax.get_visible()\n/opt/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:298: MatplotlibDeprecationWarning: \nThe colNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().colspan.start instead.\n layout[ax.rowNum, ax.colNum] = ax.get_visible()\n/opt/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:304: MatplotlibDeprecationWarning: \nThe rowNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().rowspan.start instead.\n if not layout[ax.rowNum + 1, ax.colNum]:\n/opt/anaconda3/lib/python3.7/site-packages/pandas/plotting/_matplotlib/tools.py:304: MatplotlibDeprecationWarning: \nThe colNum attribute was deprecated in Matplotlib 3.2 and will be removed two minor releases later. Use ax.get_subplotspec().colspan.start instead.\n if not layout[ax.rowNum + 1, ax.colNum]:\n" ], [ "df_titanic_features.boxplot(column='Age', figsize=(7,7))", "_____no_output_____" ], [ "# fill missing values with the median\nmedian_age = df_titanic_features['Age'].median()\nprint (median_age)\n28.0\n \ndf_titanic_features[\"Age\"].fillna(median_age, inplace=True) ", "28.0\n" ], [ "# fill missing values of the Embarked attribute\n# with the most common value in the column\n\nembarked_value_counts = df_titanic_features['Embarked'].value_counts(dropna=True)\nmost_common_value = embarked_value_counts.index[0]\n \nprint (most_common_value)\n \ndf_titanic_features[\"Embarked\"].fillna(most_common_value, inplace=True) ", "S\n" ], [ "# create a boolean feature 'CabinIsKnown'\n# which will have True if the Cabin column\n# does not have missing data\ndf_titanic_features['CabinIsKnown'] = ~df_titanic_features.Cabin.isnull()\n \n# drop the Cabin column from the dataframe\ndf_titanic_features.drop(['Cabin'], axis=1, inplace=True)", "_____no_output_____" ], [ "# display the columns of the dataframe.\nprint (df_titanic_features.columns.values)\n \n# display number of missing values in the columns\ndf_titanic_features.isnull().sum()", "['Pclass' 'Name' 'Sex' 'Age' 'SibSp' 'Parch' 'Ticket' 'Fare' 'Embarked'\n 'CabinIsKnown']\n" ], [ "# create a numeric feature called FamilySize that is\n# the sum of the SibSp and Parch features.\ndf_titanic_features['FamilySize'] = df_titanic_features.SibSp + df_titanic_features.Parch", "_____no_output_____" ], [ "# generate new categorical feature AgeCategory\nbins_age = [0,20,30,40,50,150]\nlabels_age = ['<20','20-30','30-40','40-50','>50']\n \ndf_titanic_features['AgeCategory'] = pd.cut(df_titanic_features.Age,\n bins=bins_age,\n labels=labels_age,\n include_lowest=True)", "_____no_output_____" ], [ "df_titanic_features.head()", "_____no_output_____" ], [ "# generate new categorical feature FareCategory\ndf_titanic_features['FareCategory'] = pd.qcut(df_titanic_features.Fare,\n q=4,\n labels=['Q1', 'Q2', 'Q3', 'Q4']) ", "_____no_output_____" ], [ "# use one-hot encoding to convert categorical attributes\n# into binary numeric attributes\ndf_titanic_features = pd.get_dummies(df_titanic_features, columns=['Sex','Embarked','CabinIsKnown','AgeCategory','FareCategory'])\n \n# display the columns of the dataframe.\nprint (df_titanic_features.columns.values)", "['Pclass' 'Name' 'Age' 'SibSp' 'Parch' 'Ticket' 'Fare' 'FamilySize'\n 'Sex_female' 'Sex_male' 'Embarked_C' 'Embarked_Q' 'Embarked_S'\n 'CabinIsKnown_False' 'CabinIsKnown_True' 'AgeCategory_<20'\n 'AgeCategory_20-30' 'AgeCategory_30-40' 'AgeCategory_40-50'\n 'AgeCategory_>50' 'FareCategory_Q1' 'FareCategory_Q2' 'FareCategory_Q3'\n 'FareCategory_Q4']\n" ], [ "df_titanic_features.head()", "_____no_output_____" ], [ "# strong negative correlation between Sex_male and Sex_female.\n# one of these can be dropped.\ncorr_matrix = df_titanic_features[['Sex_male', 'Sex_female']].corr()\nprint(corr_matrix)", " Sex_male Sex_female\nSex_male 1.0 -1.0\nSex_female -1.0 1.0\n" ], [ "# drop the Name, Ticket, Sex_female, CabinIsKnown_False features\n# to get a dataframe that can be used for linear or logistic regression\ndf_titanic_features_numeric = df_titanic_features.drop(['Name', 'Ticket', 'Sex_female', 'CabinIsKnown_False'], axis=1)", "_____no_output_____" ], [ "df_titanic_features_numeric.head()", "_____no_output_____" ], [ "df_titanic_features_numeric.shape", "_____no_output_____" ], [ "####################### pre-processing Test #######################", "_____no_output_____" ], [ "input_file = '/Users/aurelianosancho/Google Drive/Pre_Processing/train.csv'\ndf_titanic_test = pd.read_csv(input_file)\ndf_titanic_test.set_index(\"PassengerId\", inplace=True)\n\ndf_titanic_test_target = df_titanic_test.loc[:,['Survived']]\ndf_titanic_test_features = df_titanic_test.drop(['Survived'], axis=1)\n\nmedian_age = df_titanic_test_features['Age'].median()\ndf_titanic_test_features[\"Age\"].fillna(median_age, inplace=True)\n\nembarked_value_counts = df_titanic_test_features['Embarked'].value_counts(dropna=True)\nmost_common_value = embarked_value_counts.index[0]\ndf_titanic_test_features[\"Embarked\"].fillna(most_common_value, inplace=True) \n\ndf_titanic_test_features['CabinIsKnown'] = ~df_titanic_test_features.Cabin.isnull()\ndf_titanic_test_features.drop(['Cabin'], axis=1, inplace=True)\n\ndf_titanic_test_features['FamilySize'] = df_titanic_test_features.SibSp + df_titanic_test_features.Parch\n\nbins_age = [0,20,30,40,50,150]\nlabels_age = ['<20','20-30','30-40','40-50','>50']\n \ndf_titanic_test_features['AgeCategory'] = pd.cut(df_titanic_test_features.Age,\n bins=bins_age,\n labels=labels_age,\n include_lowest=True)\n\ndf_titanic_test_features['FareCategory'] = pd.qcut(df_titanic_test_features.Fare,\n q=4,\n labels=['Q1', 'Q2', 'Q3', 'Q4']) \n\n\ndf_titanic_test_features = pd.get_dummies(df_titanic_test_features, columns=['Sex','Embarked','CabinIsKnown','AgeCategory','FareCategory'])\n \ndf_titanic_test_features_numeric = df_titanic_test_features.drop(['Name', 'Ticket', 'Sex_female', 'CabinIsKnown_False'], axis=1)", "_____no_output_____" ] ], [ [ "* titanic_features_train = df_titanic_features_numeric\n\n* titanic_features_test = df_titanic_test_features_numeric\n\n* titanic_target_train = df_titanic_test_target\n\n* titanic_target_test = df_titanic_target", "_____no_output_____" ] ], [ [ "df_titanic_test_features_numeric.shape", "_____no_output_____" ], [ "titanic_features_train = df_titanic_features_numeric\ntitanic_features_test = df_titanic_test_features_numeric\ntitanic_target_train = df_titanic_target\ntitanic_target_test = df_titanic_test_target ", "_____no_output_____" ], [ "from sklearn.svm import SVC\nsvc_model = SVC(kernel='rbf', C=1, gamma='auto', probability=True)\nsvc_model.fit(titanic_features_train, titanic_target_train.values.ravel())\n \n# train a logistic regression model on the diabetes dataset\nfrom sklearn.linear_model import LogisticRegression\nlogit_model = LogisticRegression(penalty='l2', fit_intercept=True, solver='liblinear')\nlogit_model.fit(titanic_features_train, titanic_target_train.values.ravel())\n \n# train a decision tree based binary classifier.\nfrom sklearn.tree import DecisionTreeClassifier\n \ndtree_model = DecisionTreeClassifier(max_depth=4)\ndtree_model.fit(titanic_features_train, titanic_target_train.values.ravel())\n \n# use the models to create predictions on the diabetes test set\nsvc_predictions = svc_model.predict(titanic_features_test)\nlogit_predictions = logit_model.predict(titanic_features_test)\ndtree_predictions = dtree_model.predict(titanic_features_test)\n \n# simplistic metric - the percentage of correct predictions\nsvc_correct = svc_predictions == titanic_target_test.values.ravel()\nsvc_correct_percent = np.count_nonzero(svc_correct) / svc_predictions.size * 100\n\nlogit_correct = logit_predictions == titanic_target_test.values.ravel()\nlogit_correct_percent = np.count_nonzero(logit_correct) / logit_predictions.size * 100\n \ndtree_correct = dtree_predictions == titanic_target_test.values.ravel()\ndtree_correct_percent = np.count_nonzero(dtree_correct) / dtree_predictions.size * 100", "_____no_output_____" ], [ "print ('SVC', svc_correct_percent, 'Logistic Regression', logit_correct_percent, 'DecisionTree', dtree_correct_percent)\n", "SVC 84.73625140291807 Logistic Regression 81.03254769921436 DecisionTree 83.61391694725027\n" ], [ "from sklearn.metrics import confusion_matrix\ncm_svc = confusion_matrix(titanic_target_test.values.ravel(), svc_predictions)\ncm_logit = confusion_matrix(titanic_target_test.values.ravel(), logit_predictions)\ncm_dtree = confusion_matrix(titanic_target_test.values.ravel(), dtree_predictions)", "_____no_output_____" ], [ "cm_svc", "_____no_output_____" ], [ "cm_logit", "_____no_output_____" ], [ "cm_dtree", "_____no_output_____" ], [ "tn_svc, fp_svc, fn_svc, tp_svc = cm_svc.ravel()\ntn_logit, fp_logit, fn_logit, tp_logit = cm_logit.ravel()\ntn_dtree, fp_dtree, fn_dtree, tp_dtree = cm_dtree.ravel()", "_____no_output_____" ], [ "print (tn_svc, fp_svc, fn_svc, tp_svc)\n\nprint (tn_logit, fp_logit, fn_logit, tp_logit)\n\nprint (tn_dtree, fp_dtree, fn_dtree, tp_dtree)", "513 36 100 242\n472 77 92 250\n499 50 96 246\n" ], [ "accuracy_svc = (tp_svc + tn_svc) / (tn_svc + fp_svc + fn_svc + tp_svc)\naccuracy_logit = (tp_logit + tn_logit) / (tn_logit + fp_logit + fn_logit + tp_logit)\naccuracy_dtree = (tp_dtree + tn_dtree) / (tn_dtree + fp_dtree + fn_dtree + tp_dtree)\n \nprecision_svc = tp_svc / (tp_svc + fp_svc)\nprecision_logit = tp_logit / (tp_logit + fp_logit)\nprecision_dtree = tp_dtree / (tp_dtree + fp_dtree)\n \nrecall_svc = tp_svc / (tp_svc + fn_svc)\nrecall_logit = tp_logit / (tp_svc + fn_logit)\nrecall_dtree = tp_dtree / (tp_dtree + fn_dtree)", "_____no_output_____" ], [ "print('Accuracy SVC:',accuracy_svc, 'Accuracy REG:', accuracy_logit, 'Accuracy DTREE:', accuracy_dtree)\n \nprint('Precision SVC:',precision_svc, 'Precision REG:',precision_logit, 'Precision DTREE:',precision_dtree)\n \nprint('Recall SVC:', recall_svc, 'Recall REG:',recall_logit, 'Recall DTREE:',recall_dtree)", "Accuracy SVC: 0.8473625140291807 Accuracy REG: 0.8103254769921436 Accuracy DTREE: 0.8361391694725028\nPrecision SVC: 0.8705035971223022 Precision REG: 0.764525993883792 Precision DTREE: 0.831081081081081\nRecall SVC: 0.7076023391812866 Recall REG: 0.7485029940119761 Recall DTREE: 0.7192982456140351\n" ], [ "# plot ROC curves for the three classifiers.\n \n# compute prediction probabilities\nsvc_probabilities = svc_model.predict_proba(titanic_features_test)\nlogit_probabilities = logit_model.predict_proba(titanic_features_test)\ndtree_probabilities = dtree_model.predict_proba(titanic_features_test)\n \n# calculate the FPR and TPR for all thresholds of the SVC model\nimport sklearn.metrics as metrics\nsvc_fpr, svc_tpr, svc_thresholds = metrics.roc_curve(titanic_target_test.values.ravel(),\n svc_probabilities[:,1],\n pos_label=1,\n drop_intermediate=False)\n\n\nlogit_fpr, logit_tpr, logit_thresholds = metrics.roc_curve(titanic_target_test.values.ravel(),\n logit_probabilities[:,1],pos_label=1,\n drop_intermediate=False)\n \n# calculate the FPR and TPR for all thresholds of the decision tree model\ndtree_fpr, dtree_tpr, dtree_thresholds = metrics.roc_curve(titanic_target_test.values.ravel(),\n dtree_probabilities[:,1],pos_label=1,\n drop_intermediate=False)\n \n \n \nfig, axes = plt.subplots(1, 3, figsize=(18,6))\n \naxes[0].set_title('ROC curve: SVC model')\naxes[0].set_xlabel(\"True Positive Rate\")\naxes[0].set_ylabel(\"False Positive Rate\")\naxes[0].plot(svc_fpr, svc_tpr)\naxes[0].axhline(y=0, color='k')\naxes[0].axvline(x=0, color='k')\n \naxes[1].set_title('ROC curve: Logit model')\n\naxes[1].set_xlabel(\"True Positive Rate\")\naxes[1].set_ylabel(\"False Positive Rate\")\naxes[1].plot(logit_fpr, logit_tpr)\naxes[1].axhline(y=0, color='k')\naxes[1].axvline(x=0, color='k')\n \naxes[2].set_title('ROC curve: Tree model')\naxes[2].set_xlabel(\"True Positive Rate\")\naxes[2].set_ylabel(\"False Positive Rate\")\naxes[2].plot(dtree_fpr, dtree_tpr)\naxes[2].axhline(y=0, color='k')\naxes[2].axvline(x=0, color='k')", "_____no_output_____" ], [ "svc_auc = metrics.auc(svc_fpr, svc_tpr)\nlogit_auc = metrics.auc(logit_fpr, logit_tpr)\ndtree_auc = metrics.auc(dtree_fpr, dtree_tpr)", "_____no_output_____" ], [ "print (svc_auc, logit_auc, dtree_auc)", "0.9137373640537287 0.8673744926980475 0.8772728725274022\n" ] ], [ [ "The following code snippet uses the GridSearchCV class to try different hyperparameter \ncombinations for a multi-class decision tree classifier on the Iris flowers dataset and returns the hyperparameters that result in the best precision score:", "_____no_output_____" ] ], [ [ "# use grid search to find the hyperparameters that result\n# in the best accuracy score for a decision tree\n# based classifier on the Iris Flowers dataset\nfrom sklearn.model_selection import GridSearchCV\nfrom sklearn.tree import DecisionTreeClassifier\n \ngrid_params = {\n 'criterion': ['gini', 'entropy'],\n 'splitter': ['best', 'random'],\n 'max_depth': [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12],\n 'min_samples_split': [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12],\n 'max_features': ['auto', 'sqrt', 'log2'],\n 'presort': [True, False]\n}\n \ngrid_search = GridSearchCV(estimator=DecisionTreeClassifier(),\n param_grid=grid_params, scoring='accuracy',\n cv=5, n_jobs=-1)\n \ngrid_search.fit(titanic_features_train.values, titanic_target_train)", "/opt/anaconda3/lib/python3.7/site-packages/sklearn/tree/_classes.py:319: FutureWarning: The parameter 'presort' is deprecated and has no effect. It will be removed in v0.24. You can suppress this warning by not passing any value to the 'presort' parameter.\n FutureWarning)\n" ], [ "best_parameters = grid_search.best_params_\nprint(best_parameters)\n \nbest_accuracy = grid_search.best_score_\nprint(best_accuracy)", "{'criterion': 'gini', 'max_depth': 12, 'max_features': 'auto', 'min_samples_split': 12, 'presort': False, 'splitter': 'best'}\n0.8204381394764922\n" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]

[ [ [ "# The BioBB REST API\n\nThe **[BioBB REST API](https://mmb.irbbarcelona.org/biobb-api)** allows the execution of the **[BioExcel Building Blocks](https://mmb.irbbarcelona.org/biobb/)** in a remote server.\n\n## Documentation\n\nFor an extense documentation section, please go to the **[BioBB REST API website help](https://mmb.irbbarcelona.org/biobb-api/rest)**.\n\n## Settings\n\n### Auxiliar libraries used\n\n* [requests](https://pypi.org/project/requests/): Requests allows you to send *organic, grass-fed* HTTP/1.1 requests, without the need for manual labor.\n* [nb_conda_kernels](https://github.com/Anaconda-Platform/nb_conda_kernels): Enables a Jupyter Notebook or JupyterLab application in one conda environment to access kernels for Python, R, and other languages found in other environments.\n* [nglview](http://nglviewer.org/#nglview): Jupyter/IPython widget to interactively view molecular structures and trajectories in notebooks.\n* [ipywidgets](https://github.com/jupyter-widgets/ipywidgets): Interactive HTML widgets for Jupyter notebooks and the IPython kernel.\n* [plotly](https://plot.ly/python/offline/): Python interactive graphing library integrated in Jupyter notebooks.\n\n### Conda Installation and Launch\n\n```console\ngit clone https://github.com/bioexcel/biobb_REST_API_documentation.git\ncd biobb_REST_API_documentation\nconda env create -f conda_env/environment.yml\nconda activate biobb_REST_API_documentation\njupyter-nbextension enable --py --user widgetsnbextension\njupyter-nbextension enable --py --user nglview\njupyter-notebook biobb_REST_API_documentation/notebooks/biobb_REST_API_documentation.ipynb\n```\n\n***\n\n## Index\n\n * [Behaviour](#behaviour)\n * [Tools information](#tools_info)\n * [List of packages](#list_pckg)\n * [List of tools](#list_tools)\n * [Tool's properties](#tools_prop)\n * [Launch tool](#launch_tool)\n * [Retrieve status](#retrieve_status)\n * [Retrieve data](#retrieve_data)\n * [Sample files](#sample_files)\n * [All sample files](#all_sample)\n * [Package sample files](#pckg_sample)\n * [Tool sample files](#tool_sample)\n * [Single sample file](#sample)\n * [Examples](#examples)\n * [Tools information](#tools_info_ex)\n * [List of packages](#list_pckg_ex)\n * [List of tools from a specific package](#list_tools_ex)\n * [Tool's properties](#tools_prop_ex)\n * [Launch tool](#launch_tool_ex)\n * [Launch job with a YAML file config](#tool_yml_ex)\n * [Launch job with a JSON file config](#tool_json_ex)\n * [Launch job with a piython dictionary config](#tool_dict_ex)\n * [Retrieve status](#retrieve_status_ex)\n * [Retrieve data](#retrieve_data_ex)\n * [Practical cases](#practical_cases)\n * [Example 1: download PDB file from RSCB database](#example1)\n * [Example 2: extract heteroatom from a given structure](#example2)\n * [Example 3: extract energy components from a given GROMACS energy file](#example3)\n\n***\n<img src=\"https://bioexcel.eu/wp-content/uploads/2019/04/Bioexcell_logo_1080px_transp.png\" alt=\"Bioexcel2 logo\"\n\ttitle=\"Bioexcel2 logo\" width=\"400\" />\n***\n\n<a id=\"behaviour\"></a>\n## Behaviour\n\nThe **BioBB REST API** works as an asynchronous launcher of jobs, as these jobs can last from a few seconds to several minutes, there are some steps that must be performed for having the complete results of every tool.\n\n**BioExcel Building Blocks** are structured in **[packages and tools](http://mmb.irbbarcelona.org/biobb/availability/source)**. Every call to the **BioBB REST API** executes one single tool and returns the output file(s) related to this specific tool.\n\n<a id=\"tools_info\"></a>\n### Tools information\n\n<a id=\"list_pckg\"></a>\n#### List of packages\n\nIn order to get a complete **list of available packages**, we must do a **GET** request to the following endpoint:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch`\n\nThis endpoint returns a **JSON HTTP response** with status `200`. More information in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest/#/List%20of%20Services/getPckgList).\n\n<a id=\"list_tools\"></a>\n#### List of tools\n\nIf there is need for a **list of tools for a single package**, we must do a **GET** request to the following endpoint:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch/{package}`\n\nThis endpoint returns a **JSON HTTP response** with status `200` or a `404` status if the package id is incorrect. More information in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest/#/List%20of%20Services/getToolsList).\n\n<a id=\"tools_prop\"></a>\n#### Tool's properties\n\nIf there is only need for the **information of a single tool**, we must do a **GET** request to the following endpoint:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch/{package}/{tool}`\n\nThis endpoint returns a **JSON HTTP response** with status `200` or a `404` status if the package id and / or the tool id are incorrect. The reason for failure should be detailed in the JSON response. More information in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest/#/Launch%20Tool/getLaunchTool).\n\n<a id=\"launch_tool\"></a>\n### Launch tool\n\nFor **launching a tool**, we must do a **POST** request to the following endpoint:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch/{package}/{tool}`\n\nIn the body of this POST request, **we must add the file(s) needed as input** (included the properties config file in **JSON** or **YAML** format) and the name for the output(s). The detailed list of inputs and outputs with its respectives properties can be found in the **GET** request of this same endpoint.\n\nThis endpoint returns a **JSON HTTP response** with the following possible status:\n\n* `303`: **The job has been successfully launched** and the user must save the token provided and follow to the next endpoint (defined in the same JSON response)\n* `404`: **There was some error launching the tool.** The reason for failure should be detailed in the JSON response.\n* `500`: The job has been launched, but **some internal server error** has occurred during the execution.\n\nMore information for a generic call in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest/#/Launch%20Tool/postLaunchTool). The documentation for all the tools is available in the [BioBB REST API Tools Documentation section](https://mmb.irbbarcelona.org/biobb-api/tools-documentation?docExpansion=none). Interactive examples for all the tools are available in the [BioBB REST API Tools Execution section](https://mmb.irbbarcelona.org/biobb-api/tools-execution).\n\n<a id=\"retrieve_status\"></a>\n### Retrieve status\n\nIf the previous endpoint returned a `303` status, we must do a **GET** request to the following endpoint providing the given token in the path:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/retrieve/status/{token}`\n\nThis endpoint checks the state of the job and returns a **JSON HTTP response** with the following possible status:\n\n* `200`: **The job has finished successfully** and in the JSON response we can found a list of output files generated by the job with its correspondent id for retrieving them on the next endpoint (defined in the same JSON message).\n* `202`: The job is **still running**.\n* `404`: **Token incorrect, job unexisting or expired.**\n* `500`: Some **internal server error** has occurred during the execution.\n\nMore information in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest/#/Retrieve/getRetrieveStatus).\n\n<a id=\"retrieve_data\"></a>\n### Retrieve data\n\nOnce the previous endpoint returns a `200` status, the output file(s) are ready for its retrieval, so we must do a **GET** request to the following endpoint providing the given **file id** in the path:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/retrieve/data/{id}`\n\nThis endpoint returns the **requested file** with a `200` status or a `404` status if the provided id is incorrect, the file doesn't exist or it has expired. More information in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest/#/Retrieve/getRetrieveData).\n\nNote that if we have executed a job that returns multiple output files, a call to this endpoint must be done **for each of the output files** generated by the job.\n\n<a id=\"sample_files\"></a>\n### Sample files\n\nThe **BioBB REST API** provides sample files for most of the inputs and outputs of each tool. Files can be accessed thought the whole **BioBB REST API** hierarchical range.\n\n<a id=\"all_sample\"></a>\n#### All sample files\n\nIn order to download **all the sample files**, we must do a **GET** request to the following endpoint:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/sample`\n\nThis endpoint returns the **requested file** with a `200` status. More information in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/Sample%20Files/getSample).\n\n<a id=\"pckg_sample\"></a>\n#### Package sample files\n\nIn order to download **all the sample files of a package**, we must do a **GET** request to the following endpoint:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/sample/{package}`\n\nThis endpoint returns the **requested file** with a `200` status or a `404` status if the package id is incorrect. More information in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/Sample%20Files/getPackageSample).\n\n<a id=\"tool_sample\"></a>\n#### Tool sample files\n\nIn order to download **all the sample files of a tool**, we must do a **GET** request to the following endpoint:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/sample/{package}/{tool}`\n\nThis endpoint returns the **requested file** with a `200` status or a `404` status if the package id and / or the tool id are incorrect. The reason for failure should be detailed in the JSON response. More information in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/Sample%20Files/getToolSample).\n\n<a id=\"sample\"></a>\n#### Single sample file\n\nIn order to download **a single sample file**, we must do a **GET** request to the following endpoint:\n\n`https://mmb.irbbarcelona.org/biobb-api/rest/v1/sample/{package}/{tool}/{id}`\n\nThis endpoint returns the **requested file** with a `200` status or a `404` status if the package id and / or the tool id and / or the file id are incorrect. The reason for failure should be detailed in the JSON response. More information in the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/Sample%20Files/getSingleSample).", "_____no_output_____" ], [ "<a id=\"examples\"></a>\n## Examples\n\nBelow we will do **calls to all the previously defined endpoints** and define some **functions** for make easier the connection to the **BioBB REST API** through **Jupyter Notebook**.\n\nFirst off, we will import the Python requests and json library and set the root URI for the **BioBB REST API**.", "_____no_output_____" ] ], [ [ "import requests\nimport json\n\napiURL = \"https://mmb.irbbarcelona.org/biobb-api/rest/v1/\"", "_____no_output_____" ] ], [ [ "<a id=\"tools_info_ex\"></a>\n### Tools information", "_____no_output_____" ], [ "Definition of simple GET / POST request functions and a class Response:", "_____no_output_____" ] ], [ [ "# Class for returning response status and json content of a requested URL\nclass Response:\n def __init__(self, status, json):\n self.status = status\n self.json = json\n\n# Perform GET request\ndef get_data(url):\n r = requests.get(url)\n return Response(r.status_code, json.loads(r.text))\n\n# Perform POST request\ndef post_data(url, d, f):\n r = requests.post(url, data = d, files = f)\n return Response(r.status_code, json.loads(r.text))", "_____no_output_____" ] ], [ [ "<a id=\"list_pckg_ex\"></a>\n#### List of packages\n\nFor more information about this endpoint, please visit the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/List%20of%20Services/getPckgList).\n\n##### Endpoint", "_____no_output_____" ], [ "**GET** `https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch`", "_____no_output_____" ], [ "##### Code", "_____no_output_____" ] ], [ [ "url = apiURL + 'launch'\nresponse = get_data(url)\n\nprint(json.dumps(response.json, indent=2))", "{\n \"packages\": [\n {\n \"id\": \"biobb_analysis\",\n \"tools\": [\n {\n \"id\": \"gmx_cluster\",\n \"description\": \"Creates cluster structures from a given GROMACS compatible trajectory\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_cluster tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_cluster.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_pdb_path\",\n \"required\": true,\n \"description\": \"Path to the output cluster file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_cluster.pdb\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_rms\",\n \"description\": \"Performs a Root Mean Square deviation (RMSd) analysis from a given GROMACS compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_rms tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_rms.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_xvg_path\",\n \"required\": true,\n \"description\": \"Path to the XVG output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_rms.xvg\",\n \"formats\": [\n \".*\\\\.xvg$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_rgyr\",\n \"description\": \"Computes the radius of gyration (Rgyr) of a molecule about the x-, y- and z-axes, as a function of time, from a given GROMACS compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_rgyr tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_rgyr.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_xvg_path\",\n \"required\": true,\n \"description\": \"Path to the XVG output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_rgyr.xvg\",\n \"formats\": [\n \".*\\\\.xvg$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_energy\",\n \"description\": \"Extracts energy components from a given GROMACS energy file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_energy tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_energy.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_energy_path\",\n \"required\": true,\n \"description\": \"Path to the input EDR file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/energy.edr\",\n \"formats\": [\n \".*\\\\.edr$\"\n ]\n },\n {\n \"id\": \"output_xvg_path\",\n \"required\": true,\n \"description\": \"Path to the XVG output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_energy.xvg\",\n \"formats\": [\n \".*\\\\.xvg$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_image\",\n \"description\": \"Corrects periodicity (image) from a given GROMACS compatible trajectory file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_image tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_image.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS input topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_traj_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_image.xtc\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_trjconv_str\",\n \"description\": \"Converts between GROMACS compatible structure file formats and/or extracts a selection of atoms.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_trjconv_str tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_trjconv_str.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS input topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_str_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_trjconv.str.pdb\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_trjconv_str_ens\",\n \"description\": \"Extracts an ensemble of frames containing a selection of atoms from GROMACS compatible trajectory files.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_trjconv_str_ens tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_trjconv_str_ens.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS input topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_str_ens_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_trjconv.str.ens.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_trjconv_trj\",\n \"description\": \"Converts between GROMACS compatible trajectory file formats and/or extracts a selection of atoms.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_trjconv_trj tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_trjconv_trj.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_traj_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_trjconv.trj.xtc\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_average\",\n \"description\": \"Calculates a structure average of a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_average tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_average.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed structure\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.average.pdb\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_bfactor\",\n \"description\": \"Calculates the Bfactor fluctuations of a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_bfactor tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"input_exp_path\",\n \"required\": false,\n \"description\": \"Path to the experimental reference file (required if reference = experimental)\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/experimental.1e5t.pdb\",\n \"formats\": null\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed analysis\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.bfactor.first.dat\",\n \"formats\": [\n \".*\\\\.dat$\",\n \".*\\\\.agr$\",\n \".*\\\\.xmgr$\",\n \".*\\\\.gnu$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_rms\",\n \"description\": \"Calculates the Root Mean Square deviation (RMSd) of a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_rms tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"input_exp_path\",\n \"required\": false,\n \"description\": \"Path to the experimental reference file (required if reference = experimental)\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/experimental.1e5t.pdb\",\n \"formats\": null\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed analysis\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.rms.first.dat\",\n \"formats\": [\n \".*\\\\.dat$\",\n \".*\\\\.agr$\",\n \".*\\\\.xmgr$\",\n \".*\\\\.gnu$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_rmsf\",\n \"description\": \"Calculates the Root Mean Square fluctuations (RMSf) of a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_rmsf tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"input_exp_path\",\n \"required\": false,\n \"description\": \"Path to the experimental reference file (required if reference = experimental)\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/experimental.1e5t.pdb\",\n \"formats\": null\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed analysis\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.rmsf.first.dat\",\n \"formats\": [\n \".*\\\\.dat$\",\n \".*\\\\.agr$\",\n \".*\\\\.xmgr$\",\n \".*\\\\.gnu$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_rgyr\",\n \"description\": \"Computes the radius of gyration (Rgyr) from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_rgyr tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_rgyr.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output analysis\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.rgyr.dat\",\n \"formats\": [\n \".*\\\\.dat$\",\n \".*\\\\.agr$\",\n \".*\\\\.xmgr$\",\n \".*\\\\.gnu$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_dry\",\n \"description\": \"Dehydrates a given cpptraj compatible trajectory stripping out solvent molecules and ions.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_dry tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_dry.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.dry.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_strip\",\n \"description\": \"Strips a defined set of atoms (mask) from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_strip tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_strip.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.strip.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_snapshot\",\n \"description\": \"Extracts a particular snapshot from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_snapshot tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_snapshot.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed structure\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.snapshot.pdb\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_slice\",\n \"description\": \"Extracts a particular trajectory slice from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_slice tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_slice.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.slice.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_convert\",\n \"description\": \"Converts between cpptraj compatible trajectory file formats and/or extracts a selection of atoms or frames.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_convert tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_convert.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.convert.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_mask\",\n \"description\": \"Extracts a selection of atoms from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_mask tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_mask.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.mask.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_image\",\n \"description\": \"Corrects periodicity (image) from a given cpptraj trajectory file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_image tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_image.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.image.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n }\n ]\n },\n {\n \"id\": \"biobb_chemistry\",\n \"tools\": [\n {\n \"id\": \"acpype_params_ac\",\n \"description\": \"Small molecule parameterization for AMBER MD package.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the acpype_params_ac tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_chemistry/master/biobb_chemistry/test/data/config/config_acpype_params_ac.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/acpype/acpype.params.mol2\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.mdl$\",\n \".*\\\\.mol2$\"\n ]\n },\n {\n \"id\": \"output_path_frcmod\",\n \"required\": true,\n \"description\": \"Path to the FRCMOD output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.ac.frcmod\",\n \"formats\": [\n \".*\\\\.frcmod$\"\n ]\n },\n {\n \"id\": \"output_path_inpcrd\",\n \"required\": true,\n \"description\": \"Path to the INPCRD output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.ac.inpcrd\",\n \"formats\": [\n \".*\\\\.inpcrd$\"\n ]\n },\n {\n \"id\": \"output_path_lib\",\n \"required\": true,\n \"description\": \"Path to the LIB output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.ac.lib\",\n \"formats\": [\n \".*\\\\.lib$\"\n ]\n },\n {\n \"id\": \"output_path_prmtop\",\n \"required\": true,\n \"description\": \"Path to the PRMTOP output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.ac.prmtop\",\n \"formats\": [\n \".*\\\\.prmtop$\"\n ]\n }\n ]\n },\n {\n \"id\": \"acpype_params_cns\",\n \"description\": \"Small molecule parameterization for CNS/XPLOR MD package.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the acpype_params_cns tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_chemistry/master/biobb_chemistry/test/data/config/config_acpype_params_cns.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/acpype/acpype.params.mol2\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.mdl$\",\n \".*\\\\.mol2$\"\n ]\n },\n {\n \"id\": \"output_path_par\",\n \"required\": true,\n \"description\": \"Path to the PAR output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.cns.par\",\n \"formats\": [\n \".*\\\\.par$\"\n ]\n },\n {\n \"id\": \"output_path_inp\",\n \"required\": true,\n \"description\": \"Path to the INP output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.cns.inp\",\n \"formats\": [\n \".*\\\\.inp$\"\n ]\n },\n {\n \"id\": \"output_path_top\",\n \"required\": true,\n \"description\": \"Path to the TOP output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.cns.top\",\n \"formats\": [\n \".*\\\\.top$\"\n ]\n }\n ]\n },\n {\n \"id\": \"acpype_params_gmx\",\n \"description\": \"Small molecule parameterization for GROMACS MD package.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the acpype_params_gmx tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_chemistry/master/biobb_chemistry/test/data/config/config_acpype_params_gmx.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/acpype/acpype.params.mol2\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.mdl$\",\n \".*\\\\.mol2$\"\n ]\n },\n {\n \"id\": \"output_path_gro\",\n \"required\": true,\n \"description\": \"Path to the GRO output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.gmx.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_path_itp\",\n \"required\": true,\n \"description\": \"Path to the ITP output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.gmx.itp\",\n \"formats\": [\n \".*\\\\.itp$\"\n ]\n },\n {\n \"id\": \"output_path_top\",\n \"required\": true,\n \"description\": \"Path to the TOP output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.gmx.top\",\n \"formats\": [\n \".*\\\\.top$\"\n ]\n }\n ]\n },\n {\n \"id\": \"acpype_params_gmx_opls\",\n \"description\": \"Small molecule parameterization for OPLS/AA MD package.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the acpype_params_gmx_opls tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_chemistry/master/biobb_chemistry/test/data/config/config_acpype_params_gmx_opls.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/acpype/acpype.params.mol2\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.mdl$\",\n \".*\\\\.mol2$\"\n ]\n },\n {\n \"id\": \"output_path_itp\",\n \"required\": true,\n \"description\": \"Path to the ITP output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.gmx.opls.itp\",\n \"formats\": [\n \".*\\\\.itp$\"\n ]\n },\n {\n \"id\": \"output_path_top\",\n \"required\": true,\n \"description\": \"Path to the TOP output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/acpype/ref_acpype.gmx.opls.top\",\n \"formats\": [\n \".*\\\\.top$\"\n ]\n }\n ]\n },\n {\n \"id\": \"babel_convert\",\n \"description\": \"Small molecule format conversion.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the babel_convert tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_chemistry/master/biobb_chemistry/test/data/config/config_babel_convert.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/babel/babel.smi\",\n \"formats\": [\n \".*\\\\.abinit$\",\n \".*\\\\.acesout$\",\n \".*\\\\.acr$\",\n \".*\\\\.adfout$\",\n \".*\\\\.alc$\",\n \".*\\\\.aoforce$\",\n \".*\\\\.arc$\",\n \".*\\\\.axsf$\",\n \".*\\\\.bgf$\",\n \".*\\\\.box$\",\n \".*\\\\.bs$\",\n \".*\\\\.c09out$\",\n \".*\\\\.c3d2$\",\n \".*\\\\.caccrt$\",\n \".*\\\\.can$\",\n \".*\\\\.car$\",\n \".*\\\\.castep$\",\n \".*\\\\.ccc$\",\n \".*\\\\.cdjson$\",\n \".*\\\\.cdx$\",\n \".*\\\\.cdxml$\",\n \".*\\\\.cif$\",\n \".*\\\\.ck$\",\n \".*\\\\.cml$\",\n \".*\\\\.cmlr$\",\n \".*\\\\.CONFIG$\",\n \".*\\\\.CONTCAR$\",\n \".*\\\\.CONTFF$\",\n \".*\\\\.crk2d$\",\n \".*\\\\.crk3d$\",\n \".*\\\\.ct$\",\n \".*\\\\.cub$\",\n \".*\\\\.cube$\",\n \".*\\\\.dallog$\",\n \".*\\\\.dalmol$\",\n \".*\\\\.dat$\",\n \".*\\\\.dmol$\",\n \".*\\\\.dx$\",\n \".*\\\\.ent$\",\n \".*\\\\.exyz$\",\n \".*\\\\.fa$\",\n \".*\\\\.fasta$\",\n \".*\\\\.fch$\",\n \".*\\\\.fchk$\",\n \".*\\\\.fck$\",\n \".*\\\\.feat$\",\n \".*\\\\.fhiaims$\",\n \".*\\\\.fract$\",\n \".*\\\\.fs$\",\n \".*\\\\.fsa$\",\n \".*\\\\.g03$\",\n \".*\\\\.g09$\",\n \".*\\\\.g92$\",\n \".*\\\\.g94$\",\n \".*\\\\.g98$\",\n \".*\\\\.gal$\",\n \".*\\\\.gam$\",\n \".*\\\\.gamess$\",\n \".*\\\\.gamin$\",\n \".*\\\\.gamout$\",\n \".*\\\\.got$\",\n \".*\\\\.gpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.gukin$\",\n \".*\\\\.gukout$\",\n \".*\\\\.gzmat$\",\n \".*\\\\.hin$\",\n \".*\\\\.HISTORY$\",\n \".*\\\\.inchi$\",\n \".*\\\\.inp$\",\n \".*\\\\.ins$\",\n \".*\\\\.jin$\",\n \".*\\\\.jout$\",\n \".*\\\\.log$\",\n \".*\\\\.lpmd$\",\n \".*\\\\.mcdl$\",\n \".*\\\\.mcif$\",\n \".*\\\\.MDFF$\",\n \".*\\\\.mdl$\",\n \".*\\\\.ml2$\",\n \".*\\\\.mmcif$\",\n \".*\\\\.mmd$\",\n \".*\\\\.mmod$\",\n \".*\\\\.mol$\",\n \".*\\\\.mol2$\",\n \".*\\\\.mold$\",\n \".*\\\\.molden$\",\n \".*\\\\.molf$\",\n \".*\\\\.moo$\",\n \".*\\\\.mop$\",\n \".*\\\\.mopcrt$\",\n \".*\\\\.mopin$\",\n \".*\\\\.mopout$\",\n \".*\\\\.mpc$\",\n \".*\\\\.mpo$\",\n \".*\\\\.mpqc$\",\n \".*\\\\.mrv$\",\n \".*\\\\.msi$\",\n \".*\\\\.nwo$\",\n \".*\\\\.orca$\",\n \".*\\\\.out$\",\n \".*\\\\.outmol$\",\n \".*\\\\.output$\",\n \".*\\\\.pc$\",\n \".*\\\\.pcjson$\",\n \".*\\\\.pcm$\",\n \".*\\\\.pdb$\",\n \".*\\\\.pdbqt$\",\n \".*\\\\.png$\",\n \".*\\\\.pos$\",\n \".*\\\\.POSCAR$\",\n \".*\\\\.POSFF$\",\n \".*\\\\.pqr$\",\n \".*\\\\.pqs$\",\n \".*\\\\.prep$\",\n \".*\\\\.pwscf$\",\n \".*\\\\.qcout$\",\n \".*\\\\.res$\",\n \".*\\\\.rsmi$\",\n \".*\\\\.rxn$\",\n \".*\\\\.sd$\",\n \".*\\\\.sdf$\",\n \".*\\\\.siesta$\",\n \".*\\\\.smi$\",\n \".*\\\\.smiles$\",\n \".*\\\\.smy$\",\n \".*\\\\.sy2$\",\n \".*\\\\.t41$\",\n \".*\\\\.tdd$\",\n \".*\\\\.text$\",\n \".*\\\\.therm$\",\n \".*\\\\.tmol$\",\n \".*\\\\.txt$\",\n \".*\\\\.txyz$\",\n \".*\\\\.unixyz$\",\n \".*\\\\.VASP$\",\n \".*\\\\.vmol$\",\n \".*\\\\.xml$\",\n \".*\\\\.xsf$\",\n \".*\\\\.xtc$\",\n \".*\\\\.xyz$\",\n \".*\\\\.yob$\"\n ]\n },\n {\n \"id\": \"output_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/babel/ref_babel.convert.mol2\",\n \"formats\": [\n \".*\\\\.acesin$\",\n \".*\\\\.adf$\",\n \".*\\\\.alc$\",\n \".*\\\\.ascii$\",\n \".*\\\\.bgf$\",\n \".*\\\\.box$\",\n \".*\\\\.bs$\",\n \".*\\\\.c3d1$\",\n \".*\\\\.c3d2$\",\n \".*\\\\.cac$\",\n \".*\\\\.caccrt$\",\n \".*\\\\.cache$\",\n \".*\\\\.cacint$\",\n \".*\\\\.can$\",\n \".*\\\\.cdjson$\",\n \".*\\\\.cdxml$\",\n \".*\\\\.cht$\",\n \".*\\\\.cif$\",\n \".*\\\\.ck$\",\n \".*\\\\.cml$\",\n \".*\\\\.cmlr$\",\n \".*\\\\.com$\",\n \".*\\\\.confabreport$\",\n \".*\\\\.CONFIG$\",\n \".*\\\\.CONTCAR$\",\n \".*\\\\.CONTFF$\",\n \".*\\\\.copy$\",\n \".*\\\\.crk2d$\",\n \".*\\\\.crk3d$\",\n \".*\\\\.csr$\",\n \".*\\\\.cssr$\",\n \".*\\\\.ct$\",\n \".*\\\\.cub$\",\n \".*\\\\.cube$\",\n \".*\\\\.dalmol$\",\n \".*\\\\.dmol$\",\n \".*\\\\.dx$\",\n \".*\\\\.ent$\",\n \".*\\\\.exyz$\",\n \".*\\\\.fa$\",\n \".*\\\\.fasta$\",\n \".*\\\\.feat$\",\n \".*\\\\.fh$\",\n \".*\\\\.fhiaims$\",\n \".*\\\\.fix$\",\n \".*\\\\.fps$\",\n \".*\\\\.fpt$\",\n \".*\\\\.fract$\",\n \".*\\\\.fs$\",\n \".*\\\\.fsa$\",\n \".*\\\\.gamin$\",\n \".*\\\\.gau$\",\n \".*\\\\.gjc$\",\n \".*\\\\.gjf$\",\n \".*\\\\.gpr$\",\n \".*\\\\.gr96$\",\n \".*\\\\.gro$\",\n \".*\\\\.gukin$\",\n \".*\\\\.gukout$\",\n \".*\\\\.gzmat$\",\n \".*\\\\.hin$\",\n \".*\\\\.inchi$\",\n \".*\\\\.inchikey$\",\n \".*\\\\.inp$\",\n \".*\\\\.jin$\",\n \".*\\\\.k$\",\n \".*\\\\.lmpdat$\",\n \".*\\\\.lpmd$\",\n \".*\\\\.mcdl$\",\n \".*\\\\.mcif$\",\n \".*\\\\.MDFF$\",\n \".*\\\\.mdl$\",\n \".*\\\\.ml2$\",\n \".*\\\\.mmcif$\",\n \".*\\\\.mmd$\",\n \".*\\\\.mmod$\",\n \".*\\\\.mna$\",\n \".*\\\\.mol$\",\n \".*\\\\.mol2$\",\n \".*\\\\.mold$\",\n \".*\\\\.molden$\",\n \".*\\\\.molf$\",\n \".*\\\\.molreport$\",\n \".*\\\\.mop$\",\n \".*\\\\.mopcrt$\",\n \".*\\\\.mopin$\",\n \".*\\\\.mp$\",\n \".*\\\\.mpc$\",\n \".*\\\\.mpd$\",\n \".*\\\\.mpqcin$\",\n \".*\\\\.mrv$\",\n \".*\\\\.msms$\",\n \".*\\\\.nul$\",\n \".*\\\\.nw$\",\n \".*\\\\.orcainp$\",\n \".*\\\\.outmol$\",\n \".*\\\\.paint$\",\n \".*\\\\.pcjson$\",\n \".*\\\\.pcm$\",\n \".*\\\\.pdb$\",\n \".*\\\\.pdbqt$\",\n \".*\\\\.png$\",\n \".*\\\\.pointcloud$\",\n \".*\\\\.POSCAR$\",\n \".*\\\\.POSFF$\",\n \".*\\\\.pov$\",\n \".*\\\\.pqr$\",\n \".*\\\\.pqs$\",\n \".*\\\\.qcin$\",\n \".*\\\\.report$\",\n \".*\\\\.rsmi$\",\n \".*\\\\.rxn$\",\n \".*\\\\.sd$\",\n \".*\\\\.sdf$\",\n \".*\\\\.smi$\",\n \".*\\\\.smiles$\",\n \".*\\\\.stl$\",\n \".*\\\\.svg$\",\n \".*\\\\.sy2$\",\n \".*\\\\.tdd$\",\n \".*\\\\.text$\",\n \".*\\\\.therm$\",\n \".*\\\\.tmol$\",\n \".*\\\\.txt$\",\n \".*\\\\.txyz$\",\n \".*\\\\.unixyz$\",\n \".*\\\\.VASP$\",\n \".*\\\\.vmol$\",\n \".*\\\\.xed$\",\n \".*\\\\.xyz$\",\n \".*\\\\.yob$\",\n \".*\\\\.zin$\"\n ]\n }\n ]\n },\n {\n \"id\": \"babel_add_hydrogens\",\n \"description\": \"Adds hydrogen atoms to small molecules.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the babel_add_hydrogens tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_chemistry/master/biobb_chemistry/test/data/config/config_babel_add_hydrogens.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/babel/babel.no.H.pdb\",\n \"formats\": [\n \".*\\\\.abinit$\",\n \".*\\\\.acesout$\",\n \".*\\\\.acr$\",\n \".*\\\\.adfout$\",\n \".*\\\\.alc$\",\n \".*\\\\.aoforce$\",\n \".*\\\\.arc$\",\n \".*\\\\.axsf$\",\n \".*\\\\.bgf$\",\n \".*\\\\.box$\",\n \".*\\\\.bs$\",\n \".*\\\\.c09out$\",\n \".*\\\\.c3d2$\",\n \".*\\\\.caccrt$\",\n \".*\\\\.can$\",\n \".*\\\\.car$\",\n \".*\\\\.castep$\",\n \".*\\\\.ccc$\",\n \".*\\\\.cdjson$\",\n \".*\\\\.cdx$\",\n \".*\\\\.cdxml$\",\n \".*\\\\.cif$\",\n \".*\\\\.ck$\",\n \".*\\\\.cml$\",\n \".*\\\\.cmlr$\",\n \".*\\\\.CONFIG$\",\n \".*\\\\.CONTCAR$\",\n \".*\\\\.CONTFF$\",\n \".*\\\\.crk2d$\",\n \".*\\\\.crk3d$\",\n \".*\\\\.ct$\",\n \".*\\\\.cub$\",\n \".*\\\\.cube$\",\n \".*\\\\.dallog$\",\n \".*\\\\.dalmol$\",\n \".*\\\\.dat$\",\n \".*\\\\.dmol$\",\n \".*\\\\.dx$\",\n \".*\\\\.ent$\",\n \".*\\\\.exyz$\",\n \".*\\\\.fa$\",\n \".*\\\\.fasta$\",\n \".*\\\\.fch$\",\n \".*\\\\.fchk$\",\n \".*\\\\.fck$\",\n \".*\\\\.feat$\",\n \".*\\\\.fhiaims$\",\n \".*\\\\.fract$\",\n \".*\\\\.fs$\",\n \".*\\\\.fsa$\",\n \".*\\\\.g03$\",\n \".*\\\\.g09$\",\n \".*\\\\.g92$\",\n \".*\\\\.g94$\",\n \".*\\\\.g98$\",\n \".*\\\\.gal$\",\n \".*\\\\.gam$\",\n \".*\\\\.gamess$\",\n \".*\\\\.gamin$\",\n \".*\\\\.gamout$\",\n \".*\\\\.got$\",\n \".*\\\\.gpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.gukin$\",\n \".*\\\\.gukout$\",\n \".*\\\\.gzmat$\",\n \".*\\\\.hin$\",\n \".*\\\\.HISTORY$\",\n \".*\\\\.inchi$\",\n \".*\\\\.inp$\",\n \".*\\\\.ins$\",\n \".*\\\\.jin$\",\n \".*\\\\.jout$\",\n \".*\\\\.log$\",\n \".*\\\\.lpmd$\",\n \".*\\\\.mcdl$\",\n \".*\\\\.mcif$\",\n \".*\\\\.MDFF$\",\n \".*\\\\.mdl$\",\n \".*\\\\.ml2$\",\n \".*\\\\.mmcif$\",\n \".*\\\\.mmd$\",\n \".*\\\\.mmod$\",\n \".*\\\\.mol$\",\n \".*\\\\.mol2$\",\n \".*\\\\.mold$\",\n \".*\\\\.molden$\",\n \".*\\\\.molf$\",\n \".*\\\\.moo$\",\n \".*\\\\.mop$\",\n \".*\\\\.mopcrt$\",\n \".*\\\\.mopin$\",\n \".*\\\\.mopout$\",\n \".*\\\\.mpc$\",\n \".*\\\\.mpo$\",\n \".*\\\\.mpqc$\",\n \".*\\\\.mrv$\",\n \".*\\\\.msi$\",\n \".*\\\\.nwo$\",\n \".*\\\\.orca$\",\n \".*\\\\.out$\",\n \".*\\\\.outmol$\",\n \".*\\\\.output$\",\n \".*\\\\.pc$\",\n \".*\\\\.pcjson$\",\n \".*\\\\.pcm$\",\n \".*\\\\.pdb$\",\n \".*\\\\.pdbqt$\",\n \".*\\\\.png$\",\n \".*\\\\.pos$\",\n \".*\\\\.POSCAR$\",\n \".*\\\\.POSFF$\",\n \".*\\\\.pqr$\",\n \".*\\\\.pqs$\",\n \".*\\\\.prep$\",\n \".*\\\\.pwscf$\",\n \".*\\\\.qcout$\",\n \".*\\\\.res$\",\n \".*\\\\.rsmi$\",\n \".*\\\\.rxn$\",\n \".*\\\\.sd$\",\n \".*\\\\.sdf$\",\n \".*\\\\.siesta$\",\n \".*\\\\.smi$\",\n \".*\\\\.smiles$\",\n \".*\\\\.smy$\",\n \".*\\\\.sy2$\",\n \".*\\\\.t41$\",\n \".*\\\\.tdd$\",\n \".*\\\\.text$\",\n \".*\\\\.therm$\",\n \".*\\\\.tmol$\",\n \".*\\\\.txt$\",\n \".*\\\\.txyz$\",\n \".*\\\\.unixyz$\",\n \".*\\\\.VASP$\",\n \".*\\\\.vmol$\",\n \".*\\\\.xml$\",\n \".*\\\\.xsf$\",\n \".*\\\\.xtc$\",\n \".*\\\\.xyz$\",\n \".*\\\\.yob$\"\n ]\n },\n {\n \"id\": \"output_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/babel/ref_babel.hydrogens.pdb\",\n \"formats\": [\n \".*\\\\.acesin$\",\n \".*\\\\.adf$\",\n \".*\\\\.alc$\",\n \".*\\\\.ascii$\",\n \".*\\\\.bgf$\",\n \".*\\\\.box$\",\n \".*\\\\.bs$\",\n \".*\\\\.c3d1$\",\n \".*\\\\.c3d2$\",\n \".*\\\\.cac$\",\n \".*\\\\.caccrt$\",\n \".*\\\\.cache$\",\n \".*\\\\.cacint$\",\n \".*\\\\.can$\",\n \".*\\\\.cdjson$\",\n \".*\\\\.cdxml$\",\n \".*\\\\.cht$\",\n \".*\\\\.cif$\",\n \".*\\\\.ck$\",\n \".*\\\\.cml$\",\n \".*\\\\.cmlr$\",\n \".*\\\\.com$\",\n \".*\\\\.confabreport$\",\n \".*\\\\.CONFIG$\",\n \".*\\\\.CONTCAR$\",\n \".*\\\\.CONTFF$\",\n \".*\\\\.copy$\",\n \".*\\\\.crk2d$\",\n \".*\\\\.crk3d$\",\n \".*\\\\.csr$\",\n \".*\\\\.cssr$\",\n \".*\\\\.ct$\",\n \".*\\\\.cub$\",\n \".*\\\\.cube$\",\n \".*\\\\.dalmol$\",\n \".*\\\\.dmol$\",\n \".*\\\\.dx$\",\n \".*\\\\.ent$\",\n \".*\\\\.exyz$\",\n \".*\\\\.fa$\",\n \".*\\\\.fasta$\",\n \".*\\\\.feat$\",\n \".*\\\\.fh$\",\n \".*\\\\.fhiaims$\",\n \".*\\\\.fix$\",\n \".*\\\\.fps$\",\n \".*\\\\.fpt$\",\n \".*\\\\.fract$\",\n \".*\\\\.fs$\",\n \".*\\\\.fsa$\",\n \".*\\\\.gamin$\",\n \".*\\\\.gau$\",\n \".*\\\\.gjc$\",\n \".*\\\\.gjf$\",\n \".*\\\\.gpr$\",\n \".*\\\\.gr96$\",\n \".*\\\\.gro$\",\n \".*\\\\.gukin$\",\n \".*\\\\.gukout$\",\n \".*\\\\.gzmat$\",\n \".*\\\\.hin$\",\n \".*\\\\.inchi$\",\n \".*\\\\.inchikey$\",\n \".*\\\\.inp$\",\n \".*\\\\.jin$\",\n \".*\\\\.k$\",\n \".*\\\\.lmpdat$\",\n \".*\\\\.lpmd$\",\n \".*\\\\.mcdl$\",\n \".*\\\\.mcif$\",\n \".*\\\\.MDFF$\",\n \".*\\\\.mdl$\",\n \".*\\\\.ml2$\",\n \".*\\\\.mmcif$\",\n \".*\\\\.mmd$\",\n \".*\\\\.mmod$\",\n \".*\\\\.mna$\",\n \".*\\\\.mol$\",\n \".*\\\\.mol2$\",\n \".*\\\\.mold$\",\n \".*\\\\.molden$\",\n \".*\\\\.molf$\",\n \".*\\\\.molreport$\",\n \".*\\\\.mop$\",\n \".*\\\\.mopcrt$\",\n \".*\\\\.mopin$\",\n \".*\\\\.mp$\",\n \".*\\\\.mpc$\",\n \".*\\\\.mpd$\",\n \".*\\\\.mpqcin$\",\n \".*\\\\.mrv$\",\n \".*\\\\.msms$\",\n \".*\\\\.nul$\",\n \".*\\\\.nw$\",\n \".*\\\\.orcainp$\",\n \".*\\\\.outmol$\",\n \".*\\\\.paint$\",\n \".*\\\\.pcjson$\",\n \".*\\\\.pcm$\",\n \".*\\\\.pdb$\",\n \".*\\\\.pdbqt$\",\n \".*\\\\.png$\",\n \".*\\\\.pointcloud$\",\n \".*\\\\.POSCAR$\",\n \".*\\\\.POSFF$\",\n \".*\\\\.pov$\",\n \".*\\\\.pqr$\",\n \".*\\\\.pqs$\",\n \".*\\\\.qcin$\",\n \".*\\\\.report$\",\n \".*\\\\.rsmi$\",\n \".*\\\\.rxn$\",\n \".*\\\\.sd$\",\n \".*\\\\.sdf$\",\n \".*\\\\.smi$\",\n \".*\\\\.smiles$\",\n \".*\\\\.stl$\",\n \".*\\\\.svg$\",\n \".*\\\\.sy2$\",\n \".*\\\\.tdd$\",\n \".*\\\\.text$\",\n \".*\\\\.therm$\",\n \".*\\\\.tmol$\",\n \".*\\\\.txt$\",\n \".*\\\\.txyz$\",\n \".*\\\\.unixyz$\",\n \".*\\\\.VASP$\",\n \".*\\\\.vmol$\",\n \".*\\\\.xed$\",\n \".*\\\\.xyz$\",\n \".*\\\\.yob$\",\n \".*\\\\.zin$\"\n ]\n }\n ]\n },\n {\n \"id\": \"babel_remove_hydrogens\",\n \"description\": \"Removes hydrogen atoms to small molecules.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the babel_remove_hydrogens tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_chemistry/master/biobb_chemistry/test/data/config/config_babel_remove_hydrogens.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/babel/babel.H.pdb\",\n \"formats\": [\n \".*\\\\.abinit$\",\n \".*\\\\.acesout$\",\n \".*\\\\.acr$\",\n \".*\\\\.adfout$\",\n \".*\\\\.alc$\",\n \".*\\\\.aoforce$\",\n \".*\\\\.arc$\",\n \".*\\\\.axsf$\",\n \".*\\\\.bgf$\",\n \".*\\\\.box$\",\n \".*\\\\.bs$\",\n \".*\\\\.c09out$\",\n \".*\\\\.c3d2$\",\n \".*\\\\.caccrt$\",\n \".*\\\\.can$\",\n \".*\\\\.car$\",\n \".*\\\\.castep$\",\n \".*\\\\.ccc$\",\n \".*\\\\.cdjson$\",\n \".*\\\\.cdx$\",\n \".*\\\\.cdxml$\",\n \".*\\\\.cif$\",\n \".*\\\\.ck$\",\n \".*\\\\.cml$\",\n \".*\\\\.cmlr$\",\n \".*\\\\.CONFIG$\",\n \".*\\\\.CONTCAR$\",\n \".*\\\\.CONTFF$\",\n \".*\\\\.crk2d$\",\n \".*\\\\.crk3d$\",\n \".*\\\\.ct$\",\n \".*\\\\.cub$\",\n \".*\\\\.cube$\",\n \".*\\\\.dallog$\",\n \".*\\\\.dalmol$\",\n \".*\\\\.dat$\",\n \".*\\\\.dmol$\",\n \".*\\\\.dx$\",\n \".*\\\\.ent$\",\n \".*\\\\.exyz$\",\n \".*\\\\.fa$\",\n \".*\\\\.fasta$\",\n \".*\\\\.fch$\",\n \".*\\\\.fchk$\",\n \".*\\\\.fck$\",\n \".*\\\\.feat$\",\n \".*\\\\.fhiaims$\",\n \".*\\\\.fract$\",\n \".*\\\\.fs$\",\n \".*\\\\.fsa$\",\n \".*\\\\.g03$\",\n \".*\\\\.g09$\",\n \".*\\\\.g92$\",\n \".*\\\\.g94$\",\n \".*\\\\.g98$\",\n \".*\\\\.gal$\",\n \".*\\\\.gam$\",\n \".*\\\\.gamess$\",\n \".*\\\\.gamin$\",\n \".*\\\\.gamout$\",\n \".*\\\\.got$\",\n \".*\\\\.gpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.gukin$\",\n \".*\\\\.gukout$\",\n \".*\\\\.gzmat$\",\n \".*\\\\.hin$\",\n \".*\\\\.HISTORY$\",\n \".*\\\\.inchi$\",\n \".*\\\\.inp$\",\n \".*\\\\.ins$\",\n \".*\\\\.jin$\",\n \".*\\\\.jout$\",\n \".*\\\\.log$\",\n \".*\\\\.lpmd$\",\n \".*\\\\.mcdl$\",\n \".*\\\\.mcif$\",\n \".*\\\\.MDFF$\",\n \".*\\\\.mdl$\",\n \".*\\\\.ml2$\",\n \".*\\\\.mmcif$\",\n \".*\\\\.mmd$\",\n \".*\\\\.mmod$\",\n \".*\\\\.mol$\",\n \".*\\\\.mol2$\",\n \".*\\\\.mold$\",\n \".*\\\\.molden$\",\n \".*\\\\.molf$\",\n \".*\\\\.moo$\",\n \".*\\\\.mop$\",\n \".*\\\\.mopcrt$\",\n \".*\\\\.mopin$\",\n \".*\\\\.mopout$\",\n \".*\\\\.mpc$\",\n \".*\\\\.mpo$\",\n \".*\\\\.mpqc$\",\n \".*\\\\.mrv$\",\n \".*\\\\.msi$\",\n \".*\\\\.nwo$\",\n \".*\\\\.orca$\",\n \".*\\\\.out$\",\n \".*\\\\.outmol$\",\n \".*\\\\.output$\",\n \".*\\\\.pc$\",\n \".*\\\\.pcjson$\",\n \".*\\\\.pcm$\",\n \".*\\\\.pdb$\",\n \".*\\\\.pdbqt$\",\n \".*\\\\.png$\",\n \".*\\\\.pos$\",\n \".*\\\\.POSCAR$\",\n \".*\\\\.POSFF$\",\n \".*\\\\.pqr$\",\n \".*\\\\.pqs$\",\n \".*\\\\.prep$\",\n \".*\\\\.pwscf$\",\n \".*\\\\.qcout$\",\n \".*\\\\.res$\",\n \".*\\\\.rsmi$\",\n \".*\\\\.rxn$\",\n \".*\\\\.sd$\",\n \".*\\\\.sdf$\",\n \".*\\\\.siesta$\",\n \".*\\\\.smi$\",\n \".*\\\\.smiles$\",\n \".*\\\\.smy$\",\n \".*\\\\.sy2$\",\n \".*\\\\.t41$\",\n \".*\\\\.tdd$\",\n \".*\\\\.text$\",\n \".*\\\\.therm$\",\n \".*\\\\.tmol$\",\n \".*\\\\.txt$\",\n \".*\\\\.txyz$\",\n \".*\\\\.unixyz$\",\n \".*\\\\.VASP$\",\n \".*\\\\.vmol$\",\n \".*\\\\.xml$\",\n \".*\\\\.xsf$\",\n \".*\\\\.xtc$\",\n \".*\\\\.xyz$\",\n \".*\\\\.yob$\"\n ]\n },\n {\n \"id\": \"output_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/babel/ref_babel.nohydrogens.pdb\",\n \"formats\": [\n \".*\\\\.acesin$\",\n \".*\\\\.adf$\",\n \".*\\\\.alc$\",\n \".*\\\\.ascii$\",\n \".*\\\\.bgf$\",\n \".*\\\\.box$\",\n \".*\\\\.bs$\",\n \".*\\\\.c3d1$\",\n \".*\\\\.c3d2$\",\n \".*\\\\.cac$\",\n \".*\\\\.caccrt$\",\n \".*\\\\.cache$\",\n \".*\\\\.cacint$\",\n \".*\\\\.can$\",\n \".*\\\\.cdjson$\",\n \".*\\\\.cdxml$\",\n \".*\\\\.cht$\",\n \".*\\\\.cif$\",\n \".*\\\\.ck$\",\n \".*\\\\.cml$\",\n \".*\\\\.cmlr$\",\n \".*\\\\.com$\",\n \".*\\\\.confabreport$\",\n \".*\\\\.CONFIG$\",\n \".*\\\\.CONTCAR$\",\n \".*\\\\.CONTFF$\",\n \".*\\\\.copy$\",\n \".*\\\\.crk2d$\",\n \".*\\\\.crk3d$\",\n \".*\\\\.csr$\",\n \".*\\\\.cssr$\",\n \".*\\\\.ct$\",\n \".*\\\\.cub$\",\n \".*\\\\.cube$\",\n \".*\\\\.dalmol$\",\n \".*\\\\.dmol$\",\n \".*\\\\.dx$\",\n \".*\\\\.ent$\",\n \".*\\\\.exyz$\",\n \".*\\\\.fa$\",\n \".*\\\\.fasta$\",\n \".*\\\\.feat$\",\n \".*\\\\.fh$\",\n \".*\\\\.fhiaims$\",\n \".*\\\\.fix$\",\n \".*\\\\.fps$\",\n \".*\\\\.fpt$\",\n \".*\\\\.fract$\",\n \".*\\\\.fs$\",\n \".*\\\\.fsa$\",\n \".*\\\\.gamin$\",\n \".*\\\\.gau$\",\n \".*\\\\.gjc$\",\n \".*\\\\.gjf$\",\n \".*\\\\.gpr$\",\n \".*\\\\.gr96$\",\n \".*\\\\.gro$\",\n \".*\\\\.gukin$\",\n \".*\\\\.gukout$\",\n \".*\\\\.gzmat$\",\n \".*\\\\.hin$\",\n \".*\\\\.inchi$\",\n \".*\\\\.inchikey$\",\n \".*\\\\.inp$\",\n \".*\\\\.jin$\",\n \".*\\\\.k$\",\n \".*\\\\.lmpdat$\",\n \".*\\\\.lpmd$\",\n \".*\\\\.mcdl$\",\n \".*\\\\.mcif$\",\n \".*\\\\.MDFF$\",\n \".*\\\\.mdl$\",\n \".*\\\\.ml2$\",\n \".*\\\\.mmcif$\",\n \".*\\\\.mmd$\",\n \".*\\\\.mmod$\",\n \".*\\\\.mna$\",\n \".*\\\\.mol$\",\n \".*\\\\.mol2$\",\n \".*\\\\.mold$\",\n \".*\\\\.molden$\",\n \".*\\\\.molf$\",\n \".*\\\\.molreport$\",\n \".*\\\\.mop$\",\n \".*\\\\.mopcrt$\",\n \".*\\\\.mopin$\",\n \".*\\\\.mp$\",\n \".*\\\\.mpc$\",\n \".*\\\\.mpd$\",\n \".*\\\\.mpqcin$\",\n \".*\\\\.mrv$\",\n \".*\\\\.msms$\",\n \".*\\\\.nul$\",\n \".*\\\\.nw$\",\n \".*\\\\.orcainp$\",\n \".*\\\\.outmol$\",\n \".*\\\\.paint$\",\n \".*\\\\.pcjson$\",\n \".*\\\\.pcm$\",\n \".*\\\\.pdb$\",\n \".*\\\\.pdbqt$\",\n \".*\\\\.png$\",\n \".*\\\\.pointcloud$\",\n \".*\\\\.POSCAR$\",\n \".*\\\\.POSFF$\",\n \".*\\\\.pov$\",\n \".*\\\\.pqr$\",\n \".*\\\\.pqs$\",\n \".*\\\\.qcin$\",\n \".*\\\\.report$\",\n \".*\\\\.rsmi$\",\n \".*\\\\.rxn$\",\n \".*\\\\.sd$\",\n \".*\\\\.sdf$\",\n \".*\\\\.smi$\",\n \".*\\\\.smiles$\",\n \".*\\\\.stl$\",\n \".*\\\\.svg$\",\n \".*\\\\.sy2$\",\n \".*\\\\.tdd$\",\n \".*\\\\.text$\",\n \".*\\\\.therm$\",\n \".*\\\\.tmol$\",\n \".*\\\\.txt$\",\n \".*\\\\.txyz$\",\n \".*\\\\.unixyz$\",\n \".*\\\\.VASP$\",\n \".*\\\\.vmol$\",\n \".*\\\\.xed$\",\n \".*\\\\.xyz$\",\n \".*\\\\.yob$\",\n \".*\\\\.zin$\"\n ]\n }\n ]\n },\n {\n \"id\": \"babel_minimize\",\n \"description\": \"Energetically minimize small molecules.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the babel_minimize tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_chemistry/master/biobb_chemistry/test/data/config/config_babel_minimize.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/babel/babel.minimize.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\"\n ]\n },\n {\n \"id\": \"output_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/babel/ref_babel.minimize.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\"\n ]\n }\n ]\n },\n {\n \"id\": \"reduce_add_hydrogens\",\n \"description\": \"Adds hydrogen atoms to small molecules.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the reduce_add_hydrogens tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_chemistry/master/biobb_chemistry/test/data/config/config_reduce_add_hydrogens.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/ambertools/reduce.no.H.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/ambertools/ref_reduce.add.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"reduce_remove_hydrogens\",\n \"description\": \"Removes hydrogen atoms to small molecules.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the reduce_remove_hydrogens tool.\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_path\",\n \"required\": true,\n \"description\": \"Path to the input file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/data/ambertools/reduce.H.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_chemistry/raw/master/biobb_chemistry/test/reference/ambertools/ref_reduce.remove.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n }\n ]\n },\n {\n \"id\": \"biobb_io\",\n \"tools\": [\n {\n \"id\": \"ligand\",\n \"description\": \"Downloads a ligand file from the MMB REST API.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the ligand tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_io/master/biobb_io/test/data/config/config_ligand.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"output_pdb_path\",\n \"required\": true,\n \"description\": \"Path to the output PDB ligand file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_io/raw/master/biobb_io/test/reference/api/ligand_12d.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"pdb\",\n \"description\": \"Downloads a PDB file from the RCSB or MMB REST APIs.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the pdb tool.\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"output_pdb_path\",\n \"required\": true,\n \"description\": \"Path to the output PDB file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_io/raw/master/biobb_io/test/reference/api/pdb_1ubq.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"pdb_variants\",\n \"description\": \"Creates a text file containing a list of all the variants mapped to a RSCB PDB code from the corresponding UNIPROT entries.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the pdb_variants tool.\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"output_mutations_list_txt\",\n \"required\": true,\n \"description\": \"Path to the TXT file containing an ASCII comma separated values of the mutations\",\n \"filetype\": \"output\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.txt$\"\n ]\n }\n ]\n },\n {\n \"id\": \"pdb_cluster_zip\",\n \"description\": \"Creates a zip file containing all the PDB files in the given sequence similarity cluster percentage of the given RSCB PDB code.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the pdb_cluster_zip tool.\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"output_pdb_zip_path\",\n \"required\": true,\n \"description\": \"Path to the ZIP or PDB file containing the output PDB files\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_io/raw/master/biobb_io/test/reference/api/reference_output_pdb_zip_path.zip\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.zip$\"\n ]\n }\n ]\n }\n ]\n },\n {\n \"id\": \"biobb_md\",\n \"tools\": [\n {\n \"id\": \"pdb2gmx\",\n \"description\": \"Creates a compressed (ZIP) GROMACS topology (TOP and ITP files) from a given PDB file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the pdb2gmx tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_md/master/biobb_md/test/data/config/config_pdb2gmx.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_pdb_path\",\n \"required\": true,\n \"description\": \"Path to the input PDB file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/egfr.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_gro_path\",\n \"required\": true,\n \"description\": \"Path to the output GRO file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_pdb2gmx.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the output TOP topology in zip format\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_pdb2gmx.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n }\n ]\n },\n {\n \"id\": \"editconf\",\n \"description\": \"Creates a GROMACS structure file (GRO) adding the information of the solvent box to the input structure file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the editconf tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_gro_path\",\n \"required\": true,\n \"description\": \"Path to the input GRO file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/editconf.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_gro_path\",\n \"required\": true,\n \"description\": \"Path to the output GRO file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_editconf.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n }\n ]\n },\n {\n \"id\": \"genion\",\n \"description\": \"Creates a new compressed GROMACS topology adding ions until reaching the desired concentration to the input compressed GROMACS topology. \",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the genion tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_tpr_path\",\n \"required\": true,\n \"description\": \"Path to the input portable run input TPR file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/genion.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\"\n ]\n },\n {\n \"id\": \"output_gro_path\",\n \"required\": true,\n \"description\": \"Path to the input structure GRO file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_genion.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"input_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the input TOP topology in zip format\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/genion.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"output_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the output topology TOP and ITP files zipball\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_genion.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n }\n ]\n },\n {\n \"id\": \"genrestr\",\n \"description\": \"Creates a new GROMACS compressed topology applying the indicated force restrains to the given input compressed topology.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the genrestr tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure PDB, GRO or TPR format\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/genrestr.gro\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\",\n \".*\\\\.tpr$\"\n ]\n },\n {\n \"id\": \"input_ndx_path\",\n \"required\": true,\n \"description\": \"Path to the input GROMACS index file, NDX format\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/genrestr.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_itp_path\",\n \"required\": true,\n \"description\": \"Path the output ITP topology file with restrains\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_genrestr.itp\",\n \"formats\": [\n \".*\\\\.itp$\"\n ]\n }\n ]\n },\n {\n \"id\": \"grompp\",\n \"description\": \"Creates a GROMACS portable binary run input file (TPR) applying the desired properties from the input compressed GROMACS topology.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the grompp tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_md/master/biobb_md/test/data/config/config_grompp.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_gro_path\",\n \"required\": true,\n \"description\": \"Path to the input GROMACS structure GRO file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/grompp.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"input_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the input GROMACS topology TOP and ITP files in zip format\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/grompp.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"output_tpr_path\",\n \"required\": true,\n \"description\": \"Path to the output portable binary run file TPR\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_grompp.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\"\n ]\n },\n {\n \"id\": \"input_cpt_path\",\n \"required\": false,\n \"description\": \"Path to the input GROMACS checkpoint file CPT\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.cpt$\"\n ]\n },\n {\n \"id\": \"input_ndx_path\",\n \"required\": false,\n \"description\": \"Path to the input GROMACS index files NDX\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n }\n ]\n },\n {\n \"id\": \"mdrun\",\n \"description\": \"Performs molecular dynamics simulations from an input GROMACS TPR file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the mdrun tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_tpr_path\",\n \"required\": true,\n \"description\": \"Path to the portable binary run input file TPR\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/mdrun.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\"\n ]\n },\n {\n \"id\": \"output_trr_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS uncompressed raw trajectory file TRR\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_mdrun.trr\",\n \"formats\": [\n \".*\\\\.trr$\"\n ]\n },\n {\n \"id\": \"output_gro_path\",\n \"required\": true,\n \"description\": \"Path to the output GROMACS structure GRO file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_mdrun.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_edr_path\",\n \"required\": true,\n \"description\": \"Path to the output GROMACS portable energy file EDR\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_mdrun.edr\",\n \"formats\": [\n \".*\\\\.edr$\"\n ]\n },\n {\n \"id\": \"output_log_path\",\n \"required\": true,\n \"description\": \"Path to the output GROMACS trajectory log file LOG\",\n \"filetype\": \"output\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.log$\"\n ]\n },\n {\n \"id\": \"output_xtc_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS compressed trajectory file XTC\",\n \"filetype\": \"output\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.xtc$\"\n ]\n },\n {\n \"id\": \"output_cpt_path\",\n \"required\": false,\n \"description\": \"Path to the output GROMACS checkpoint file CPT\",\n \"filetype\": \"output\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.cpt$\"\n ]\n },\n {\n \"id\": \"output_dhdl_path\",\n \"required\": false,\n \"description\": \"Path to the output dhdl\",\n \"filetype\": \"output\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.xvg$\"\n ]\n }\n ]\n },\n {\n \"id\": \"make_ndx\",\n \"description\": \"Creates a GROMACS index file (NDX) from an input selection and an input GROMACS structure file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the make_ndx tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input GRO/PDB/TPR file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/make_ndx.tpr\",\n \"formats\": [\n \".*\\\\.gro$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tpr$\"\n ]\n },\n {\n \"id\": \"output_ndx_path\",\n \"required\": true,\n \"description\": \"Path to the output index NDX file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_make_ndx.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"input_ndx_path\",\n \"required\": false,\n \"description\": \"Path to the input index NDX file\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n }\n ]\n },\n {\n \"id\": \"select\",\n \"description\": \"Creates a GROMACS index file (NDX) from an input selection and an input GROMACS structure file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the select tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_md/master/biobb_md/test/data/config/config_select.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input GRO/PDB/TPR file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/make_ndx.tpr\",\n \"formats\": [\n \".*\\\\.gro$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tpr$\"\n ]\n },\n {\n \"id\": \"output_ndx_path\",\n \"required\": true,\n \"description\": \"Path to the output index NDX file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_select.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"input_ndx_path\",\n \"required\": false,\n \"description\": \"Path to the input index NDX file\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n }\n ]\n },\n {\n \"id\": \"solvate\",\n \"description\": \"Creates a new compressed GROMACS topology file adding solvent molecules to a given input compressed GROMACS topology file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the solvate tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_solute_gro_path\",\n \"required\": true,\n \"description\": \"Path to the input GRO file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/solvate.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_gro_path\",\n \"required\": true,\n \"description\": \"Path to the output GRO file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_solvate.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"input_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the input TOP topology in zip format\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs/solvate.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"output_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the output topology in zip format\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs/ref_solvate.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n }\n ]\n },\n {\n \"id\": \"ndx2resttop\",\n \"description\": \"Creates a new GROMACS compressed topology applying the force restrains to the input groups in the input index file to the given input compressed topology.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the ndx2resttop tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_md/master/biobb_md/test/data/config/config_ndx2resttop.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_ndx_path\",\n \"required\": true,\n \"description\": \"Path to the input NDX index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs_extra/ndx2resttop.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"input_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the input TOP topology in zip format\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs_extra/ndx2resttop.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"output_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the output TOP topology in zip format\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs_extra/ref_ndx2resttop.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n }\n ]\n },\n {\n \"id\": \"append_ligand\",\n \"description\": \"Takes a ligand ITP file and inserts it in a topology.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the append_ligand tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the input topology TOP and ITP files zipball\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs_extra/ndx2resttop.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_itp_path\",\n \"required\": true,\n \"description\": \"Path to the ligand ITP file to be inserted in the topology\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/data/gromacs_extra/pep_ligand.itp\",\n \"formats\": [\n \".*\\\\.itp$\"\n ]\n },\n {\n \"id\": \"output_top_zip_path\",\n \"required\": true,\n \"description\": \"Path/Name the output topology TOP and ITP files zipball\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_md/raw/master/biobb_md/test/reference/gromacs_extra/ref_appendligand.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n }\n ]\n }\n ]\n },\n {\n \"id\": \"biobb_model\",\n \"tools\": [\n {\n \"id\": \"fix_side_chain\",\n \"description\": \"Reconstructs the missing side chains and heavy atoms of the given PDB file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the fix_side_chain tool.\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_pdb_path\",\n \"required\": true,\n \"description\": \"Input PDB file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_model/raw/master/biobb_model/test/data/model/2ki5.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_pdb_path\",\n \"required\": true,\n \"description\": \"Output PDB file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_model/raw/master/biobb_model/test/reference/model/output_pdb_path.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"mutate\",\n \"description\": \"Creates a new PDB file performing the mutations given in a list of amino acid mutations to the input PDB file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the mutate tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_model/master/biobb_model/test/data/config/config_mutate.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_pdb_path\",\n \"required\": true,\n \"description\": \"Input PDB file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_model/raw/master/biobb_model/test/data/model/2ki5.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_pdb_path\",\n \"required\": true,\n \"description\": \"Output PDB file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_model/raw/master/biobb_model/test/reference/model/output_mutated_pdb_path.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n }\n ]\n },\n {\n \"id\": \"biobb_pmx\",\n \"tools\": [\n {\n \"id\": \"mutate\",\n \"description\": \"pmx tool to insert mutated residues in structure files for free energy simulations\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the mutate tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_pmx/master/biobb_pmx/test/data/config/config_mutate.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_pmx/raw/master/biobb_pmx/test/data/pmx/frame99.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_structure_path\",\n \"required\": true,\n \"description\": \"Path to the output structure file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_pmx/raw/master/biobb_pmx/test/reference/pmx/ref_output_structure.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"input_b_structure_path\",\n \"required\": false,\n \"description\": \"Path to the mutated input structure file\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gentop\",\n \"description\": \"pmx tool to generate hybrid GROMACS topologies: adding a B state to an .itp or .top file for a hybrid residue\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gentop tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_pmx/master/biobb_pmx/test/data/config/config_gentop.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the input GROMACS topology TOP and ITP files in zip format\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_pmx/raw/master/biobb_pmx/test/data/pmx/topology.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"output_top_zip_path\",\n \"required\": true,\n \"description\": \"Path the output TOP topology in zip format\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_pmx/raw/master/biobb_pmx/test/reference/pmx/ref_output_topology.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n }\n ]\n },\n {\n \"id\": \"analyse\",\n \"description\": \"pmx tool to calculate free energies from fast growth thermodynamic integration simulations.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the analyse tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_pmx/master/biobb_pmx/test/data/config/config_analyse.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_a_xvg_zip_path\",\n \"required\": true,\n \"description\": \"Path the zip file containing the dgdl\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_pmx/raw/master/biobb_pmx/test/data/pmx/xvg_A.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_b_xvg_zip_path\",\n \"required\": true,\n \"description\": \"Path the zip file containing the dgdl\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_pmx/raw/master/biobb_pmx/test/data/pmx/xvg_B.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"output_result_path\",\n \"required\": true,\n \"description\": \"Path to the TXT results file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_pmx/raw/master/biobb_pmx/test/reference/pmx/ref_result.txt\",\n \"formats\": [\n \".*\\\\.txt$\"\n ]\n },\n {\n \"id\": \"output_work_plot_path\",\n \"required\": true,\n \"description\": \"Path to the PNG plot results file\",\n \"filetype\": \"output\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.png$\"\n ]\n }\n ]\n }\n ]\n },\n {\n \"id\": \"biobb_structure_utils\",\n \"tools\": [\n {\n \"id\": \"cat_pdb\",\n \"description\": \"Class to concat two PDB structures in a single PDB file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cat_pdb tool.\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure1\",\n \"required\": true,\n \"description\": \"Input structure 1 file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/cat_protein.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"input_structure2\",\n \"required\": true,\n \"description\": \"Input structure 2 file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/cat_ligand.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_structure_path\",\n \"required\": true,\n \"description\": \"Output protein file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/ref_cat_pdb.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"extract_atoms\",\n \"description\": \"Class to extract atoms from a 3D structure.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the extract_atoms tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_structure_utils/master/biobb_structure_utils/test/data/config/config_extract_atoms.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Input structure file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/2vgb.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_structure_path\",\n \"required\": true,\n \"description\": \"Output structure file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/OE2_atoms.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\"\n ]\n }\n ]\n },\n {\n \"id\": \"extract_chain\",\n \"description\": \"Class to extract a chain from a 3D structure.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the extract_chain tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_structure_utils/master/biobb_structure_utils/test/data/config/config_extract_chain.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Input structure file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/extract_chain.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_structure_path\",\n \"required\": true,\n \"description\": \"Output structure file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/ref_extract_chain.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"extract_heteroatoms\",\n \"description\": \"Class to extract hetero-atoms from a 3D structure.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the extract_heteroatoms tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_structure_utils/master/biobb_structure_utils/test/data/config/config_extract_heteroatoms.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Input structure file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/extract_heteroatom.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_heteroatom_path\",\n \"required\": true,\n \"description\": \"Output heteroatom file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/ref_extract_heteroatom.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"extract_model\",\n \"description\": \"Class to extract a model from a 3D structure.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the extract_model tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_structure_utils/master/biobb_structure_utils/test/data/config/config_extract_model.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Input structure file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/extract_model.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_structure_path\",\n \"required\": true,\n \"description\": \"Output structure file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/ref_extract_model.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"extract_protein\",\n \"description\": \"Class to extract a protein from a 3D structure.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the extract_protein tool.\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Input structure file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/extract_protein.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_protein_path\",\n \"required\": true,\n \"description\": \"Output protein file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/ref_extract_protein.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"remove_ligand\",\n \"description\": \"Class to remove the selected ligand atoms from a 3D structure.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the remove_ligand tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_structure_utils/master/biobb_structure_utils/test/data/config/config_remove_ligand.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Input structure file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/WT_aq4_md_1.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_structure_path\",\n \"required\": true,\n \"description\": \"Output structure file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/WT_apo_md_1.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\"\n ]\n }\n ]\n },\n {\n \"id\": \"remove_pdb_water\",\n \"description\": \"Class to remove water molecules from PDB 3D structures.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the remove_pdb_water tool.\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_pdb_path\",\n \"required\": true,\n \"description\": \"Input PDB file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/WT_aq4_md_WAT.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n },\n {\n \"id\": \"output_pdb_path\",\n \"required\": true,\n \"description\": \"Output PDB file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/WT_apo_no_wat.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\"\n ]\n }\n ]\n },\n {\n \"id\": \"renumber_structure\",\n \"description\": \"Class to renumber atomic indexes from a 3D structure.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the renumber_structure tool.\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Input structure file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/cl3.noH.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_structure_path\",\n \"required\": true,\n \"description\": \"Output structure file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/renum_cl3_noH.pdb\",\n \"formats\": [\n \".*\\\\.pdb$\",\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_mapping_json_path\",\n \"required\": true,\n \"description\": \"Output mapping json file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/cl3_output_mapping_json_path.json\",\n \"formats\": [\n \".*\\\\.json$\"\n ]\n }\n ]\n },\n {\n \"id\": \"sort_gro_residues\",\n \"description\": \"Class to sort the selected residues from a GRO 3D structure.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the sort_gro_residues tool.\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_structure_utils/master/biobb_structure_utils/test/data/config/config_sort_gro_residues.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_gro_path\",\n \"required\": true,\n \"description\": \"Input GRO file path\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/data/utils/WT_aq4_md_1.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n },\n {\n \"id\": \"output_gro_path\",\n \"required\": true,\n \"description\": \"Output sorted GRO file path\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_structure_utils/raw/master/biobb_structure_utils/test/reference/utils/WT_aq4_md_sorted.gro\",\n \"formats\": [\n \".*\\\\.gro$\"\n ]\n }\n ]\n }\n ]\n }\n ]\n}\n" ] ], [ [ "<a id=\"list_tools_ex\"></a>\n#### List of tools from a specific package\n\nFor more information about this endpoint, please visit the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/List%20of%20Services/getToolsList).\n\n##### Endpoint", "_____no_output_____" ], [ "**GET** `https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch/{package}`", "_____no_output_____" ], [ "##### Code", "_____no_output_____" ] ], [ [ "package = 'biobb_analysis'\nurl = apiURL + 'launch/' + package\nresponse = get_data(url)\n\nprint(json.dumps(response.json, indent=2))", "{\n \"id\": \"biobb_analysis\",\n \"tools\": [\n {\n \"id\": \"gmx_cluster\",\n \"description\": \"Creates cluster structures from a given GROMACS compatible trajectory\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_cluster tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_cluster.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_pdb_path\",\n \"required\": true,\n \"description\": \"Path to the output cluster file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_cluster.pdb\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_rms\",\n \"description\": \"Performs a Root Mean Square deviation (RMSd) analysis from a given GROMACS compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_rms tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_rms.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_xvg_path\",\n \"required\": true,\n \"description\": \"Path to the XVG output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_rms.xvg\",\n \"formats\": [\n \".*\\\\.xvg$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_rgyr\",\n \"description\": \"Computes the radius of gyration (Rgyr) of a molecule about the x-, y- and z-axes, as a function of time, from a given GROMACS compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_rgyr tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_rgyr.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_xvg_path\",\n \"required\": true,\n \"description\": \"Path to the XVG output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_rgyr.xvg\",\n \"formats\": [\n \".*\\\\.xvg$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_energy\",\n \"description\": \"Extracts energy components from a given GROMACS energy file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_energy tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_energy.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_energy_path\",\n \"required\": true,\n \"description\": \"Path to the input EDR file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/energy.edr\",\n \"formats\": [\n \".*\\\\.edr$\"\n ]\n },\n {\n \"id\": \"output_xvg_path\",\n \"required\": true,\n \"description\": \"Path to the XVG output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_energy.xvg\",\n \"formats\": [\n \".*\\\\.xvg$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_image\",\n \"description\": \"Corrects periodicity (image) from a given GROMACS compatible trajectory file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_image tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_image.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS input topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_traj_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_image.xtc\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_trjconv_str\",\n \"description\": \"Converts between GROMACS compatible structure file formats and/or extracts a selection of atoms.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_trjconv_str tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_trjconv_str.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_structure_path\",\n \"required\": true,\n \"description\": \"Path to the input structure file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS input topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_str_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_trjconv.str.pdb\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_trjconv_str_ens\",\n \"description\": \"Extracts an ensemble of frames containing a selection of atoms from GROMACS compatible trajectory files.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_trjconv_str_ens tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_trjconv_str_ens.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS input topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/topology.tpr\",\n \"formats\": [\n \".*\\\\.tpr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.brk$\",\n \".*\\\\.ent$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_str_ens_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_trjconv.str.ens.zip\",\n \"formats\": [\n \".*\\\\.zip$\"\n ]\n }\n ]\n },\n {\n \"id\": \"gmx_trjconv_trj\",\n \"description\": \"Converts between GROMACS compatible trajectory file formats and/or extracts a selection of atoms.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the gmx_trjconv_trj tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_gmx_trjconv_trj.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the GROMACS trajectory file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/trajectory.trr\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.cpt$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n },\n {\n \"id\": \"input_index_path\",\n \"required\": false,\n \"description\": \"Path to the GROMACS index file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/gromacs/index.ndx\",\n \"formats\": [\n \".*\\\\.ndx$\"\n ]\n },\n {\n \"id\": \"output_traj_path\",\n \"required\": true,\n \"description\": \"Path to the output file\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/gromacs/ref_trjconv.trj.xtc\",\n \"formats\": [\n \".*\\\\.xtc$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.g96$\",\n \".*\\\\.pdb$\",\n \".*\\\\.tng$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_average\",\n \"description\": \"Calculates a structure average of a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_average tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_average.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed structure\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.average.pdb\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_bfactor\",\n \"description\": \"Calculates the Bfactor fluctuations of a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_bfactor tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"input_exp_path\",\n \"required\": false,\n \"description\": \"Path to the experimental reference file (required if reference = experimental)\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/experimental.1e5t.pdb\",\n \"formats\": null\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed analysis\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.bfactor.first.dat\",\n \"formats\": [\n \".*\\\\.dat$\",\n \".*\\\\.agr$\",\n \".*\\\\.xmgr$\",\n \".*\\\\.gnu$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_rms\",\n \"description\": \"Calculates the Root Mean Square deviation (RMSd) of a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_rms tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"input_exp_path\",\n \"required\": false,\n \"description\": \"Path to the experimental reference file (required if reference = experimental)\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/experimental.1e5t.pdb\",\n \"formats\": null\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed analysis\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.rms.first.dat\",\n \"formats\": [\n \".*\\\\.dat$\",\n \".*\\\\.agr$\",\n \".*\\\\.xmgr$\",\n \".*\\\\.gnu$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_rmsf\",\n \"description\": \"Calculates the Root Mean Square fluctuations (RMSf) of a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_rmsf tool\",\n \"filetype\": \"input\",\n \"sample\": null,\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"input_exp_path\",\n \"required\": false,\n \"description\": \"Path to the experimental reference file (required if reference = experimental)\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/experimental.1e5t.pdb\",\n \"formats\": null\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed analysis\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.rmsf.first.dat\",\n \"formats\": [\n \".*\\\\.dat$\",\n \".*\\\\.agr$\",\n \".*\\\\.xmgr$\",\n \".*\\\\.gnu$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_rgyr\",\n \"description\": \"Computes the radius of gyration (Rgyr) from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_rgyr tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_rgyr.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output analysis\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.rgyr.dat\",\n \"formats\": [\n \".*\\\\.dat$\",\n \".*\\\\.agr$\",\n \".*\\\\.xmgr$\",\n \".*\\\\.gnu$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_dry\",\n \"description\": \"Dehydrates a given cpptraj compatible trajectory stripping out solvent molecules and ions.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_dry tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_dry.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.dry.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_strip\",\n \"description\": \"Strips a defined set of atoms (mask) from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_strip tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_strip.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.strip.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_snapshot\",\n \"description\": \"Extracts a particular snapshot from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_snapshot tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_snapshot.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed structure\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.snapshot.pdb\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_slice\",\n \"description\": \"Extracts a particular trajectory slice from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_slice tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_slice.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.slice.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_convert\",\n \"description\": \"Converts between cpptraj compatible trajectory file formats and/or extracts a selection of atoms or frames.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_convert tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_convert.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.convert.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_mask\",\n \"description\": \"Extracts a selection of atoms from a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_mask tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_mask.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.mask.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n },\n {\n \"id\": \"cpptraj_image\",\n \"description\": \"Corrects periodicity (image) from a given cpptraj trajectory file.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_image tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_image.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed trajectory\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.image.netcdf\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n }\n ]\n}\n" ] ], [ [ "<a id=\"tools_prop_ex\"></a>\n#### Tool's properties\n\nFor more information about this endpoint, please visit the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/Launch%20Tool/getLaunchTool).\n\n##### Endpoint", "_____no_output_____" ], [ "**GET** `https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch/{package}/{tool}`", "_____no_output_____" ], [ "##### Code", "_____no_output_____" ] ], [ [ "package = 'biobb_analysis'\ntool = 'cpptraj_average'\nurl = apiURL + 'launch/' + package + '/' + tool\nresponse = get_data(url)\n\nprint(json.dumps(response.json, indent=2))", "{\n \"id\": \"cpptraj_average\",\n \"description\": \"Calculates a structure average of a given cpptraj compatible trajectory.\",\n \"arguments\": [\n {\n \"id\": \"config\",\n \"required\": false,\n \"description\": \"Configuration file for the cpptraj_average tool\",\n \"filetype\": \"input\",\n \"sample\": \"https://raw.githubusercontent.com/bioexcel/biobb_analysis/master/biobb_analysis/test/data/config/config_cpptraj_average.json\",\n \"formats\": [\n \".*\\\\.json$\",\n \".*\\\\.yml$\"\n ]\n },\n {\n \"id\": \"input_top_path\",\n \"required\": true,\n \"description\": \"Path to the input structure or topology file\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.parm.top\",\n \"formats\": [\n \".*\\\\.top$\",\n \".*\\\\.pdb$\",\n \".*\\\\.prmtop$\",\n \".*\\\\.parmtop$\",\n \".*\\\\.zip$\"\n ]\n },\n {\n \"id\": \"input_traj_path\",\n \"required\": true,\n \"description\": \"Path to the input trajectory to be processed\",\n \"filetype\": \"input\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/data/ambertools/cpptraj.traj.dcd\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.cdf$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.restart$\",\n \".*\\\\.ncrestart$\",\n \".*\\\\.restartnc$\",\n \".*\\\\.dcd$\",\n \".*\\\\.charmm$\",\n \".*\\\\.cor$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.trr$\",\n \".*\\\\.gro$\",\n \".*\\\\.binpos$\",\n \".*\\\\.xtc$\",\n \".*\\\\.cif$\",\n \".*\\\\.arc$\",\n \".*\\\\.sqm$\",\n \".*\\\\.sdf$\",\n \".*\\\\.conflib$\"\n ]\n },\n {\n \"id\": \"output_cpptraj_path\",\n \"required\": true,\n \"description\": \"Path to the output processed structure\",\n \"filetype\": \"output\",\n \"sample\": \"https://github.com/bioexcel/biobb_analysis/raw/master/biobb_analysis/test/reference/ambertools/ref_cpptraj.average.pdb\",\n \"formats\": [\n \".*\\\\.crd$\",\n \".*\\\\.netcdf$\",\n \".*\\\\.rst7$\",\n \".*\\\\.ncrst$\",\n \".*\\\\.dcd$\",\n \".*\\\\.pdb$\",\n \".*\\\\.mol2$\",\n \".*\\\\.binpos$\",\n \".*\\\\.trr$\",\n \".*\\\\.xtc$\",\n \".*\\\\.sqm$\"\n ]\n }\n ]\n}\n" ] ], [ [ "<a id=\"launch_tool_ex\"></a>\n### Launch tool\n\nFor more information about this endpoint, please visit the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/Launch%20Tool/postLaunchTool). The documentation for all the tools is available in the [BioBB REST API Tools Documentation section](https://mmb.irbbarcelona.org/biobb-api/tools-documentation?docExpansion=none). Interactive examples for all the tools are available in the [BioBB REST API Tools Execution section](https://mmb.irbbarcelona.org/biobb-api/tools-execution).\n\nDefinition of functions needed for launch a job:", "_____no_output_____" ] ], [ [ "from io import BytesIO\nfrom pathlib import Path\n\n# Function used for encode python dictionary to JSON file\ndef encode_config(data):\n jsonData = json.dumps(data)\n binaryData = jsonData.encode()\n return BytesIO(binaryData)\n\n# Launch job\ndef launch_job(url, **kwargs):\n data = {}\n files = {}\n # Fill data (output paths) and files (input files) objects\n for key, value in kwargs.items():\n # Inputs / Outputs\n if type(value) is str:\n if key.startswith('input'):\n files[key] = (value, open(value, 'rb'))\n elif key.startswith('output'):\n data[key] = value\n elif Path(value).is_file():\n files[key] = (value, open(value, 'rb'))\n # Properties (in case properties are provided as a dictionary instead of a file)\n if type(value) is dict:\n files['config'] = ('prop.json', encode_config(value))\n # Request URL with data and files\n response = post_data(url, data, files)\n # Print REST API response\n print(json.dumps(response.json, indent=2))\n # Save token if status == 303\n if response.status == 303:\n token = response.json['token']\n return token", "_____no_output_____" ] ], [ [ "Hereafter we will launch a job on *biobb_analysis.cpptraj_average* tool with the provided *files/* in the files folder of this same repository. The response is a JSON with the status code, the state of the job, a message and a token for checking the job status.\n\n<a id=\"tool_yml_ex\"></a>\n#### Launch job with a YAML file config\n\n##### File config", "_____no_output_____" ], [ "```yaml \nproperties:\n in_parameters:\n start: 1\n end: -1\n step: 1\n mask: c-alpha\n out_parameters:\n format: pdb\n```", "_____no_output_____" ], [ "##### Endpoint", "_____no_output_____" ], [ "**POST** `https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch/{package}/{tool}`", "_____no_output_____" ], [ "##### Code", "_____no_output_____" ], [ "The function below sends POST data and files to the *{package}/{tool}* endpoint. The config properties are sent as a YAML file.\n\nThe response is a JSON with the status code, the state of the job, a message and a token that will be used for checking the job status in the next step. ", "_____no_output_____" ] ], [ [ "# Launch BioBB on REST API with YAML config file\n\ntoken = launch_job(url = apiURL + 'launch/biobb_analysis/cpptraj_average', \n config = 'files/config.yml',\n input_top_path = 'files/cpptraj.parm.top',\n input_traj_path = 'files/cpptraj.traj.dcd',\n output_cpptraj_path = 'output.cpptraj.average.pdb')", "{\n \"code\": 303,\n \"state\": \"RUNNING\",\n \"message\": \"The requested job has has been successfully launched, please go to /retrieve/status/{token} for checking job status.\",\n \"token\": \"fe2805760eeeec0d5b8a34fbc40aa6c2a2d68c7ba1663cccb88659b1e149c898a414bbc04e37bb73efc725b7a29de2a93ffb55e6ef85cd6467f3d62a06ea5bfa\"\n}\n" ] ], [ [ "<a id=\"tool_json_ex\"></a>\n#### Launch job with a JSON file config\n\nFile config:", "_____no_output_____" ], [ "```json\n{\n\t\"in_parameters\": {\n\t\t\"start\": 1,\n\t\t\"end\": -1,\n\t\t\"step\": 1,\n\t\t\"mask\": \"c-alpha\"\n\t},\n\t\"out_parameters\": {\n\t\t\"format\": \"pdb\"\n\t}\n}\n```", "_____no_output_____" ], [ "##### Endpoint", "_____no_output_____" ], [ "**POST** `https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch/{package}/{tool}`", "_____no_output_____" ], [ "##### Code", "_____no_output_____" ], [ "The function below sends POST data and files to the *{package}/{tool}* endpoint. The config properties are sent as a JSON file.\n\nThe response is a JSON with the status code, the state of the job, a message and a token that will be used for checking the job status in the next step. ", "_____no_output_____" ] ], [ [ "# Launch BioBB on REST API with JSON config file\n\ntoken = launch_job(url = apiURL + 'launch/biobb_analysis/cpptraj_average', \n config = 'files/config.json',\n input_top_path = 'files/cpptraj.parm.top',\n input_traj_path = 'files/cpptraj.traj.dcd',\n output_cpptraj_path = 'output.cpptraj.average.pdb')", "{\n \"code\": 303,\n \"state\": \"RUNNING\",\n \"message\": \"The requested job has has been successfully launched, please go to /retrieve/status/{token} for checking job status.\",\n \"token\": \"84ab5ef63d82ab3fa4f120532949905d83f6aff65f101cb1ed5fdd5f05acb00421ddc4560098f877f26a96972a8ea8521ab222a0bb78a5ffa9d213c0ab2618c9\"\n}\n" ] ], [ [ "<a id=\"tool_dict_ex\"></a>\n#### Launch job with a python dictionary config", "_____no_output_____" ], [ "##### Endpoint", "_____no_output_____" ], [ "**POST** `https://mmb.irbbarcelona.org/biobb-api/rest/v1/launch/{package}/{tool}`", "_____no_output_____" ], [ "##### Code", "_____no_output_____" ], [ "The function below sends POST data and files to the *{package}/{tool}* endpoint. The config properties are sent as a python dictionary embedded in the code.\n\nThe response is a JSON with the status code, the state of the job, a message and a token that will be used for checking the job status in the next step. ", "_____no_output_____" ] ], [ [ "# Launch BioBB on REST API with JSON config file\n\nprop = {\n \"in_parameters\" : {\n \"start\": 1,\n \"end\": -1,\n \"step\": 1,\n \"mask\": \"c-alpha\"\n },\n \"out_parameters\" : {\n \"format\": \"pdb\"\n }\n}\n\ntoken = launch_job(url = apiURL + 'launch/biobb_analysis/cpptraj_average', \n config = prop,\n input_top_path = 'files/cpptraj.parm.top',\n input_traj_path = 'files/cpptraj.traj.dcd',\n output_cpptraj_path = 'output.cpptraj.average.pdb')", "{\n \"code\": 303,\n \"state\": \"RUNNING\",\n \"message\": \"The requested job has has been successfully launched, please go to /retrieve/status/{token} for checking job status.\",\n \"token\": \"98013d74bef397d5498db3eb1008e5e136702d63903b6ea0cb5a2db44c4a4e0adbcd1ce9999915acd90c444f8749880c052185bbbfc747c1ebc7d67d6d2c84c8\"\n}\n" ] ], [ [ "<a id=\"retrieve_status_ex\"></a>\n### Retrieve status\n\nFor more information about this endpoint, please visit the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/Retrieve/getRetrieveStatus).\n\nDefinition of functions needed for retrieve the status of a job:", "_____no_output_____" ] ], [ [ "import datetime\nfrom time import sleep\n\n# Checks status until a provided \"ok\" status is returned by the response\ndef check_status(url, ok, error):\n counter = 0\n while True:\n if counter < 10: slp = 1\n if counter >= 10 and counter < 60: slp = 10\n if counter >= 60: slp = 60\n counter = counter + slp\n sleep(slp)\n r = requests.get(url)\n if r.status_code == ok or r.status_code == error:\n return counter\n break\n\n# Function that checks the status and parses the reponse JSON for saving the output files in a list\ndef check_job(token, apiURL):\n # define retrieve status URL\n url = apiURL + 'retrieve/status/' + token\n # check status until job has finished\n counter = check_status(url, 200, 500)\n # Get content when status = 200\n response = get_data(url)\n # Save id for the generated output_files\n if response.status == 200:\n out_files = []\n for outf in response.json['output_files']:\n item = { 'id': outf['id'], 'name': outf['name'] }\n out_files.append(item)\n\n # Print REST API response\n print(\"Total elapsed time: %s\" % str(datetime.timedelta(seconds=counter)))\n print(\"REST API JSON response:\")\n print(json.dumps(response.json, indent=4))\n \n if response.status == 200: \n return out_files\n else: return None", "_____no_output_____" ] ], [ [ "##### Endpoint", "_____no_output_____" ], [ "**GET** `https://mmb.irbbarcelona.org/biobb-api/rest/v1/retrieve/status/{token}`", "_____no_output_____" ], [ "##### Code", "_____no_output_____" ], [ "The function below checks the status of a job and awaits until the response status is `200`. The response is a JSON with the status code, the state of the job, a message, a list with all the generated output files and the date of the expiration of these files. Additionally, the function also provides the elapsed time since the job has been launched until it has finished. ", "_____no_output_____" ] ], [ [ "# Check job status\nout_files = check_job(token, apiURL)", "Total elapsed time: 0:00:20\nREST API JSON response:\n{\n \"code\": 200,\n \"state\": \"FINISHED\",\n \"message\": \"The requested job has finished successfully, please go to /retrieve/data/{id} for each output_files.\",\n \"output_files\": [\n {\n \"id\": \"5e42837a40fe75.05757111\",\n \"name\": \"output.cpptraj.average.pdb\",\n \"size\": 77397,\n \"mimetype\": \"text/plain\"\n }\n ],\n \"expiration\": \"February 13, 2020 00:00 GMT+0000\"\n}\n" ] ], [ [ "<a id=\"retrieve_data_ex\"></a>\n### Retrieve data\n\nFor more information about this endpoint, please visit the [BioBB REST API Documentation section](https://mmb.irbbarcelona.org/biobb-api/rest#/Retrieve/getRetrieveData).\n\nDefinition of functions needed for retrieve the output file(s) generated by a job:", "_____no_output_____" ] ], [ [ "# Downloads to disk a file from a given URL\ndef get_file(url, filename):\n r = requests.get(url, allow_redirects=True)\n file = open(filename,'wb') \n file.write(r.content) \n file.close()\n\n# Retrieves all the files provided in the out_files list\ndef retrieve_data(out_files, apiURL):\n if not out_files:\n return \"No files provided\"\n for outf in out_files:\n get_file(apiURL + 'retrieve/data/' + outf['id'], outf['name'])", "_____no_output_____" ] ], [ [ "##### Endpoint", "_____no_output_____" ], [ "**GET** `https://mmb.irbbarcelona.org/biobb-api/rest/v1/retrieve/data/{id}`", "_____no_output_____" ], [ "##### Code", "_____no_output_____" ], [ "The function below makes a single call to the *retrieve/data* endpoint for each output file got in the *retrieve/status* endpoint and save the generated file(s) to disk.", "_____no_output_____" ] ], [ [ "# Save generated file(s) to disk\n\nretrieve_data(out_files, apiURL)", "_____no_output_____" ] ], [ [ "<a id=\"practical_cases\"></a>\n## Practical cases", "_____no_output_____" ], [ "Now we will execute some Bioexcel Building Blocks through the BioBB REST API and with the results we will do some interactions with other python libraries such as [plotly](https://plot.ly/python/offline/) or [nglview](http://nglviewer.org/#nglview).", "_____no_output_____" ], [ "<a id=\"example1\"></a>\n### Example 1: download PDB file from RSCB database", "_____no_output_____" ], [ "Launch the *biobb_io.pdb* job that downloads a PDB file from the RSCB database:", "_____no_output_____" ] ], [ [ "# Downloading desired PDB file\n\n# Create properties dict and inputs/outputs\ndownloaded_pdb = '3EBP.pdb'\nprop = {\n 'pdb_code': '3EBP',\n 'filter': False\n}\n\n# Launch bb on REST API\ntoken = launch_job(url = apiURL + 'launch/biobb_io/pdb', \n config = prop,\n output_pdb_path = downloaded_pdb)\n", "{\n \"code\": 303,\n \"state\": \"RUNNING\",\n \"message\": \"The requested job has has been successfully launched, please go to /retrieve/status/{token} for checking job status.\",\n \"token\": \"af60d733db949a71167f3aa6a7a793fc520b5a4176b57a770bed4798654a79be2a47b81e6a77de4eb285de84f9b768b119004f0bfbbe4be9e5ff1ffe31b81fd9\"\n}\n" ], [ "# Check job status\nout_files = check_job(token, apiURL)", "Total elapsed time: 0:00:06\nREST API JSON response:\n{\n \"code\": 200,\n \"state\": \"FINISHED\",\n \"message\": \"The requested job has finished successfully, please go to /retrieve/data/{id} for each output_files.\",\n \"output_files\": [\n {\n \"id\": \"5e428389eeafa3.49051362\",\n \"name\": \"3EBP.pdb\",\n \"size\": 609120,\n \"mimetype\": \"text/plain\"\n }\n ],\n \"expiration\": \"February 13, 2020 00:00 GMT+0000\"\n}\n" ], [ "# Save generated file to disk\nretrieve_data(out_files, apiURL)", "_____no_output_____" ] ], [ [ "Visualize downloaded PDB in NGLView:", "_____no_output_____" ] ], [ [ "import nglview\n\n# Show protein\nview = nglview.show_structure_file(downloaded_pdb)\nview.add_representation(repr_type='ball+stick', selection='het')\nview._remote_call('setSize', target='Widget', args=['','600px'])\nview", "_____no_output_____" ], [ "view.render_image()\nview.download_image(filename='ngl1.png')", "_____no_output_____" ] ], [ [ "<img src='ngl1.png'></img>", "_____no_output_____" ], [ "<a id=\"example2\"></a>\n### Example 2: extract heteroatom from a given structure", "_____no_output_____" ], [ "Launch the *biobb_structure_utils.extract_heteroatoms* job that extracts a heteroatom from a PDB file.", "_____no_output_____" ] ], [ [ "# Extracting heteroatom from a given structure\n\n# Create properties dict and inputs/outputs\nheteroatom = 'CPB.pdb'\nprop = {\n 'heteroatoms': [{\n 'name': 'CPB'\n }]\n}\n\n# Launch bb on REST API\ntoken = launch_job(url = apiURL + 'launch/biobb_structure_utils/extract_heteroatoms', \n config = prop,\n input_structure_path = downloaded_pdb,\n output_heteroatom_path = heteroatom)\n", "{\n \"code\": 303,\n \"state\": \"RUNNING\",\n \"message\": \"The requested job has has been successfully launched, please go to /retrieve/status/{token} for checking job status.\",\n \"token\": \"740c9ac30767ba996e445e4ed05c151ee903fed2234c0cc7ace6ec3ba4e1fa8bdcd5a3c6835c7f1038530eb81c4cc319674f235cf55b38863903181dce09a8d1\"\n}\n" ], [ "# Check job status\nout_files = check_job(token, apiURL)", "Total elapsed time: 0:00:20\nREST API JSON response:\n{\n \"code\": 200,\n \"state\": \"FINISHED\",\n \"message\": \"The requested job has finished successfully, please go to /retrieve/data/{id} for each output_files.\",\n \"output_files\": [\n {\n \"id\": \"5e4283986555a0.86371712\",\n \"name\": \"CPB.pdb\",\n \"size\": 2268,\n \"mimetype\": \"text/plain\"\n }\n ],\n \"expiration\": \"February 13, 2020 00:00 GMT+0000\"\n}\n" ], [ "# Save generated file to disk\nretrieve_data(out_files, apiURL)", "_____no_output_____" ] ], [ [ "Visualize generated extracted heteroatom in NGLView:", "_____no_output_____" ] ], [ [ "# Show protein\nview = nglview.show_structure_file(heteroatom)\nview.add_representation(repr_type='ball+stick', selection='het')\nview._remote_call('setSize', target='Widget', args=['','600px'])\nview", "_____no_output_____" ], [ "view.render_image()\nview.download_image(filename='ngl2.png')", "_____no_output_____" ] ], [ [ "<img src='ngl2.png'></img>", "_____no_output_____" ], [ "<a id=\"example3\"></a>\n### Example 3: extract energy components from a given GROMACS energy file", "_____no_output_____" ] ], [ [ "# GMXEnergy: Getting system energy by time \n\n# Create prop dict and inputs/outputs\noutput_min_ene_xvg ='file_min_ene.xvg'\noutput_min_edr = 'files/1AKI_min.edr'\nprop = {\n 'terms': [\"Potential\"]\n}\n\n# Launch bb on REST API\ntoken = launch_job(url = apiURL + 'launch/biobb_analysis/gmx_energy',\n config = prop,\n input_energy_path = output_min_edr,\n output_xvg_path = output_min_ene_xvg)", "{\n \"code\": 303,\n \"state\": \"RUNNING\",\n \"message\": \"The requested job has has been successfully launched, please go to /retrieve/status/{token} for checking job status.\",\n \"token\": \"170e8e2645d179eaa40e2de652f3e6dec909ef1df4642526ba789ed21806bab917bb4ce7f9fb730dfe635155f52ac9f3869c4429afcc767bd190f01337a8a718\"\n}\n" ], [ "# Check job status\nout_files = check_job(token, apiURL)", "Total elapsed time: 0:00:08\nREST API JSON response:\n{\n \"code\": 200,\n \"state\": \"FINISHED\",\n \"message\": \"The requested job has finished successfully, please go to /retrieve/data/{id} for each output_files.\",\n \"output_files\": [\n {\n \"id\": \"5e4283a6c70143.38956052\",\n \"name\": \"file_min_ene.xvg\",\n \"size\": 54143,\n \"mimetype\": \"text/plain\"\n }\n ],\n \"expiration\": \"February 13, 2020 00:00 GMT+0000\"\n}\n" ], [ "# Save generated file to disk\nretrieve_data(out_files, apiURL)", "_____no_output_____" ] ], [ [ "Visualize generated energy file in plotly:", "_____no_output_____" ] ], [ [ "import plotly\nimport plotly.graph_objs as go\n\n#Read data from file and filter energy values higher than 1000 Kj/mol^-1\nwith open(output_min_ene_xvg,'r') as energy_file:\n x,y = map(\n list,\n zip(*[\n (float(line.split()[0]),float(line.split()[1]))\n for line in energy_file \n if not line.startswith((\"#\",\"@\")) \n if float(line.split()[1]) < 1000 \n ])\n )\n\nplotly.offline.init_notebook_mode(connected=True)\n\nfig = {\n \"data\": [go.Scatter(x=x, y=y)],\n \"layout\": go.Layout(title=\"Energy Minimization\",\n xaxis=dict(title = \"Energy Minimization Step\"),\n yaxis=dict(title = \"Potential Energy KJ/mol-1\")\n )\n}\n\nplotly.offline.iplot(fig)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ] ]

[ "CC-BY-3.0" ]

[ "CC-BY-3.0" ]

[ "CC-BY-3.0" ]

[ [ [ "An illustration of the metric and non-metric MDS on generated noisy data.\n\nThe reconstructed points using the metric MDS and non metric MDS are slightly shifted to avoid overlapping.", "_____no_output_____" ], [ "#### New to Plotly?\nPlotly's Python library is free and open source! [Get started](https://plot.ly/python/getting-started/) by downloading the client and [reading the primer](https://plot.ly/python/getting-started/).\n<br>You can set up Plotly to work in [online](https://plot.ly/python/getting-started/#initialization-for-online-plotting) or [offline](https://plot.ly/python/getting-started/#initialization-for-offline-plotting) mode, or in [jupyter notebooks](https://plot.ly/python/getting-started/#start-plotting-online).\n<br>We also have a quick-reference [cheatsheet](https://images.plot.ly/plotly-documentation/images/python_cheat_sheet.pdf) (new!) to help you get started!", "_____no_output_____" ], [ "### Version", "_____no_output_____" ] ], [ [ "import sklearn\nsklearn.__version__", "_____no_output_____" ] ], [ [ "### Imports", "_____no_output_____" ] ], [ [ "print(__doc__)\n\nimport plotly.plotly as py\nimport plotly.graph_objs as go\n\nimport numpy as np\n\nfrom sklearn import manifold\nfrom sklearn.metrics import euclidean_distances\nfrom sklearn.decomposition import PCA", "Automatically created module for IPython interactive environment\n" ] ], [ [ "### Calculations", "_____no_output_____" ] ], [ [ "n_samples = 20\nseed = np.random.RandomState(seed=3)\nX_true = seed.randint(0, 20, 2 * n_samples).astype(np.float)\nX_true = X_true.reshape((n_samples, 2))\n# Center the data\nX_true -= X_true.mean()\n\nsimilarities = euclidean_distances(X_true)\n\n# Add noise to the similarities\nnoise = np.random.rand(n_samples, n_samples)\nnoise = noise + noise.T\nnoise[np.arange(noise.shape[0]), np.arange(noise.shape[0])] = 0\nsimilarities += noise\n\nmds = manifold.MDS(n_components=2, max_iter=3000, eps=1e-9, random_state=seed,\n dissimilarity=\"precomputed\", n_jobs=1)\npos = mds.fit(similarities).embedding_\n\nnmds = manifold.MDS(n_components=2, metric=False, max_iter=3000, eps=1e-12,\n dissimilarity=\"precomputed\", random_state=seed, n_jobs=1,\n n_init=1)\nnpos = nmds.fit_transform(similarities, init=pos)\n\n# Rescale the data\npos *= np.sqrt((X_true ** 2).sum()) / np.sqrt((pos ** 2).sum())\nnpos *= np.sqrt((X_true ** 2).sum()) / np.sqrt((npos ** 2).sum())\n\n# Rotate the data\nclf = PCA(n_components=2)\nX_true = clf.fit_transform(X_true)\n\npos = clf.fit_transform(pos)\n\nnpos = clf.fit_transform(npos)", "_____no_output_____" ] ], [ [ "### Plot Results", "_____no_output_____" ] ], [ [ "data = []\np1 = go.Scatter(x=X_true[:, 0], y=X_true[:, 1], \n mode='markers+lines',\n marker=dict(color='navy', size=10),\n line=dict(width=1),\n name='True Position')\ndata.append(p1)\np2 = go.Scatter(x=pos[:, 0], y=pos[:, 1],\n mode='markers+lines',\n marker=dict(color='turquoise', size=10),\n line=dict(width=1),\n name='MDS')\ndata.append(p2)\np3 = go.Scatter(x=npos[:, 0], y=npos[:, 1], \n mode='markers+lines',\n marker=dict(color='orange', size=10),\n line=dict(width=1),\n name='NMDS')\ndata.append(p3)\n\nsimilarities = similarities.max() / similarities * 100\nsimilarities[np.isinf(similarities)] = 0\n\n# Plot the edges\nstart_idx, end_idx = np.where(pos)\n# a sequence of (*line0*, *line1*, *line2*), where::\n# linen = (x0, y0), (x1, y1), ... (xm, ym)\nsegments = [[X_true[i, :], X_true[j, :]]\n for i in range(len(pos)) for j in range(len(pos))]\nvalues = np.abs(similarities)\nfor i in range(len(segments)):\n p4 = go.Scatter(x=[segments[i][0][0],segments[i][1][0]],\n y=[segments[i][0][1],segments[i][1][1]],\n mode = 'lines',\n showlegend=False,\n line = dict(\n color = 'lightblue',\n width = 0.5))\n data.append(p4)\n \nlayout = go.Layout(xaxis=dict(zeroline=False, showgrid=False,\n ticks='', showticklabels=False),\n yaxis=dict(zeroline=False, showgrid=False,\n ticks='', showticklabels=False),\n height=900, hovermode='closest')\nfig = go.Figure(data=data, layout=layout)", "_____no_output_____" ], [ "py.iplot(fig)", "_____no_output_____" ] ], [ [ "### License", "_____no_output_____" ], [ "Author: \n \n Nelle Varoquaux <[email protected]>\n\nLicense:\n \n BSD\n", "_____no_output_____" ] ], [ [ "from IPython.display import display, HTML\n\ndisplay(HTML('<link href=\"//fonts.googleapis.com/css?family=Open+Sans:600,400,300,200|Inconsolata|Ubuntu+Mono:400,700\" rel=\"stylesheet\" type=\"text/css\" />'))\ndisplay(HTML('<link rel=\"stylesheet\" type=\"text/css\" href=\"http://help.plot.ly/documentation/all_static/css/ipython-notebook-custom.css\">'))\n\n! pip install git+https://github.com/plotly/publisher.git --upgrade\nimport publisher\npublisher.publish(\n 'Multi-Dimensional Scaling.ipynb', 'scikit-learn/plot-mds/', 'Multi-Dimensional Scaling | plotly',\n '',\n title = 'Multi-Dimensional Scaling | plotly',\n name = 'Multi-Dimensional Scaling',\n has_thumbnail='true', thumbnail='thumbnail/mds.jpg', \n language='scikit-learn', page_type='example_index',\n display_as='manifold_learning', order=2,\n ipynb= '~Diksha_Gabha/3320')", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# Web Coverage Service (WCS) Download Example\n## Introduction\nWe'll demonstrate how to download a GeoTIFF data file from a public WCS service using Python 3. \n### WCS Data Service\nFor this demonstration we'll use Landfire (LF_1.4.0): https://www.landfire.gov/data_access.php\nFor Landfire LF_1.4.0 we see that the base URL is https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\n## WCS Requests\nThere are three types of WCS requests:\n- GetCapabilities\n- DescribeCoverage\n- GetCoverage\n\nGenerally, you first do a GetCapabilities request to obtain high-level information about what data you can ask for from the WCS server. \nYou then perform a DescribeCoverage request to get information specific to the coverage you want to get data from. \nFinally, you perform a GetCoverage to obtain the data itself.\n### GetCapabilities\nLet's perform a GetCapabilities request on the Landfire WCS server to see what the service can do.", "_____no_output_____" ] ], [ [ "import requests\n\n# base WCS server URL\nwcs_base_url = \"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"\n\n# add on to the base WCS server URL to define GetCapabilities URL\nwcs_get_capabilities_url = wcs_base_url + \"?REQUEST=GetCapabilities&SERVICE=WCS\"\n\n# perform an HTTP GET request with the GetCapabilities URL\nwcs_get_capabilities_response = requests.get(wcs_get_capabilities_url)\n\n# show the resulting body of the GetCapabilities request\nprint(\"GetCapabilities Response:\")\nprint(wcs_get_capabilities_response.text)", "GetCapabilities Response:\n<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<Capabilities xmlns=\"http://www.opengis.net/wcs/1.1\"\n xmlns:ows=\"http://www.opengis.net/ows/1.1\"\n xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\"\n xmlns:xlink=\"http://www.w3.org/1999/xlink\"\n xsi:schemaLocation=\"http://www.opengis.net/wcs/1.1 http://schemas.opengis.net/wcs/1.1/wcsGetCapabilities.xsd http://www.opengis.net/ows/1.1/ http://schemas.opengis.net/ows/1.1.0/owsAll.xsd\" version=\"1.1.2\">\n <ows:ServiceIdentification>\n <ows:Title>US_140</ows:Title>\n <ows:ServiceType>WCS</ows:ServiceType>\n <ows:ServiceTypeVersion>1.0.0</ows:ServiceTypeVersion>\n <ows:ServiceTypeVersion>1.1.0</ows:ServiceTypeVersion>\n <ows:ServiceTypeVersion>1.1.1</ows:ServiceTypeVersion>\n <ows:ServiceTypeVersion>1.1.2</ows:ServiceTypeVersion>\n <ows:Fees>NONE</ows:Fees>\n <ows:AccessConstraints>NONE</ows:AccessConstraints>\n </ows:ServiceIdentification>\n <ows:ServiceProvider>\n <ows:ProviderName></ows:ProviderName>\n <ows:ServiceContact>\n <ows:ContactInfo>\n <ows:Phone>\n </ows:Phone>\n <ows:Address>\n </ows:Address>\n <ows:OnlineResource xlink:href=\"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"/>\n </ows:ContactInfo>\n </ows:ServiceContact>\n </ows:ServiceProvider>\n <ows:OperationsMetadata>\n <ows:Operation name=\"GetCapabilities\">\n <ows:DCP>\n <ows:HTTP>\n <ows:Get xlink:href=\"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"/>\n <ows:Post xlink:href=\"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"/>\n </ows:HTTP>\n </ows:DCP>\n <ows:Parameter name=\"service\">\n <ows:AllowedValues>\n <ows:Value>WCS</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n <ows:Parameter name=\"AcceptVersions\">\n <ows:AllowedValues>\n <ows:Value>1.0.0</ows:Value>\n <ows:Value>1.1.0</ows:Value>\n <ows:Value>1.1.1</ows:Value>\n <ows:Value>1.1.2</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n </ows:Operation>\n <ows:Operation name=\"DescribeCoverage\">\n <ows:DCP>\n <ows:HTTP>\n <ows:Get xlink:href=\"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"/>\n <ows:Post xlink:href=\"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"/>\n </ows:HTTP>\n </ows:DCP>\n <ows:Parameter name=\"service\">\n <ows:AllowedValues>\n <ows:Value>WCS</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n <ows:Parameter name=\"version\">\n <ows:AllowedValues>\n <ows:Value>1.0.0</ows:Value>\n <ows:Value>1.1.0</ows:Value>\n <ows:Value>1.1.1</ows:Value>\n <ows:Value>1.1.2</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n <ows:Parameter name=\"Identifier\">\n <ows:AllowedValues>\n <ows:Value>1</ows:Value>\n <ows:Value>2</ows:Value>\n <ows:Value>3</ows:Value>\n <ows:Value>4</ows:Value>\n <ows:Value>5</ows:Value>\n <ows:Value>6</ows:Value>\n <ows:Value>7</ows:Value>\n <ows:Value>8</ows:Value>\n <ows:Value>9</ows:Value>\n <ows:Value>10</ows:Value>\n <ows:Value>11</ows:Value>\n <ows:Value>12</ows:Value>\n <ows:Value>13</ows:Value>\n <ows:Value>14</ows:Value>\n <ows:Value>15</ows:Value>\n <ows:Value>16</ows:Value>\n <ows:Value>17</ows:Value>\n <ows:Value>18</ows:Value>\n <ows:Value>19</ows:Value>\n <ows:Value>20</ows:Value>\n <ows:Value>21</ows:Value>\n <ows:Value>22</ows:Value>\n <ows:Value>23</ows:Value>\n <ows:Value>24</ows:Value>\n <ows:Value>25</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n </ows:Operation>\n <ows:Operation name=\"GetCoverage\">\n <ows:DCP>\n <ows:HTTP>\n <ows:Get xlink:href=\"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"/>\n <ows:Post xlink:href=\"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"/>\n </ows:HTTP>\n </ows:DCP>\n <ows:Parameter name=\"service\">\n <ows:AllowedValues>\n <ows:Value>WCS</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n <ows:Parameter name=\"version\">\n <ows:AllowedValues>\n <ows:Value>1.0.0</ows:Value>\n <ows:Value>1.1.0</ows:Value>\n <ows:Value>1.1.1</ows:Value>\n <ows:Value>1.1.2</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n <ows:Parameter name=\"Identifier\">\n <ows:AllowedValues>\n <ows:Value>1</ows:Value>\n <ows:Value>2</ows:Value>\n <ows:Value>3</ows:Value>\n <ows:Value>4</ows:Value>\n <ows:Value>5</ows:Value>\n <ows:Value>6</ows:Value>\n <ows:Value>7</ows:Value>\n <ows:Value>8</ows:Value>\n <ows:Value>9</ows:Value>\n <ows:Value>10</ows:Value>\n <ows:Value>11</ows:Value>\n <ows:Value>12</ows:Value>\n <ows:Value>13</ows:Value>\n <ows:Value>14</ows:Value>\n <ows:Value>15</ows:Value>\n <ows:Value>16</ows:Value>\n <ows:Value>17</ows:Value>\n <ows:Value>18</ows:Value>\n <ows:Value>19</ows:Value>\n <ows:Value>20</ows:Value>\n <ows:Value>21</ows:Value>\n <ows:Value>22</ows:Value>\n <ows:Value>23</ows:Value>\n <ows:Value>24</ows:Value>\n <ows:Value>25</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n <ows:Parameter name=\"InterpolationType\">\n <ows:AllowedValues>\n <ows:Value>nearest</ows:Value>\n <ows:Value>bilinear</ows:Value>\n <ows:Value>bicubic</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n <ows:Parameter name=\"format\">\n <ows:AllowedValues>\n <ows:Value>image/GeoTIFF</ows:Value>\n <ows:Value>image/NITF</ows:Value>\n <ows:Value>image/JPEG</ows:Value>\n <ows:Value>image/PNG</ows:Value>\n <ows:Value>image/JPEG2000</ows:Value>\n <ows:Value>image/HDF</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n <ows:Parameter name=\"store\">\n <ows:AllowedValues>\n <ows:Value>False</ows:Value>\n <ows:Value>True</ows:Value>\n </ows:AllowedValues>\n </ows:Parameter>\n </ows:Operation>\n </ows:OperationsMetadata>\n <Contents>\n <CoverageSummary>\n <ows:Title>US_DIST2013</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_DIST2013</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_DIST2014</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_DIST2014</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140CBD</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140CBD</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140CBH</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140CBH</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140CH</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140CH</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140EVH</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140EVH</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140VDEP</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140VDEP</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140FBFM40</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140FBFM40</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140FBFM13</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140FBFM13</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140MFRI</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98737642604233 22.765512000364211</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681016233757</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140MFRI</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140FRG</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98737642604233 22.765512000364211</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681016233757</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140FRG</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140ESP</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140ESP</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140BPS</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98737642604233 22.765512000364211</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681016233757</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140BPS</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140SCLASS</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140SCLASS</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140CC</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140CC</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140EVC</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140EVC</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140EVT</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140EVT</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140PRS</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98737642604233 22.765512000364211</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681016233757</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140PRS</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140PMS</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98737642604233 22.765512000364211</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681016233757</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140PMS</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140PLS</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98737642604233 22.765512000364211</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681016233757</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140PLS</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_VDIST2014</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_VDIST2014</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_VTM2014</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_VTM2014</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_FDIST2014</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_FDIST2014</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140VCC</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765446426860603</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140VCC</Identifier>\n </CoverageSummary>\n <CoverageSummary>\n <ows:Title>US_140FCCS</ows:Title>\n <ows:Abstract>tempname</ows:Abstract>\n <ows:WGS84BoundingBox>\n <ows:LowerCorner>-127.98775263969655 22.765714837805092</ows:LowerCorner>\n <ows:UpperCorner>-65.254445466369276 51.649681015029245</ows:UpperCorner>\n </ows:WGS84BoundingBox>\n <Identifier>US_140FCCS</Identifier>\n </CoverageSummary>\n <SupportedCRS>urn:ogc:def:crs:EPSG::4326</SupportedCRS>\n <SupportedCRS>urn:ogc:def:crs:EPSG::102039</SupportedCRS>\n <SupportedFormat>image/GeoTIFF</SupportedFormat>\n <SupportedFormat>image/NITF</SupportedFormat>\n <SupportedFormat>image/JPEG</SupportedFormat>\n <SupportedFormat>image/PNG</SupportedFormat>\n <SupportedFormat>image/JPEG2000</SupportedFormat>\n <SupportedFormat>image/HDF</SupportedFormat>\n </Contents>\n</Capabilities>\n\n" ] ], [ [ "#### GetCapabilities results\nThe GetCapabilities request returns a lot of information for us. \n\nWe can see that the WCS server supports WCS version `1.0.0`, `1.1.0`, `1.1.1`, and `1.1.2` . \nFor the purposes of this guide, we're going to stick to WCS 1.0.0.\n\nWithin the contents section, we can see all of the available coverages. A coverage is just another word for a data layer.\nFor this guide, let's pick an arbitrary data layer: `US_VDIST2014`\n\n### DescribeCoverage\n\nWe'll use the results we got from GetCoverage to perform a DescribeCoverage request to get more specific information about how to ultimately perform the GetCoverage Operation. We'll use WCS version `1.0.0` with coverage `US_VDIST2014` to perform the request.", "_____no_output_____" ] ], [ [ "import requests\n\n# base WCS server URL\nwcs_base_url = \"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"\n\n# add on to the base WCS server URL to define DescribeCoverage URL\nwcs_describe_coverage_url = wcs_base_url + \"?SERVICE=WCS&VERSION=1.0.0&REQUEST=DescribeCoverage&COVERAGE=US_VDIST2014\"\n\n# perform an HTTP GET request with the DescribeCoverage URL\nwcs_describe_coverage_response = requests.get(wcs_describe_coverage_url)\n\n# show the resulting body of the DescribeCoverage request\nprint(\"DescribeCoverage Response:\")\nprint(wcs_describe_coverage_response.text)", "DescribeCoverage Response:\n<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<CoverageDescription xmlns=\"http://www.opengis.net/wcs\"\n xmlns:gml=\"http://www.opengis.net/gml\"\n xmlns:xlink=\"http://www.w3.org/1999/xlink\"\n xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\"\n xsi:schemaLocation=\"http://www.opengis.net/wcs http://schemas.opengeospatial.net/wcs/1.0.0/describeCoverage.xsd\" version=\"1.0.0\">\n<CoverageOffering>\n <description>tempname</description>\n <name>US_VDIST2014</name>\n <label>US_VDIST2014</label>\n <lonLatEnvelope srsName=\"urn:ogc:def:crs:OGC:1.3:CRS84\">\n <gml:pos dimension=\"2\">-127.98775263969655 22.765446426860603</gml:pos>\n <gml:pos dimension=\"2\">-65.254445466369276 51.649681015029245</gml:pos>\n </lonLatEnvelope>\n <domainSet>\n <spatialDomain>\n <gml:Envelope srsName=\"EPSG:102039\">\n <gml:pos dimension=\"2\">-2362425 258915</gml:pos>\n <gml:pos dimension=\"2\">2263815 3177435</gml:pos>\n </gml:Envelope>\n <gml:RectifiedGrid srsName=\"EPSG:102039\" dimension=\"2\">\n <gml:limits>\n <gml:GridEnvelope>\n <gml:low>0 0</gml:low> \n <gml:high>154207 97283</gml:high>\n </gml:GridEnvelope>\n </gml:limits>\n <gml:axisName>Raster_Pixel_Columns(X-axis)</gml:axisName>\n <gml:axisName>Raster_Pixel_Rows(Y-axis)</gml:axisName>\n <gml:origin>\n <gml:pos>-2362410 3177420</gml:pos>\n </gml:origin>\n <gml:offsetVector>30 0</gml:offsetVector>\n <gml:offsetVector>0 -30</gml:offsetVector>\n </gml:RectifiedGrid>\n </spatialDomain>\n </domainSet>\n <rangeSet>\n <RangeSet>\n <name>RangeSet_21</name>\n <label>US_VDIST2014 RangeSet</label>\n <axisDescription>\n <AxisDescription>\n <name>Band</name>\n <label>Band Numbers</label>\n <values>\n <singleValue>1</singleValue>\n </values>\n </AxisDescription>\n </axisDescription>\n </RangeSet>\n </rangeSet>\n <supportedCRSs>\n <requestResponseCRSs>EPSG:4326</requestResponseCRSs>\n <requestResponseCRSs>EPSG:102039</requestResponseCRSs>\n <nativeCRSs>EPSG:102039</nativeCRSs>\n </supportedCRSs>\n <supportedFormats nativeFormat=\"GeoTIFF\">\n <formats>GeoTIFF</formats>\n <formats>NITF</formats>\n <formats>HDF</formats>\n <formats>JPEG2000</formats>\n </supportedFormats>\n <supportedInterpolations default=\"nearest neighbor\">\n <interpolationMethod>nearest neighbor</interpolationMethod>\n <interpolationMethod>bilinear</interpolationMethod>\n <interpolationMethod>bicubic</interpolationMethod>\n </supportedInterpolations>\n</CoverageOffering>\n</CoverageDescription>\n\n" ] ], [ [ "#### DescribeCoverage results\nFrom the DescribeCoverage response, we can see information about the available bounding box for the layer, the Coordinate Reference System (CRS), the download data types (e.g. GeoTiff), and more. We'll use this information to help formulate a successful WCS GetCoverage request to download some actual data.\n\n### GetCoverage\nWith the results from DescribeCoverage, we now know everything we need to know in order to download some data using a GetCoverage request.\nOne unfortunate aspect of WCS version 1.0.0 is that you must provide the resolution or number of pixels of your downloaded output file; you cannot omit this and simply ask for the download file to be in the native resolution of the source data. This means that you may need to perform some additional math to calculate the native resolution on your own before downloading the file. For the purposes of this tutorial, we'll simply download a 500 by 100 pixel image of the continental United States. ", "_____no_output_____" ] ], [ [ "import requests\n\n# base WCS server URL\nwcs_base_url = \"https://landfire.cr.usgs.gov/arcgis/services/Landfire/US_140/MapServer/WCSServer\"\n\n# add on to the base WCS server URL to define GetCoverage URL\nwcs_get_coverage_url = wcs_base_url + \"?SERVICE=WCS&VERSION=1.0.0&REQUEST=GetCoverage&FORMAT=GeoTIFF&COVERAGE=US_VDIST2014&BBOX=-127.98775263969655,22.765446426860603,-65.254445466369276,51.649681015029245&CRS=EPSG:4326&WIDTH=500&HEIGHT=100\"\n\n# perform an HTTP GET request with the DescribeCoverage URL\nwcs_get_coverage_response = requests.get(wcs_get_coverage_url)\n\n# download the resulting response image to disk\nif wcs_get_coverage_response.status_code == 200:\n with open(\"wcs-example.tif\", 'wb') as f:\n f.write(wcs_get_coverage_response.content)", "_____no_output_____" ] ], [ [ "#### GetCoverage results\nAt this point you should have a downloaded GeoTiff file called `wcs-example.tif`. Once you use the GetCapabilities and DescribeCoverage requests to dial in the types of GetCoverage requests you want to perform, you can simply tweak your GetCoverage requests to perform these types of requests in an automated fashion on your code.", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Method-I:-For-Loop\" data-toc-modified-id=\"Method-I:-For-Loop-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>Method I: For Loop</a></span></li><li><span><a href=\"#Method-II:-List-Comprehension\" data-toc-modified-id=\"Method-II:-List-Comprehension-2\"><span class=\"toc-item-num\">2&nbsp;&nbsp;</span>Method II: List Comprehension</a></span></li><li><span><a href=\"#Method-III:-Combining\" data-toc-modified-id=\"Method-III:-Combining-3\"><span class=\"toc-item-num\">3&nbsp;&nbsp;</span>Method III: Combining</a></span></li></ul></div>", "_____no_output_____" ] ], [ [ "import numpy as np\nimport os\nimport random\nimport matplotlib.pyplot as plt", "_____no_output_____" ], [ "arr=[1,2,3,4,5,6,6,7,7,7,7,8,5,3,2]\nnp_arr=np.array(arr)", "_____no_output_____" ], [ "np_arr", "_____no_output_____" ] ], [ [ "Note: () for functions, [] for calling elements of a list or array.", "_____no_output_____" ] ], [ [ "def my_function(x):\n return x**2", "_____no_output_____" ], [ "my_function(10)", "_____no_output_____" ] ], [ [ "# Method I: For Loop", "_____no_output_____" ] ], [ [ "arr1=[]\nfor i in range(100):\n arr1.append(i**(1/2)+1+i**1.2)\n ", "_____no_output_____" ] ], [ [ "plt.plot(Xvalues,Yvalues,....properties)\n\nplt.show()", "_____no_output_____" ] ], [ [ "plt.plot(range(100),arr1)\nplt.show()", "_____no_output_____" ], [ "arr2=[]\nfor i in range(1000):\n arr2.append(random.random())", "_____no_output_____" ], [ "plt.plot(range(1000),arr2)\nplt.show()", "_____no_output_____" ] ], [ [ "# Method II: List Comprehension", "_____no_output_____" ] ], [ [ "arr3=[i**(1/2) for i in range(100)]", "_____no_output_____" ], [ "plt.plot(range(100),arr3)", "_____no_output_____" ], [ "arr4=[random.random() for i in range(100)]", "_____no_output_____" ], [ "arr4", "_____no_output_____" ] ], [ [ "# Method III: Combining", "_____no_output_____" ] ], [ [ "len(arr3+arr4), len(arr3),len(arr4)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "# Python datetime module\n\nWe will look at an important standard library, the [datetime library][1] which contains many powerful functions to support date, time and datetime manipulation. Pandas does not rely on this object and instead creates its own, a `Timestamp`, discussed in other notebooks.\n\nThe datetime library is part of the standard library, so it comes shipped along with every Python installation. Let's get started by importing it into our namespace.\n\n[1]: https://docs.python.org/3/library/datetime.html", "_____no_output_____" ] ], [ [ "import datetime", "_____no_output_____" ] ], [ [ "## Create a date, a time and a datetime\nThe datetime module provides three separate objects for dates, times, and datetimes. Let's use the `date` type to construct a date. It takes three integers, the year, month and day. Here we create the date April 11, 2016.", "_____no_output_____" ] ], [ [ "my_date = datetime.date(2016, 4, 11)\nmy_date", "_____no_output_____" ], [ "type(my_date)", "_____no_output_____" ] ], [ [ "Use the `time` type to construct a time. It takes 4 integers - hours, minutes, seconds, and microseconds (one millionth of a second). Here we create the time 10:54:32.034512", "_____no_output_____" ] ], [ [ "my_time = datetime.time(10, 54, 32, 34512)\nmy_time", "_____no_output_____" ], [ "type(my_time)", "_____no_output_____" ] ], [ [ "Only the hour component is mandatory. For instance, we can create the time 5:44 with this:", "_____no_output_____" ] ], [ [ "datetime.time(5, 44)", "_____no_output_____" ] ], [ [ "Or you can specify just a particular component of time.", "_____no_output_____" ] ], [ [ "datetime.time(second=34)", "_____no_output_____" ] ], [ [ "Finally, we can construct a datetime with the `datetime` type, which takes up to 7 parameters - three for the date, and four for the time.", "_____no_output_____" ] ], [ [ "my_datetime = datetime.datetime(2016, 4, 11, 10, 54, 32, 34512)\nmy_datetime", "_____no_output_____" ], [ "type(my_datetime)", "_____no_output_____" ] ], [ [ "### Format changes when printed to the screen\nPrinting the objects from above to the screen provides a more readable view.", "_____no_output_____" ] ], [ [ "print(my_date)\nprint(my_time)\nprint(my_datetime)", "_____no_output_____" ] ], [ [ "## Attributes of date, time, and datetimes\nEach individual component of the date, time, and datetime is available as an attribute.", "_____no_output_____" ] ], [ [ "my_date.year", "_____no_output_____" ], [ "my_date.month", "_____no_output_____" ], [ "my_date.day", "_____no_output_____" ], [ "my_time.hour", "_____no_output_____" ], [ "my_time.minute", "_____no_output_____" ], [ "my_time.second", "_____no_output_____" ], [ "my_datetime.day", "_____no_output_____" ], [ "my_datetime.microsecond", "_____no_output_____" ], [ "my_date.weekday()", "_____no_output_____" ] ], [ [ "## Methods of date, time, and datetimes\nSeveral methods exist for each of these objects. The methods that begin with ISO represent the [International Standards Organization][1] formatting rules for dates, times, and datetimes. The particular standard here is [ISO 8601][2]. Python will return according to this standard.\n\n[1]: https://www.iso.org/home.html\n[2]: https://en.wikipedia.org/wiki/ISO_8601", "_____no_output_____" ] ], [ [ "my_date.weekday()", "_____no_output_____" ], [ "my_date.isoformat()", "_____no_output_____" ], [ "my_date.isocalendar()", "_____no_output_____" ], [ "my_time.isoformat()", "_____no_output_____" ], [ "my_datetime.isoformat()", "_____no_output_____" ], [ "# get the date from a datetime\nmy_datetime.date()", "_____no_output_____" ], [ "# get the time from a datetime\nmy_datetime.time()", "_____no_output_____" ] ], [ [ "## Alternate Constructors\nYou can create dates and datetimes from a single integer which represents the number of seconds since the Unix epoch, January 1, 1970 UTC. UTC is the timezone, [Coordinated Universal Time][1] and is 0 degrees longitude or 5 hours ahead of Easter Standard Time.\n\nPassing the integer 0 to the `fromtimestamp` datetime constructor will return a datetime at the Unix epoch adjusted to your local timezone. If you are located in EST, then you will get returned December 31, 1969 7 p.m.\n\n[1]: https://en.wikipedia.org/wiki/Coordinated_Universal_Time", "_____no_output_____" ] ], [ [ "datetime.datetime.fromtimestamp(0)", "_____no_output_____" ], [ "# 1 billion seconds from the unix epoch\ndatetime.datetime.fromtimestamp(10 ** 9)", "_____no_output_____" ] ], [ [ "The date type also has this constructor, but not time.", "_____no_output_____" ] ], [ [ "# also works for date\ndatetime.date.fromtimestamp(10 ** 9)", "_____no_output_____" ] ], [ [ "Can get todays date or datetime:", "_____no_output_____" ] ], [ [ "datetime.date.today()", "_____no_output_____" ], [ "datetime.datetime.now()", "_____no_output_____" ] ], [ [ "### Constructing from strings\nThe `strptime` alternate datetime constructor has the ability to convert a string into a datetime. In addition to the string, you must pass it a specific **format** to alert the constructor which part of the string corresponds to which component of the datetime. There are special character codes called **directives** which must be used to form this correspondence.\n\n## Directives\nAll the directives can be found in the official [Python documentation for the datetime module][1]. Below are some common ones.\n\n* **%y** - two digit year\n* **%Y** - four digit year\n* **%m** - Month\n* **%d** - Day of the month \n* **%H** - Hour (24-hour clock)\n* **%I** - Hour (12-hour clock)\n* **%M** - Minute\n\n[1]: https://docs.python.org/3/library/datetime.html#strftime-strptime-behavior", "_____no_output_____" ], [ "### Examples of string parsing to datetimes\nThe `strptime` alternate constructor stands for string parse time (though it only parses datetimes). You must create a string with the correct directives that represents the format of the date string you are trying to convert to a datetime. For instance, the string '2016-10-22' can use the format '%Y-%m-%d' to parse it correctly.", "_____no_output_____" ] ], [ [ "s = '2016-10-22'\nfmt = '%Y-%m-%d'\ndatetime.datetime.strptime(s, fmt)", "_____no_output_____" ], [ "s = '2016/1/22 5:32:44'\nfmt = '%Y/%m/%d %H:%M:%S'\ndatetime.datetime.strptime(s, fmt)", "_____no_output_____" ], [ "s = 'January 23, 2019 5:22 PM'\nfmt = '%B %d, %Y %H:%M %p'\ndatetime.datetime.strptime(s, fmt)", "_____no_output_____" ], [ "s = 'On January the 23rd 2019 at 5:22 PM'\nfmt = 'On %B the %drd %Y at %H:%M %p'\ndatetime.datetime.strptime(s, fmt)", "_____no_output_____" ] ], [ [ "### Converting datetimes to string\nThe **strftime** method converts a date, time, or datetime to a string. It stands for **string format time**. Begin with a date, time, or datetime and use a string with directives to make the conversion.", "_____no_output_____" ] ], [ [ "# Convert directly into a string of your choice. Lookup directives online\nmy_date.strftime(\"%Y-%m-%d\")", "_____no_output_____" ], [ "# Another more involved directive\nmy_date.strftime(\"Remembering back to %A, %B %d, %Y.... What a fantastic day that was.\")", "_____no_output_____" ] ], [ [ "## Date and Datetime addition\nIt's possible to add an amount of time to a date or datetime object using the timedelta function. timedelta simply produces some amount of time measured in days, seconds and microseconds. You can then use this object to add to date or datetime objects.\n\n**`timedelta`** objects are constructed with the following definition: \n\n**`timedelta(days=0, seconds=0, microseconds=0, milliseconds=0, minutes=0, hours=0, weeks=0)`**", "_____no_output_____" ] ], [ [ "my_timedelta = datetime.timedelta(seconds=5000) \nmy_timedelta", "_____no_output_____" ], [ "type(my_timedelta)", "_____no_output_____" ], [ "# add to datetime\nmy_datetime + my_timedelta", "_____no_output_____" ], [ "# original\nmy_datetime", "_____no_output_____" ], [ "# add to date\nmy_date + my_timedelta", "_____no_output_____" ], [ "# original date. Nothing changed since 5000 seconds wasn't long enough to make an extra day\nmy_date", "_____no_output_____" ], [ "# now there is a change\nmy_date + datetime.timedelta(days = 5)", "_____no_output_____" ], [ "# add weeks\na = my_datetime + datetime.timedelta(weeks=72, days=4, hours=44)", "_____no_output_____" ], [ "# the difference between the underlying string representation and the print function\nprint(a.__repr__())\nprint(a)", "_____no_output_____" ], [ "datetime.timedelta(weeks=72, days=4, hours=44)", "_____no_output_____" ] ], [ [ "## Third-Party library `dateutil`\nFor improved datetime handling, you can use [dateutil][1], a more advanced third-party library. Pandas actually uses this library to for its complex date handling. Two of it's most useful features are string parsing and datetime addition.\n\n### Advanced string handling\n\nThe `parse` function handles a wide variety of strings. It returns the same datetime type from above. See [many more examples][2] in the documentation.\n\n[1]: https://dateutil.readthedocs.io/en/stable/\n[2]: https://dateutil.readthedocs.io/en/stable/examples.html#parse-examples", "_____no_output_____" ] ], [ [ "from dateutil.parser import parse", "_____no_output_____" ], [ "parse('Jan 3, 2003 and 5:22')", "_____no_output_____" ] ], [ [ "Pandas uses this under the hood.", "_____no_output_____" ] ], [ [ "import pandas as pd\npd.Timestamp('Jan 3, 2003 and 5:22')", "_____no_output_____" ] ], [ [ "### Advanced datetime addition\nAn upgrade to the **`timedelta`** class exists with the **`relativedelta`** class. Check [this stackoverflow][1] post for more detail or see the [documentation for examples][2].\n\n[1]: http://stackoverflow.com/questions/12433233/what-is-the-difference-between-datetime-timedelta-and-dateutil-relativedelta\n[2]: https://dateutil.readthedocs.io/en/stable/relativedelta.html#examples", "_____no_output_____" ] ], [ [ "from dateutil.relativedelta import relativedelta", "_____no_output_____" ] ], [ [ "There are two ways to use it. First, you can pass it two datetimes to find the difference between the two.", "_____no_output_____" ] ], [ [ "dt1 = datetime.datetime(2016, 1, 20, 5, 33)\ndt2 = datetime.datetime(2018, 3, 20, 6, 22)\nrelativedelta(dt1, dt2)", "_____no_output_____" ] ], [ [ "Second, create an amount of time with the parameters years, months, weeks, days, etc... and then add that to a datetime.", "_____no_output_____" ] ], [ [ "rd = relativedelta(months=3)", "_____no_output_____" ], [ "dt1 + rd", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import numpy as np\nimport pandas as pd\nimport time\nimport os\n\nfrom pyspark.ml.clustering import KMeans\nfrom pyspark.ml.evaluation import ClusteringEvaluator\nfrom pyspark.ml.linalg import Vectors\n\nfrom matplotlib import pyplot as plt\n\nfrom pyspark.sql import SparkSession\n# from pyspark.ml.clustering import KMeans, KMeansModel\n\nimport networkx as nx # thư viện để tạo, thao tác, học cấu trúc.... của các mạng phức tạp\nimport seaborn as sns # thư viện để trực quan hóa dữ liệu\n\nsns.set_style('darkgrid', {'axes.facecolor': '.9'})\nsns.set_palette(palette='deep')\nsns_c = sns.color_palette(palette='deep')\n\nfloat_formatter = lambda x: \"%.6f\" % x\n\nnp.set_printoptions(formatter={'float_kind':float_formatter})\n", "_____no_output_____" ], [ "def draw_graph(G):\n pos = nx.spring_layout(G)\n nx.draw_networkx_nodes(G, pos, node_size=10)\n # nx.draw_networkx_labels(G, pos)\n nx.draw_networkx_edges(G, pos, width=0.1, alpha=0.5)\n\n\ndef draw_graph_cluster(G, labels):\n pos = nx.spring_layout(G)\n nx.draw(\n G,\n pos,\n node_color=node_colors,\n node_size=10,\n width=0.1,\n alpha=0.5,\n with_labels=False,\n )\n\ndef get_node_color(label):\n switcher = {\n 0: 'red',\n 1: 'blue',\n 2: 'orange',\n 3: 'gray',\n 4: 'violet',\n 5: 'pink',\n 6: 'purple',\n 7: 'brown',\n 8: 'yellow',\n 9: 'lime',\n 10: 'cyan'\n }\n return switcher.get(label, 'Invalid label')", "_____no_output_____" ], [ "spark = SparkSession.builder \\\n .master(\"local\") \\\n .appName(\"CLustering\") \\\n .config(\"spark.some.config.option\", \"some-value\") \\\n .getOrCreate()\nspark", "_____no_output_____" ], [ "base_path = os.getcwd()\n# file_input = base_path + \"/facebook_combined.txt\" \nfile_input = base_path + \"/ChG-Miner_miner-chem-gene.tsv\"\nfile_input", "_____no_output_____" ], [ "pdf = pd.read_table(file_input, sep='\\t', names=['src', 'dst'])\npdf.head()", "_____no_output_____" ], [ "pdf = pdf.to_numpy()", "_____no_output_____" ], [ "G = nx.Graph()\nG.add_edges_from(pdf)", "_____no_output_____" ], [ "len(G.nodes())", "_____no_output_____" ], [ "len(G.edges())", "_____no_output_____" ], [ "# adjacency matrix\nW = nx.adjacency_matrix(G)\nprint(W.todense())", "[[0 1 0 ... 0 0 0]\n [1 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n ...\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]\n [0 0 0 ... 0 0 0]]\n" ], [ "# degree matrix\nD = np.diag(np.sum(np.array(W.todense()), axis=1))\nprint(D)", "[[ 5 0 0 ... 0 0 0]\n [ 0 15 0 ... 0 0 0]\n [ 0 0 3 ... 0 0 0]\n ...\n [ 0 0 0 ... 1 0 0]\n [ 0 0 0 ... 0 1 0]\n [ 0 0 0 ... 0 0 1]]\n" ], [ "# Laplacian matrix\nL = D - W\nprint(L)", "[[ 5 -1 0 ... 0 0 0]\n [-1 15 0 ... 0 0 0]\n [ 0 0 3 ... 0 0 0]\n ...\n [ 0 0 0 ... 1 0 0]\n [ 0 0 0 ... 0 1 0]\n [ 0 0 0 ... 0 0 1]]\n" ], [ "# eigenvalues, eigenvector\neigenvals, eigenvcts = np.linalg.eigh(L)", "_____no_output_____" ], [ "eigenvcts", "_____no_output_____" ], [ "eigenvals", "_____no_output_____" ], [ "eigenvals_sorted_indices = np.argsort(eigenvals)\neigenvals_sorted = eigenvals[eigenvals_sorted_indices]", "_____no_output_____" ], [ "eigenvals_sorted_indices", "_____no_output_____" ], [ "eigenvals_sorted", "_____no_output_____" ], [ "fig, ax = plt.subplots(figsize=(10, 6))\nsns.lineplot(x=range(1, eigenvals_sorted_indices.size + 1), y=eigenvals_sorted, ax=ax)\nax.set(title='Sorted Eigenvalues Graph Laplacian', xlabel='index', ylabel=r'$\\lambda$')", "_____no_output_____" ], [ "index_lim = 250\n\nfig, ax = plt.subplots(figsize=(10, 6))\nsns.scatterplot(x=range(1, eigenvals_sorted_indices[: index_lim].size + 1), y=eigenvals_sorted[: index_lim], s=80, ax=ax)\nsns.lineplot(x=range(1, eigenvals_sorted_indices[: index_lim].size + 1), y=eigenvals_sorted[: index_lim], alpha=0.5, ax=ax)\nax.axvline(x=1, color=sns_c[3], label='zero eigenvalues', linestyle='--')\nax.legend()\nax.set(title=f'Sorted Eigenvalues Graph Laplacian (First {index_lim})', xlabel='index', ylabel=r'$\\lambda$')", "_____no_output_____" ], [ "zero_eigenvals_index = np.argwhere(abs(eigenvals) < 0.02)\nzero_eigenvals_index.squeeze()", "_____no_output_____" ], [ "proj_df = pd.DataFrame(eigenvcts[:, zero_eigenvals_index.squeeze()[206]])\n# proj_df = proj_df.transpose()\nproj_df = proj_df.rename(columns={0: 'features'})\nproj_df", "_____no_output_____" ], [ "U = []\nfor x in proj_df['features']:\n U.append(Vectors.dense(x))\npdf_train = pd.DataFrame(U, columns=['features'])\ndf = spark.createDataFrame(pdf_train)\ndisplay(df)", "_____no_output_____" ], [ "# train k-means model\ncost = np.zeros(15)\nfor i in range(2,15):\n kmeans = KMeans(k=i, seed=1)\n model = kmeans.fit(df)\n cost[i] = model.computeCost(df) # requires Spark 2.0 or later", "_____no_output_____" ], [ "fig, ax = plt.subplots(1,1, figsize =(8,6))\nax.plot(range(2,15),cost[2:15])\nax.set_xlabel('k')\nax.set_ylabel('cost')\nplt.show()", "_____no_output_____" ], [ "# train\nkmeans = KMeans(k=9, seed=1)\nmodel = kmeans.fit(df)\n\n# Make predictions\npredictions = model.transform(df)\nrows = predictions.select(\"features\",\"prediction\").collect()\n\n# Evaluate clustering by computing Silhouette score\nevaluator = ClusteringEvaluator()\nsilhouette = evaluator.evaluate(predictions)\nsilhouette", "_____no_output_____" ], [ "rows", "_____no_output_____" ], [ "node_colors = []\nfor label in rows:\n node_colors.append(get_node_color(label.prediction))", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

[ "MIT" ]

[ "MIT" ]

[ "MIT" ]

[ [ [ "import os\nos.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' \n\nfrom tensorflow import keras\nfrom sklearn.model_selection import train_test_split\nimport matplotlib.pyplot as plt\nimport numpy as np", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code" ] ]

[ "ADSL" ]

[ "ADSL" ]

[ "ADSL" ]

[ [ [ "# VacationPy\n----\n\n#### Note\n* Keep an eye on your API usage. Use https://developers.google.com/maps/reporting/gmp-reporting as reference for how to monitor your usage and billing.\n\n* Instructions have been included for each segment. You do not have to follow them exactly, but they are included to help you think through the steps.", "_____no_output_____" ] ], [ [ "# Dependencies and Setup\nimport matplotlib.pyplot as plt\nimport pandas as pd\nimport numpy as np\nimport requests\nimport gmaps\nimport os\n\n# Import API key\nfrom api_keys import g_key", "_____no_output_____" ] ], [ [ "### Store Part I results into DataFrame\n* Load the csv exported in Part I to a DataFrame", "_____no_output_____" ] ], [ [ "# Import World Happiness Report Data 2021\ncity_data = pd.read_csv(\"../WeatherPy/city_data.csv\")\ncity_data\n", "_____no_output_____" ] ], [ [ "### Humidity Heatmap\n* Configure gmaps.\n* Use the Lat and Lng as locations and Humidity as the weight.\n* Add Heatmap layer to map.", "_____no_output_____" ] ], [ [ "# Configure gmaps\ngmaps.configure(api_key=g_key)\n\n# Store 'Lat' and 'Lng' into locations \nlocations = city_data[[\"Lat\", \"Lng\"]].astype(float)\n\nhumidity = city_data['Humidity']", "_____no_output_____" ], [ "# Create a Humidity Heatmap\nfig = gmaps.figure(center=(0, 0), zoom_level=2)\n\nhumidity_heatmap = gmaps.heatmap_layer(locations, weights=humidity, \n dissipating=False, max_intensity=100,\n point_radius = 1)\n\n\nfig.add_layer(humidity_heatmap)\n\nfig", "_____no_output_____" ] ], [ [ "### Create new DataFrame fitting weather criteria\n* Narrow down the cities to fit weather conditions.\n* Drop any rows will null values.", "_____no_output_____" ] ], [ [ "# Dropna in city_data\ncity_data_narrowed = city_data.dropna()\n\n# Narrow down the cities to fit weather conditions\ncity_data_narrowed = city_data_narrowed[(city_data_narrowed['Max Temp'] > 70) & (city_data_narrowed['Max Temp'] < 80)]\ncity_data_narrowed = city_data_narrowed[city_data_narrowed['Wind Speed'] <= 10]\ncity_data_narrowed = city_data_narrowed[city_data_narrowed['Cloudiness'] == 0]\n\ncity_data_narrowed", "_____no_output_____" ] ], [ [ "### Hotel Map\n* Store into variable named `hotel_df`.\n* Add a \"Hotel Name\" column to the DataFrame.\n* Set parameters to search for hotels with 5000 meters.\n* Hit the Google Places API for each city's coordinates.\n* Store the first Hotel result into the DataFrame.\n* Plot markers on top of the heatmap.", "_____no_output_____" ] ], [ [ "# Store into variable named hotel_df\nhotel_df = city_data_narrowed\n\n# Add \"Hotel Name\" column\nhotel_df['Hotel Name'] = \"\"\n\nhotel_df", "_____no_output_____" ], [ "# Set parameters to search for hotels with 5000 meters\nbase_url = \"https://maps.googleapis.com/maps/api/place/nearbysearch/json?\"\n\nparams = {\"type\" : \"hotel\",\n \"keyword\" : \"hotel\",\n \"radius\" : 5000,\n \"key\" : g_key}\n", "_____no_output_____" ], [ "# Hit the Google Places API for each city's coordinates\n\n# Import time function \nimport time\n\n# Print header\nprint('Beginning Data Retrieval')\nprint('------------------------------------------------------------')\n\n# Loop through the hotel_df Dataframe with the iterrows function\nfor index, row in hotel_df.iterrows():\n lat = row['Lat']\n lng = row['Lng']\n city_name = row['City']\n \n params['location'] = f'{lat},{lng}'\n \n print('Processing Record for {} (Index {}):'.format(city_name, index))\n \n response = requests.get(base_url, params=params).json()\n results = response['results']\n \n try:\n hotel_df.loc[index, 'Hotel Name'] = results[0]['name']\n print('The closest hotel in {} is {}'.format(city_name, results[0]['name']))\n \n except (KeyError, IndexError):\n print('City not found. Skipping...')\n print(\"------------------------------------------------------------\")\n \n # Limit waiting time\n time.sleep(1)\n \n pass\n \n print(\"------------------------------------------------------------\")\n \n# Print footer\nprint('------------------------------------------------------------')\nprint('Data Retrieval Complete ')\nprint('------------------------------------------------------------')", "Beginning Data Retrieval\n------------------------------------------------------------\nProcessing Record for Boueni (Index 90):\nThe closest hotel in Boueni is LE CHISSIOUA\n------------------------------------------------------------\nProcessing Record for Barkhan (Index 335):\nThe closest hotel in Barkhan is National Bank of Pakistan (NBP)\n------------------------------------------------------------\nProcessing Record for Riyadh (Index 339):\nThe closest hotel in Riyadh is Four Seasons Hotel Riyadh At Kingdom Center\n------------------------------------------------------------\nProcessing Record for Suzhou (Index 381):\nThe closest hotel in Suzhou is Suzhou Marriott Hotel\n------------------------------------------------------------\nProcessing Record for Algiers (Index 396):\nThe closest hotel in Algiers is Sofitel Algiers Hamma Garden\n------------------------------------------------------------\n------------------------------------------------------------\nData Retrieval Complete \n------------------------------------------------------------\n" ], [ "# NOTE: Do not change any of the code in this cell\n\n# Using the template add the hotel marks to the heatmap\ninfo_box_template = \"\"\"\n<dl>\n<dt>Name</dt><dd>{Hotel Name}</dd>\n<dt>City</dt><dd>{City}</dd>\n<dt>Country</dt><dd>{Country}</dd>\n</dl>\n\"\"\"\n# Store the DataFrame Row\n# NOTE: be sure to update with your DataFrame name\nhotel_info = [info_box_template.format(**row) for index, row in hotel_df.iterrows()]\nlocations = hotel_df[[\"Lat\", \"Lng\"]]", "_____no_output_____" ], [ "# Add marker layer ontop of heat map\nmarkers = gmaps.marker_layer(locations, info_box_content = hotel_info)\nfig.add_layer(markers)\n\n# Display figure\nfig\n", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ] ]

[ "Apache-2.0" ]

[ "Apache-2.0" ]