path
stringlengths 13
17
| screenshot_names
sequencelengths 1
873
| code
stringlengths 0
40.4k
| cell_type
stringclasses 1
value |
|---|---|---|---|
18140562/cell_2 | [
"text_plain_output_1.png"
] | import os
import os
print(os.listdir('../input')) | code |
18140562/cell_19 | [
"text_html_output_1.png"
] | from nltk.corpus import stopwords
from nltk.stem import WordNetLemmatizer
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.model_selection import cross_val_score, train_test_split
from sklearn.svm import LinearSVC
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import re
import seaborn as sns
from imblearn.over_sampling import SMOTE
from nltk.tokenize import word_tokenize
from nltk.corpus import stopwords
stop_words = set(stopwords.words('english'))
from sklearn.preprocessing import LabelEncoder
from xgboost import XGBClassifier
from nltk.stem import WordNetLemmatizer
lemmatizer = WordNetLemmatizer()
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.svm import LinearSVC
from sklearn.model_selection import cross_val_score, train_test_split
from sklearn.metrics import confusion_matrix, accuracy_score
df_train = pd.read_excel('../input/Data_Train.xlsx')
df_test = pd.read_excel('../input/Data_Test.xlsx')
df_train.sample(5)
df_train.isna().sum()
X = df_train['STORY']
y = df_train['SECTION']
tfidf_vec = TfidfVectorizer(ngram_range=(1, 2), stop_words=stop_words)
tfidf_vec.fit(X)
x_vec = tfidf_vec.transform(X)
svc = LinearSVC(C=10.0)
print(cross_val_score(svc, x_vec, y, cv=10, verbose=False).mean()) | code |
18140562/cell_1 | [
"application_vnd.jupyter.stderr_output_1.png"
] | from nltk.corpus import stopwords
from nltk.stem import WordNetLemmatizer
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import re
import seaborn as sns
from imblearn.over_sampling import SMOTE
from nltk.tokenize import word_tokenize
from nltk.corpus import stopwords
stop_words = set(stopwords.words('english'))
from sklearn.preprocessing import LabelEncoder
from xgboost import XGBClassifier
from nltk.stem import WordNetLemmatizer
lemmatizer = WordNetLemmatizer()
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.svm import LinearSVC
from sklearn.model_selection import cross_val_score, train_test_split
from sklearn.metrics import confusion_matrix, accuracy_score | code |
18140562/cell_7 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df_train = pd.read_excel('../input/Data_Train.xlsx')
df_test = pd.read_excel('../input/Data_Test.xlsx')
df_train.sample(5)
df_train.isna().sum()
plt.figure(figsize=(8, 5))
labels = list(set(df_train['SECTION']))
counts = []
for label in labels:
counts.append(np.count_nonzero(df_train['SECTION'] == label))
plt.pie(counts, labels=labels, autopct='%1.1f%%')
plt.show() | code |
18140562/cell_16 | [
"text_plain_output_1.png"
] | from imblearn.over_sampling import SMOTE
from nltk.corpus import stopwords
from nltk.stem import WordNetLemmatizer
from sklearn.feature_extraction.text import TfidfVectorizer
import matplotlib.pyplot as plt
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import re
import seaborn as sns
from imblearn.over_sampling import SMOTE
from nltk.tokenize import word_tokenize
from nltk.corpus import stopwords
stop_words = set(stopwords.words('english'))
from sklearn.preprocessing import LabelEncoder
from xgboost import XGBClassifier
from nltk.stem import WordNetLemmatizer
lemmatizer = WordNetLemmatizer()
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.svm import LinearSVC
from sklearn.model_selection import cross_val_score, train_test_split
from sklearn.metrics import confusion_matrix, accuracy_score
df_train = pd.read_excel('../input/Data_Train.xlsx')
df_test = pd.read_excel('../input/Data_Test.xlsx')
df_train.sample(5)
df_train.isna().sum()
labels = list(set(df_train['SECTION']))
counts = []
for label in labels:
counts.append(np.count_nonzero(df_train['SECTION'] == label))
X = df_train['STORY']
y = df_train['SECTION']
tfidf_vec = TfidfVectorizer(ngram_range=(1, 2), stop_words=stop_words)
tfidf_vec.fit(X)
x_vec = tfidf_vec.transform(X)
sm = SMOTE(random_state=42)
x_vec, y = sm.fit_sample(x_vec, y)
labels = list(set(y))
counts = []
for label in labels:
counts.append(np.count_nonzero(y == label))
plt.pie(counts, labels=labels, autopct='%1.1f%%')
plt.show() | code |
18140562/cell_22 | [
"image_output_1.png"
] | from nltk.corpus import stopwords
from nltk.stem import WordNetLemmatizer
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics import confusion_matrix, accuracy_score
from sklearn.model_selection import cross_val_score, train_test_split
from sklearn.svm import LinearSVC
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import re
import seaborn as sns
from imblearn.over_sampling import SMOTE
from nltk.tokenize import word_tokenize
from nltk.corpus import stopwords
stop_words = set(stopwords.words('english'))
from sklearn.preprocessing import LabelEncoder
from xgboost import XGBClassifier
from nltk.stem import WordNetLemmatizer
lemmatizer = WordNetLemmatizer()
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.svm import LinearSVC
from sklearn.model_selection import cross_val_score, train_test_split
from sklearn.metrics import confusion_matrix, accuracy_score
df_train = pd.read_excel('../input/Data_Train.xlsx')
df_test = pd.read_excel('../input/Data_Test.xlsx')
df_train.sample(5)
df_train.isna().sum()
X = df_train['STORY']
y = df_train['SECTION']
tfidf_vec = TfidfVectorizer(ngram_range=(1, 2), stop_words=stop_words)
tfidf_vec.fit(X)
x_vec = tfidf_vec.transform(X)
svc = LinearSVC(C=10.0)
svc = LinearSVC(C=10.0)
svc.fit(x_train_vec, y_train)
y_preds = svc.predict(x_eval_vec)
confusion_matrix(y_eval, y_preds) | code |
89142870/cell_6 | [
"text_plain_output_1.png"
] | from pytorch_lightning.callbacks import Callback
from pytorch_lightning.callbacks.early_stopping import EarlyStopping
from pytorch_lightning.callbacks.model_checkpoint import ModelCheckpoint
from sklearn.model_selection import KFold
from torch.utils.data import DataLoader, TensorDataset, Subset
import gc
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import pytorch_lightning as pl
import torch
import torch.nn as nn
import torch.nn.functional as F
train = pd.read_parquet('../input/train-small/train_small.parquet')
float_feature_names = train.drop(['target', 'row_id', 'time_id', 'investment_id'], axis=1).columns
float_input = train[float_feature_names].values
investment_id = train[['investment_id']].values.astype(int)
time_id = train[['time_id']].values.astype(int)
targets = train[['target']].values
del train
float_input = torch.FloatTensor(float_input)
investment_id = torch.LongTensor(investment_id)
time_id = torch.LongTensor(time_id)
target = torch.FloatTensor(targets)
dataset = TensorDataset(time_id, investment_id, float_input, target)
gc.collect()
class UbiquantRegressor(pl.LightningModule):
def __init__(self):
super(UbiquantRegressor, self).__init__()
# Embedding of investment_id to 11 float features.
# As the number of unseen investment_ids is unknown, a large margin is selected (10000).
self.id_embedding = nn.Embedding(10000,11)
# credits to sahil112: https://www.kaggle.com/sahil112/whyonlykeras-easy-pytorch-competitive-dnn for this architecture
self.layers = nn.Sequential(nn.Linear(311, 64),
nn.BatchNorm1d(64),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(64, 128),
nn.BatchNorm1d(128),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(128, 256),
nn.BatchNorm1d(256),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(256, 512),
nn.BatchNorm1d(512),
nn.SiLU(0.1),
nn.Dropout(0.4),
nn.Linear(512, 256),
nn.BatchNorm1d(256),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(256, 128),
nn.BatchNorm1d(128),
nn.SiLU(0.1),
nn.Dropout(0.4),
nn.Linear(128, 8),
nn.BatchNorm1d(8),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(8, 1))
def forward(self, time_id, investment_id, f_features):
# Embedding of the investment_id
invest_embedding = self.id_embedding(investment_id).squeeze(dim=1)
# Concat embedding and features.
# Open question: should the network have access to the time_id?
# The final test set will consist of time_id never seen in the train set
# Nevertheless, it can be easily added here...
#dnn_input = torch.cat((invest_embedding, time_id, f_features), axis=-1)
dnn_input = torch.cat((invest_embedding, f_features), axis=-1)
return self.layers(dnn_input)
def training_step(self, batch, batch_nb):
time_id, investment_id, float_input, target = batch
out = self(time_id, investment_id, float_input)
loss = F.mse_loss(out, target)
self.log('train_loss', loss)
return loss
def validation_step(self, batch, batch_nb):
time_id, investment_id, float_input, target = batch
result = self(time_id, investment_id, float_input)
loss = F.mse_loss(result, target)
dict = {'val_loss': loss,
'result': result,
'target': target,
'time_id': time_id,
'investment_id': investment_id,
}
return dict
def validation_epoch_end(self, outputs):
val_losses = [x['val_loss'] for x in outputs]
result = torch.cat([x['result'] for x in outputs])
target = torch.cat([x['target'] for x in outputs])
time_ids = torch.cat([x['time_id'] for x in outputs])
investment_ids = torch.cat([x['investment_id'] for x in outputs])
corrs = []
for t in torch.unique(time_ids):
t_results = result[time_ids == t]
t_target = target[time_ids == t]
# corr = torch.corrcoef(torch.stack((t_results, t_target)))[0,1] # use this when pytorch>=1.10
corr = np.corrcoef(torch.stack((t_results, t_target)).cpu().numpy())[0, 1]
corrs.append(corr)
# mean_corr = torch.mean(torch.stack(corrs)) # use this when pytorch>=1.10
mean_corr = np.nanmean(corrs)
epoch_loss = torch.stack(val_losses).mean() # Combine losses
self.log('val_loss', epoch_loss, prog_bar=True)
self.log('mean_corr', mean_corr, prog_bar=True)
dict = {'val_loss': epoch_loss,
'corrs': mean_corr}
return dict
def epoch_end(self, epoch, result):
pass
def test_step(self, batch, batch_nb):
pass
def configure_optimizers(self):
return torch.optim.Adam(self.parameters(), lr=0.001)
class MetricTracker(Callback):
def __init__(self):
self.val_losses = []
self.corrs = []
def on_validation_epoch_end(self, trainer, module):
self.val_losses.append(trainer._results['validation_epoch_end.val_loss'].value.cpu().numpy()) # track them
self.corrs.append(trainer._results['validation_epoch_end.mean_corr'].value.cpu().numpy()) # track them
if 0: #index==1:
# live plotting of results during training, switched off
ax.plot(self.val_losses, color="orange")
ax.set_ylabel("Val loss", color="orange", fontsize=14)
ax2 = ax.twinx()
ax2.plot(self.corrs, color="blue")
ax2.set_ylabel("Mean daily corr 2 target", color="blue", fontsize=14)
plt.show()
n_splits=10
kf = KFold(n_splits=n_splits, shuffle=True)
val_losses = []
mean_corrs = []
models = [] # A list of all final models
index=0
for train_index, test_index in kf.split(dataset):
index+=1
print("CV run {}...".format(index))
train_ds, val_ds = Subset(dataset, train_index), Subset(dataset, test_index)
train_loader = DataLoader(train_ds, 32768)
val_loader = DataLoader(val_ds, 32768)
uq_regressor = UbiquantRegressor()
metricTracker = MetricTracker()
trainer = pl.Trainer(gpus=1,
callbacks=[metricTracker,
EarlyStopping(monitor="mean_corr", mode="max", patience=3),
ModelCheckpoint(save_top_k=1, monitor="mean_corr", mode="max", save_on_train_epoch_end=False)],
max_epochs=21,
num_sanity_val_steps=0,)
trainer.fit(uq_regressor, train_loader, val_loader)
# Load best model based on mean daily correlation with target
uq_regressor = UbiquantRegressor().load_from_checkpoint(trainer.checkpoint_callback.best_model_path)
uq_regressor.eval()
models.append(uq_regressor)
# Show val results
val_result = trainer.validate(model=uq_regressor, dataloaders=val_loader)
val_losses.append(val_result[0]['val_loss'])
mean_corrs.append(val_result[0]['mean_corr'])
fig, ax = plt.subplots()
ax.plot(metricTracker.val_losses, color="orange")
ax.set_ylabel("Val loss", color="orange", fontsize=14)
ax2 = ax.twinx()
ax2.plot(metricTracker.corrs, color="blue")
ax2.set_ylabel("Mean daily corr 2 target", color="blue", fontsize=14)
plt.show()
val_result = trainer.validate(model=uq_regressor, dataloaders=val_loader) | code |
89142870/cell_8 | [
"application_vnd.jupyter.stderr_output_2.png",
"text_html_output_2.png",
"text_html_output_1.png",
"text_plain_output_1.png",
"text_html_output_3.png"
] | from pytorch_lightning.callbacks import Callback
from pytorch_lightning.callbacks.early_stopping import EarlyStopping
from pytorch_lightning.callbacks.model_checkpoint import ModelCheckpoint
from sklearn.model_selection import KFold
from torch.utils.data import DataLoader, TensorDataset, Subset
import gc
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import pytorch_lightning as pl
import torch
import torch.nn as nn
import torch.nn.functional as F
import ubiquant
train = pd.read_parquet('../input/train-small/train_small.parquet')
float_feature_names = train.drop(['target', 'row_id', 'time_id', 'investment_id'], axis=1).columns
float_input = train[float_feature_names].values
investment_id = train[['investment_id']].values.astype(int)
time_id = train[['time_id']].values.astype(int)
targets = train[['target']].values
del train
float_input = torch.FloatTensor(float_input)
investment_id = torch.LongTensor(investment_id)
time_id = torch.LongTensor(time_id)
target = torch.FloatTensor(targets)
dataset = TensorDataset(time_id, investment_id, float_input, target)
gc.collect()
class UbiquantRegressor(pl.LightningModule):
def __init__(self):
super(UbiquantRegressor, self).__init__()
# Embedding of investment_id to 11 float features.
# As the number of unseen investment_ids is unknown, a large margin is selected (10000).
self.id_embedding = nn.Embedding(10000,11)
# credits to sahil112: https://www.kaggle.com/sahil112/whyonlykeras-easy-pytorch-competitive-dnn for this architecture
self.layers = nn.Sequential(nn.Linear(311, 64),
nn.BatchNorm1d(64),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(64, 128),
nn.BatchNorm1d(128),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(128, 256),
nn.BatchNorm1d(256),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(256, 512),
nn.BatchNorm1d(512),
nn.SiLU(0.1),
nn.Dropout(0.4),
nn.Linear(512, 256),
nn.BatchNorm1d(256),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(256, 128),
nn.BatchNorm1d(128),
nn.SiLU(0.1),
nn.Dropout(0.4),
nn.Linear(128, 8),
nn.BatchNorm1d(8),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(8, 1))
def forward(self, time_id, investment_id, f_features):
# Embedding of the investment_id
invest_embedding = self.id_embedding(investment_id).squeeze(dim=1)
# Concat embedding and features.
# Open question: should the network have access to the time_id?
# The final test set will consist of time_id never seen in the train set
# Nevertheless, it can be easily added here...
#dnn_input = torch.cat((invest_embedding, time_id, f_features), axis=-1)
dnn_input = torch.cat((invest_embedding, f_features), axis=-1)
return self.layers(dnn_input)
def training_step(self, batch, batch_nb):
time_id, investment_id, float_input, target = batch
out = self(time_id, investment_id, float_input)
loss = F.mse_loss(out, target)
self.log('train_loss', loss)
return loss
def validation_step(self, batch, batch_nb):
time_id, investment_id, float_input, target = batch
result = self(time_id, investment_id, float_input)
loss = F.mse_loss(result, target)
dict = {'val_loss': loss,
'result': result,
'target': target,
'time_id': time_id,
'investment_id': investment_id,
}
return dict
def validation_epoch_end(self, outputs):
val_losses = [x['val_loss'] for x in outputs]
result = torch.cat([x['result'] for x in outputs])
target = torch.cat([x['target'] for x in outputs])
time_ids = torch.cat([x['time_id'] for x in outputs])
investment_ids = torch.cat([x['investment_id'] for x in outputs])
corrs = []
for t in torch.unique(time_ids):
t_results = result[time_ids == t]
t_target = target[time_ids == t]
# corr = torch.corrcoef(torch.stack((t_results, t_target)))[0,1] # use this when pytorch>=1.10
corr = np.corrcoef(torch.stack((t_results, t_target)).cpu().numpy())[0, 1]
corrs.append(corr)
# mean_corr = torch.mean(torch.stack(corrs)) # use this when pytorch>=1.10
mean_corr = np.nanmean(corrs)
epoch_loss = torch.stack(val_losses).mean() # Combine losses
self.log('val_loss', epoch_loss, prog_bar=True)
self.log('mean_corr', mean_corr, prog_bar=True)
dict = {'val_loss': epoch_loss,
'corrs': mean_corr}
return dict
def epoch_end(self, epoch, result):
pass
def test_step(self, batch, batch_nb):
pass
def configure_optimizers(self):
return torch.optim.Adam(self.parameters(), lr=0.001)
class MetricTracker(Callback):
def __init__(self):
self.val_losses = []
self.corrs = []
def on_validation_epoch_end(self, trainer, module):
self.val_losses.append(trainer._results['validation_epoch_end.val_loss'].value.cpu().numpy()) # track them
self.corrs.append(trainer._results['validation_epoch_end.mean_corr'].value.cpu().numpy()) # track them
if 0: #index==1:
# live plotting of results during training, switched off
ax.plot(self.val_losses, color="orange")
ax.set_ylabel("Val loss", color="orange", fontsize=14)
ax2 = ax.twinx()
ax2.plot(self.corrs, color="blue")
ax2.set_ylabel("Mean daily corr 2 target", color="blue", fontsize=14)
plt.show()
n_splits=10
kf = KFold(n_splits=n_splits, shuffle=True)
val_losses = []
mean_corrs = []
models = [] # A list of all final models
index=0
for train_index, test_index in kf.split(dataset):
index+=1
print("CV run {}...".format(index))
train_ds, val_ds = Subset(dataset, train_index), Subset(dataset, test_index)
train_loader = DataLoader(train_ds, 32768)
val_loader = DataLoader(val_ds, 32768)
uq_regressor = UbiquantRegressor()
metricTracker = MetricTracker()
trainer = pl.Trainer(gpus=1,
callbacks=[metricTracker,
EarlyStopping(monitor="mean_corr", mode="max", patience=3),
ModelCheckpoint(save_top_k=1, monitor="mean_corr", mode="max", save_on_train_epoch_end=False)],
max_epochs=21,
num_sanity_val_steps=0,)
trainer.fit(uq_regressor, train_loader, val_loader)
# Load best model based on mean daily correlation with target
uq_regressor = UbiquantRegressor().load_from_checkpoint(trainer.checkpoint_callback.best_model_path)
uq_regressor.eval()
models.append(uq_regressor)
# Show val results
val_result = trainer.validate(model=uq_regressor, dataloaders=val_loader)
val_losses.append(val_result[0]['val_loss'])
mean_corrs.append(val_result[0]['mean_corr'])
fig, ax = plt.subplots()
ax.plot(metricTracker.val_losses, color="orange")
ax.set_ylabel("Val loss", color="orange", fontsize=14)
ax2 = ax.twinx()
ax2.plot(metricTracker.corrs, color="blue")
ax2.set_ylabel("Mean daily corr 2 target", color="blue", fontsize=14)
plt.show()
import ubiquant
env = ubiquant.make_env()
iter_test = env.iter_test()
for test_df, sample_prediction_df in iter_test:
time_id = test_df.row_id.str.split('_', expand=True)[0].values.astype(int)
investment_id = test_df[['investment_id']].values.astype(int)
float_input = test_df[float_feature_names].values
float_input = torch.FloatTensor(float_input)
investment_id = torch.LongTensor(investment_id)
time_id = torch.LongTensor(time_id).unsqueeze(-1)
sample_prediction_df['target'] = 0
for uq_regressor in models:
predictions = uq_regressor(time_id, investment_id, float_input).squeeze()
sample_prediction_df['target'] += predictions.detach().cpu().numpy() / n_splits
env.predict(sample_prediction_df)
display(sample_prediction_df) | code |
89142870/cell_3 | [
"text_plain_output_1.png"
] | from torch.utils.data import DataLoader, TensorDataset, Subset
import gc
import pandas as pd
import torch
train = pd.read_parquet('../input/train-small/train_small.parquet')
float_feature_names = train.drop(['target', 'row_id', 'time_id', 'investment_id'], axis=1).columns
float_input = train[float_feature_names].values
investment_id = train[['investment_id']].values.astype(int)
time_id = train[['time_id']].values.astype(int)
targets = train[['target']].values
del train
float_input = torch.FloatTensor(float_input)
investment_id = torch.LongTensor(investment_id)
time_id = torch.LongTensor(time_id)
target = torch.FloatTensor(targets)
dataset = TensorDataset(time_id, investment_id, float_input, target)
gc.collect() | code |
89142870/cell_5 | [
"text_plain_output_5.png",
"text_plain_output_15.png",
"text_plain_output_9.png",
"text_plain_output_20.png",
"text_plain_output_4.png",
"text_plain_output_13.png",
"image_output_5.png",
"text_plain_output_14.png",
"text_plain_output_10.png",
"text_plain_output_6.png",
"image_output_7.png",
"text_plain_output_18.png",
"text_plain_output_3.png",
"image_output_4.png",
"text_plain_output_7.png",
"image_output_8.png",
"text_plain_output_16.png",
"text_plain_output_8.png",
"image_output_6.png",
"text_plain_output_2.png",
"text_plain_output_1.png",
"image_output_3.png",
"text_plain_output_19.png",
"image_output_2.png",
"image_output_1.png",
"image_output_10.png",
"text_plain_output_17.png",
"text_plain_output_11.png",
"text_plain_output_12.png",
"image_output_9.png"
] | from pytorch_lightning.callbacks import Callback
from pytorch_lightning.callbacks.early_stopping import EarlyStopping
from pytorch_lightning.callbacks.model_checkpoint import ModelCheckpoint
from sklearn.model_selection import KFold
from torch.utils.data import DataLoader, TensorDataset, Subset
import gc
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import pytorch_lightning as pl
import torch
import torch.nn as nn
import torch.nn.functional as F
train = pd.read_parquet('../input/train-small/train_small.parquet')
float_feature_names = train.drop(['target', 'row_id', 'time_id', 'investment_id'], axis=1).columns
float_input = train[float_feature_names].values
investment_id = train[['investment_id']].values.astype(int)
time_id = train[['time_id']].values.astype(int)
targets = train[['target']].values
del train
float_input = torch.FloatTensor(float_input)
investment_id = torch.LongTensor(investment_id)
time_id = torch.LongTensor(time_id)
target = torch.FloatTensor(targets)
dataset = TensorDataset(time_id, investment_id, float_input, target)
gc.collect()
class UbiquantRegressor(pl.LightningModule):
def __init__(self):
super(UbiquantRegressor, self).__init__()
# Embedding of investment_id to 11 float features.
# As the number of unseen investment_ids is unknown, a large margin is selected (10000).
self.id_embedding = nn.Embedding(10000,11)
# credits to sahil112: https://www.kaggle.com/sahil112/whyonlykeras-easy-pytorch-competitive-dnn for this architecture
self.layers = nn.Sequential(nn.Linear(311, 64),
nn.BatchNorm1d(64),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(64, 128),
nn.BatchNorm1d(128),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(128, 256),
nn.BatchNorm1d(256),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(256, 512),
nn.BatchNorm1d(512),
nn.SiLU(0.1),
nn.Dropout(0.4),
nn.Linear(512, 256),
nn.BatchNorm1d(256),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(256, 128),
nn.BatchNorm1d(128),
nn.SiLU(0.1),
nn.Dropout(0.4),
nn.Linear(128, 8),
nn.BatchNorm1d(8),
nn.SiLU(),
nn.Dropout(0.4),
nn.Linear(8, 1))
def forward(self, time_id, investment_id, f_features):
# Embedding of the investment_id
invest_embedding = self.id_embedding(investment_id).squeeze(dim=1)
# Concat embedding and features.
# Open question: should the network have access to the time_id?
# The final test set will consist of time_id never seen in the train set
# Nevertheless, it can be easily added here...
#dnn_input = torch.cat((invest_embedding, time_id, f_features), axis=-1)
dnn_input = torch.cat((invest_embedding, f_features), axis=-1)
return self.layers(dnn_input)
def training_step(self, batch, batch_nb):
time_id, investment_id, float_input, target = batch
out = self(time_id, investment_id, float_input)
loss = F.mse_loss(out, target)
self.log('train_loss', loss)
return loss
def validation_step(self, batch, batch_nb):
time_id, investment_id, float_input, target = batch
result = self(time_id, investment_id, float_input)
loss = F.mse_loss(result, target)
dict = {'val_loss': loss,
'result': result,
'target': target,
'time_id': time_id,
'investment_id': investment_id,
}
return dict
def validation_epoch_end(self, outputs):
val_losses = [x['val_loss'] for x in outputs]
result = torch.cat([x['result'] for x in outputs])
target = torch.cat([x['target'] for x in outputs])
time_ids = torch.cat([x['time_id'] for x in outputs])
investment_ids = torch.cat([x['investment_id'] for x in outputs])
corrs = []
for t in torch.unique(time_ids):
t_results = result[time_ids == t]
t_target = target[time_ids == t]
# corr = torch.corrcoef(torch.stack((t_results, t_target)))[0,1] # use this when pytorch>=1.10
corr = np.corrcoef(torch.stack((t_results, t_target)).cpu().numpy())[0, 1]
corrs.append(corr)
# mean_corr = torch.mean(torch.stack(corrs)) # use this when pytorch>=1.10
mean_corr = np.nanmean(corrs)
epoch_loss = torch.stack(val_losses).mean() # Combine losses
self.log('val_loss', epoch_loss, prog_bar=True)
self.log('mean_corr', mean_corr, prog_bar=True)
dict = {'val_loss': epoch_loss,
'corrs': mean_corr}
return dict
def epoch_end(self, epoch, result):
pass
def test_step(self, batch, batch_nb):
pass
def configure_optimizers(self):
return torch.optim.Adam(self.parameters(), lr=0.001)
class MetricTracker(Callback):
def __init__(self):
self.val_losses = []
self.corrs = []
def on_validation_epoch_end(self, trainer, module):
self.val_losses.append(trainer._results['validation_epoch_end.val_loss'].value.cpu().numpy()) # track them
self.corrs.append(trainer._results['validation_epoch_end.mean_corr'].value.cpu().numpy()) # track them
if 0: #index==1:
# live plotting of results during training, switched off
ax.plot(self.val_losses, color="orange")
ax.set_ylabel("Val loss", color="orange", fontsize=14)
ax2 = ax.twinx()
ax2.plot(self.corrs, color="blue")
ax2.set_ylabel("Mean daily corr 2 target", color="blue", fontsize=14)
plt.show()
n_splits = 10
kf = KFold(n_splits=n_splits, shuffle=True)
val_losses = []
mean_corrs = []
models = []
index = 0
for train_index, test_index in kf.split(dataset):
index += 1
print('CV run {}...'.format(index))
train_ds, val_ds = (Subset(dataset, train_index), Subset(dataset, test_index))
train_loader = DataLoader(train_ds, 32768)
val_loader = DataLoader(val_ds, 32768)
uq_regressor = UbiquantRegressor()
metricTracker = MetricTracker()
trainer = pl.Trainer(gpus=1, callbacks=[metricTracker, EarlyStopping(monitor='mean_corr', mode='max', patience=3), ModelCheckpoint(save_top_k=1, monitor='mean_corr', mode='max', save_on_train_epoch_end=False)], max_epochs=21, num_sanity_val_steps=0)
trainer.fit(uq_regressor, train_loader, val_loader)
uq_regressor = UbiquantRegressor().load_from_checkpoint(trainer.checkpoint_callback.best_model_path)
uq_regressor.eval()
models.append(uq_regressor)
val_result = trainer.validate(model=uq_regressor, dataloaders=val_loader)
val_losses.append(val_result[0]['val_loss'])
mean_corrs.append(val_result[0]['mean_corr'])
fig, ax = plt.subplots()
ax.plot(metricTracker.val_losses, color='orange')
ax.set_ylabel('Val loss', color='orange', fontsize=14)
ax2 = ax.twinx()
ax2.plot(metricTracker.corrs, color='blue')
ax2.set_ylabel('Mean daily corr 2 target', color='blue', fontsize=14)
plt.show() | code |
90106156/cell_21 | [
"text_html_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
data[data['from_lat'] == data['from_lat'].median()].shape | code |
90106156/cell_9 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.info() | code |
90106156/cell_30 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
data[data['to_lat'].isnull()]['from_area_id'].value_counts() | code |
90106156/cell_20 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
data[data['from_lat'] == data['from_lat'].median()]['from_area_id'] | code |
90106156/cell_6 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.head() | code |
90106156/cell_29 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
data[data['to_lat'].isnull()].head() | code |
90106156/cell_11 | [
"text_html_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean() | code |
90106156/cell_19 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
print(data[data['from_lat'] == data['from_lat'].median()]['from_area_id'].max())
print(data['from_lat'].median())
print(data['from_long'].median()) | code |
90106156/cell_1 | [
"text_plain_output_1.png"
] | import os
import numpy as np
import pandas as pd
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
90106156/cell_7 | [
"text_html_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape | code |
90106156/cell_32 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
data.info() | code |
90106156/cell_28 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
data.info() | code |
90106156/cell_15 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data['vehicle_model_id'].value_counts(normalize=True) * 100 | code |
90106156/cell_17 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
data.head() | code |
90106156/cell_46 | [
"text_html_output_1.png"
] | from geopy import distance
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
traveltype = pd.get_dummies(data['travel_type_id'], drop_first=True)
data = pd.concat([data, traveltype], axis=1)
data = data.drop(['travel_type_id'], axis=1)
data.rename(columns={2: 'traveltype_pointtopoint', 3: 'traveltype_hourly'}, inplace=True)
def cal_distance(from_lat, from_long, to_lat, to_long):
return distance.distance((from_lat, from_long), (to_lat, to_long)).km
data['distance'] = data.apply(lambda row: cal_distance(row['from_lat'], row['from_long'], row['to_lat'], row['to_long']), axis=1)
data = data.drop(['from_lat', 'from_long', 'to_lat', 'to_long'], axis=1)
data[data['time_diff'] < 0].head() | code |
90106156/cell_24 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape
data = data.drop(['vehicle_model_id'], axis=1)
data.info() | code |
90106156/cell_14 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('/kaggle/input/yourcabs/YourCabs_training.csv')
df.shape
data = df.drop(['Car_Cancellation', 'Cost_of_error'], axis=1)
target = df[['Car_Cancellation']]
data.isnull().mean()
data = data.drop(['id', 'user_id', 'package_id', 'to_area_id', 'from_city_id', 'to_city_id', 'to_date'], axis=1)
data.shape | code |
32069353/cell_2 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import os
import warnings
import math
import numpy as np
import pandas as pd
from tqdm import tqdm
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
import warnings
warnings.filterwarnings('ignore')
PUBLIC_PRIVATE = 1
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
32069353/cell_7 | [
"application_vnd.jupyter.stderr_output_1.png"
] | from datetime import datetime
from datetime import datetime, timedelta
from scipy.optimize.minpack import curve_fit
from tqdm import tqdm
import math
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import time
import warnings
import math
import numpy as np
import pandas as pd
from tqdm import tqdm
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
import warnings
warnings.filterwarnings('ignore')
PUBLIC_PRIVATE = 1
import os
df_train = pd.read_csv('/kaggle/input/covid19-global-forecasting-week-4/train.csv')
df_test = pd.read_csv('/kaggle/input/covid19-global-forecasting-week-4/test.csv')
sub_example = pd.read_csv('/kaggle/input/covid19-global-forecasting-week-4/submission.csv')
regr_ensemble_sub = pd.read_csv('/kaggle/input/regressors-ensemble-submission-week-4/submission (4).csv')
df_train['ForecastId'] = -1
df_train['DataType'] = 0
df_test['DataType'] = 2
df_test['ConfirmedCases'] = -1
df_test['Fatalities'] = -1
df_test['Id'] = df_test['ForecastId'] + df_train['Id'].max()
df_intersection = df_train[df_train['Date'] >= df_test['Date'].min()]
df_intersection['DataType'] = 1
df_intersection['ForecastId'] = df_test[df_test['Date'] <= df_train['Date'].max()]['ForecastId'].values
df_train = df_train[df_train['Date'] < df_test['Date'].min()]
df_test = df_test[df_test['Date'] > df_intersection['Date'].max()]
if not PUBLIC_PRIVATE:
df_intersection['ConfirmedCases'] = -1
df_intersection['Fatalities'] = -1
df_full = pd.concat([df_train, df_intersection, df_test], sort=False, axis=0)
else:
df_full = pd.concat([df_train, df_intersection, df_test], sort=False, axis=0)
df_full['Province_State'] = df_full['Province_State'].fillna('')
df_full['Location'] = ['_'.join(x) for x in zip(df_full['Country_Region'], df_full['Province_State'])]
df_full.drop(columns=['Province_State', 'Country_Region'], inplace=True)
df_full.shape
import time
from datetime import datetime
df_full['Date'] = pd.to_datetime(df_full['Date'])
df_full['Date'] = df_full['Date'].apply(lambda s: time.mktime(s.timetuple()))
min_timestamp = np.min(df_full['Date'])
df_full['Date'] = df_full['Date'].apply(lambda s: (s - min_timestamp) / 86400.0)
import random
from scipy.optimize.minpack import curve_fit
from sklearn.metrics import r2_score
from scipy.special import expit
def Gompertz(a, c, t, t0):
Q = a * np.exp(-np.exp(-c * (t - t0)))
return Q
ref = df_full[df_full['DataType'].isin([0, 1]) & (df_full['Location'] == 'China_Anhui')]
ref['ConfirmedCases'] /= ref['ConfirmedCases'].max()
ref['Fatalities'] /= ref['Fatalities'].max()
def getMultiplier(date):
try:
return 1.0 / ref[ref['Date'] == date]['ConfirmedCases'].values[0]
except:
return 3.5
locations = list(set(df_full['Location']))
location_sample = ['Brazil_']
train = df_full[df_full['DataType'] == 0]
valid = df_full[df_full['DataType'] == 1]
test = df_full[df_full['DataType'] == 2]
for location in tqdm(locations):
_train = train[train['Location'] == location]
_valid = valid[valid['Location'] == location]
_test = test[test['Location'] == location]
n_train_days = _train.Date.nunique()
n_valid_days = _valid.Date.nunique()
n_test_days = _test.Date.nunique()
x_train = range(n_train_days)
x_test = range(n_train_days + n_valid_days + n_test_days + 200)
y_train_f = _train['Fatalities']
y_train_c = _train['ConfirmedCases']
the_first_one = _train[_train['ConfirmedCases'] > 0.05 * _train['ConfirmedCases'].max()]['Date'].min()
first_cases = _train[_train['ConfirmedCases'] > 0.4 * _train['ConfirmedCases'].max()]['Date'].min()
if math.isnan(the_first_one):
the_first_one = _train['Date'].max()
first_cases = _train['Date'].max() + 8
current = _train['Date'].max() - first_cases
if location.startswith('China'):
lower_c = [0, 0.02, 0]
upper_c = [2 * y_train_c.max() + 1, 0.15, 25]
else:
lower_c = [0, 0.02, the_first_one]
upper_c = [getMultiplier(current) * np.max(y_train_c) + 1, 0.15, first_cases + 28]
popt_c, pcov_c = curve_fit(Gompertz, x_train, y_train_c, method='trf', bounds=(lower_c, upper_c))
a_max_c, estimated_c_c, estimated_t0_c = popt_c
y_predict_c = Gompertz(a_max_c, estimated_c_c, x_test, estimated_t0_c)
y_predict_c_at_t0 = Gompertz(a_max_c, estimated_c_c, estimated_t0_c, estimated_t0_c)
the_first_one = _train[_train['Fatalities'] > 0.05 * _train['Fatalities'].max()]['Date'].min()
first_cases = _train[_train['Fatalities'] > 0.37 * _train['Fatalities'].max()]['Date'].min()
if math.isnan(the_first_one):
the_first_one = _train['Date'].max()
first_cases = _train['Date'].max() + 8
current = _train['Date'].max() - first_cases
if location.startswith('China'):
lower = [0, 0.02, 0]
upper = [2 * y_train_f.max() + 1, 0.15, 25]
else:
lower = [0, 0.02, the_first_one]
upper = [getMultiplier(current) * np.max(y_train_f) + 1, 0.15, first_cases + 28]
popt_f, pcov_f = curve_fit(Gompertz, x_train, y_train_f, method='trf', bounds=(lower, upper))
a_max, estimated_c, estimated_t0 = popt_f
y_predict_f = Gompertz(a_max, estimated_c, x_test, estimated_t0)
y_predict_f_at_t0 = Gompertz(a_max, estimated_c, estimated_t0, estimated_t0)
from datetime import datetime, timedelta
initial_date = datetime(2020, 1, 22)
dates_train = list(x_train)
dates_test = list(x_test)
for i in range(len(dates_train)):
dates_train[i] = initial_date + timedelta(days=dates_train[i])
for i in range(len(dates_test)):
dates_test[i] = initial_date + timedelta(days=dates_test[i])
values = y_predict_c[df_full[df_full['DataType'] == 2]['Date'].astype(int).min():df_full[df_full['DataType'] == 2]['Date'].astype(int).max() + 1]
df_full.loc[(df_full['DataType'] == 2) & (df_full['Location'] == location), 'ConfirmedCases'] = values
values = y_predict_f[df_full[df_full['DataType'] == 2]['Date'].astype(int).min():df_full[df_full['DataType'] == 2]['Date'].astype(int).max() + 1]
df_full.loc[(df_full['DataType'] == 2) & (df_full['Location'] == location), 'Fatalities'] = values | code |
32069353/cell_3 | [
"text_html_output_1.png"
] | import pandas as pd
df_train = pd.read_csv('/kaggle/input/covid19-global-forecasting-week-4/train.csv')
df_test = pd.read_csv('/kaggle/input/covid19-global-forecasting-week-4/test.csv')
sub_example = pd.read_csv('/kaggle/input/covid19-global-forecasting-week-4/submission.csv')
regr_ensemble_sub = pd.read_csv('/kaggle/input/regressors-ensemble-submission-week-4/submission (4).csv')
regr_ensemble_sub.head() | code |
32069353/cell_5 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import os
import pandas as pd
import warnings
import math
import numpy as np
import pandas as pd
from tqdm import tqdm
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
import warnings
warnings.filterwarnings('ignore')
PUBLIC_PRIVATE = 1
import os
df_train = pd.read_csv('/kaggle/input/covid19-global-forecasting-week-4/train.csv')
df_test = pd.read_csv('/kaggle/input/covid19-global-forecasting-week-4/test.csv')
sub_example = pd.read_csv('/kaggle/input/covid19-global-forecasting-week-4/submission.csv')
regr_ensemble_sub = pd.read_csv('/kaggle/input/regressors-ensemble-submission-week-4/submission (4).csv')
df_train['ForecastId'] = -1
df_train['DataType'] = 0
df_test['DataType'] = 2
df_test['ConfirmedCases'] = -1
df_test['Fatalities'] = -1
df_test['Id'] = df_test['ForecastId'] + df_train['Id'].max()
df_intersection = df_train[df_train['Date'] >= df_test['Date'].min()]
df_intersection['DataType'] = 1
df_intersection['ForecastId'] = df_test[df_test['Date'] <= df_train['Date'].max()]['ForecastId'].values
df_train = df_train[df_train['Date'] < df_test['Date'].min()]
df_test = df_test[df_test['Date'] > df_intersection['Date'].max()]
if not PUBLIC_PRIVATE:
df_intersection['ConfirmedCases'] = -1
df_intersection['Fatalities'] = -1
df_full = pd.concat([df_train, df_intersection, df_test], sort=False, axis=0)
else:
df_full = pd.concat([df_train, df_intersection, df_test], sort=False, axis=0)
df_full['Province_State'] = df_full['Province_State'].fillna('')
df_full['Location'] = ['_'.join(x) for x in zip(df_full['Country_Region'], df_full['Province_State'])]
df_full.drop(columns=['Province_State', 'Country_Region'], inplace=True)
df_full.shape | code |
18114340/cell_6 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import *
from sklearn.metrics import *
regr = LinearRegression()
regr.fit(X_train, y_train)
pred = regr.predict(X_test)
mean_squared_error(pred, y_test) | code |
18114340/cell_1 | [
"application_vnd.jupyter.stderr_output_1.png"
] | import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import os
from keras.optimizers import *
import keras
from keras.layers import *
from keras.models import *
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import * | code |
18114340/cell_8 | [
"text_html_output_1.png"
] | from sklearn.model_selection import train_test_split
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
train_df = pd.read_csv('../input/train.csv')
train_df['Test'] = False
test_df = pd.read_csv('../input/test.csv')
test_df['Test'] = True
df = pd.concat([train_df, test_df], sort=False)
corr = abs(train_df.corr())
price_corr = corr['SalePrice'].sort_values()
columns = list(price_corr[price_corr > 0.51].index)
columns.append('Test')
columns.append('Id')
prepared_data = df[columns].copy()
id_col = df['Id']
uniq = prepared_data.apply(lambda x: x.nunique())
idxs = np.array((uniq <= 10) & (uniq > 2))
dummies_columns = prepared_data.iloc[:, idxs].columns
cont_cols = set(prepared_data.columns) - set(dummies_columns)
prepared_data = pd.get_dummies(prepared_data, columns=dummies_columns)
prepared_data.fillna(0, inplace=True)
price_multy = prepared_data['SalePrice'].max()
scaler = MinMaxScaler()
prepared_data.loc[:, cont_cols] /= prepared_data.loc[:, cont_cols].max()
prepared_data[['Id', 'Test']] = df[['Id', 'Test']]
train_data = prepared_data.loc[prepared_data['Test'] != True]
train_y = train_data['SalePrice']
train_x = train_data.drop(columns=['SalePrice', 'Id'])
X_train, X_test, y_train, y_test = train_test_split(train_x, train_y, test_size=0.2, random_state=42)
from sklearn.linear_model import *
from sklearn.metrics import *
regr = LinearRegression()
regr.fit(X_train, y_train)
pred = regr.predict(X_test)
mean_squared_error(pred, y_test)
result_df = prepared_data.loc[prepared_data['Test'] == True]
id_col = result_df['Id']
result_df = result_df.drop(columns=['SalePrice', 'Id'])
predictions = regr.predict(result_df)
result_df = prepared_data.loc[prepared_data['Test'] == True]
result_df = prepared_data.loc[prepared_data['Test'] == True]
id_col = result_df['Id']
result_df = result_df.drop(columns=['SalePrice', 'Id'])
predictions = regr.predict(result_df)
res_df = pd.DataFrame(predictions * price_multy, columns=['SalePrice'])
res_df['Id'] = id_col
res_df.to_csv('sub.csv', index=None, header=True)
res_df | code |
18127692/cell_21 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LinearRegression
from sklearn.metrics import r2_score
from sklearn.model_selection import train_test_split
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
headbrain = pd.read_csv('../input/headbrain.csv')
headbrain = headbrain.values
X = headbrain[:, 2]
Y = headbrain[:, 3]
(X.shape, Y.shape)
X = X.reshape(len(X), 1)
X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.3)
reg = LinearRegression()
reg.fit(X_train, y_train)
y_predictions = reg.predict(X_test)
print('R-squared :', r2_score(y_test, y_predictions)) | code |
18127692/cell_4 | [
"text_plain_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
print('Reading the csv file and looking at the first five rows :\n')
headbrain = pd.read_csv('../input/headbrain.csv')
print(headbrain.head()) | code |
18127692/cell_6 | [
"text_plain_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
headbrain = pd.read_csv('../input/headbrain.csv')
print('HeadBrain Info :\n')
print(headbrain.info()) | code |
18127692/cell_2 | [
"text_plain_output_1.png"
] | import os
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
from sklearn.metrics import classification_report
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.metrics import r2_score
import os
print(os.listdir('../input')) | code |
18127692/cell_19 | [
"text_plain_output_1.png"
] | from sklearn.model_selection import train_test_split
import matplotlib.pyplot as plt
import numpy as np # linear algebra
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
headbrain = pd.read_csv('../input/headbrain.csv')
headbrain = headbrain.values
X = headbrain[:, 2]
Y = headbrain[:, 3]
(X.shape, Y.shape)
def Linear_Regression(X, Y):
mean_x = np.mean(X)
mean_y = np.mean(Y)
n = len(X)
numerator = 0
denominator = 0
for i in range(n):
numerator += (X[i] - mean_x) * (Y[i] - mean_y)
denominator += (X[i] - mean_x) ** 2
m = numerator / denominator
c = mean_y - m * mean_x
return (m, c)
def predict(X, m, c):
pred_y = []
for i in range(len(X)):
pred_y.append(c + m * X[i])
return pred_y
def r2score(y_obs, y_pred):
yhat = np.mean(y_obs)
ss_res = 0.0
ss_tot = 0.0
for i in range(len(y_obs)):
ss_tot += (y_obs[i] - yhat) ** 2
ss_res += (y_obs[i] - y_pred[i]) ** 2
r2 = 1 - ss_res / ss_tot
return r2
plt.title('Linear Regression Plot of HeadSize Vs Brain Weight')
X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.3)
m, c = Linear_Regression(X_train, y_train)
print('slope = ', m)
print('intercept = ', c)
y_pred = predict(X_test, m, c)
print('R-squared :', r2score(y_test, y_pred))
plt.plot(X_test, y_pred, color='red', label='Linear Regression')
plt.scatter(X_train, y_train, c='b', label='Scatter Plot')
plt.xlabel('Head Size')
plt.ylabel('Brain Weight')
plt.legend()
plt.show() | code |
18127692/cell_7 | [
"text_plain_output_1.png",
"image_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
headbrain = pd.read_csv('../input/headbrain.csv')
print('Checking for any null values:\n')
print(headbrain.isnull().any()) | code |
18127692/cell_8 | [
"text_plain_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
headbrain = pd.read_csv('../input/headbrain.csv')
print('Checking for unique values in each column:\n')
print(headbrain.nunique()) | code |
18127692/cell_10 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
headbrain = pd.read_csv('../input/headbrain.csv')
plt.figure(figsize=(10, 10))
sns.scatterplot(y='Brain Weight(grams)', x='Head Size(cm^3)', data=headbrain)
plt.show() | code |
18127692/cell_12 | [
"text_plain_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
headbrain = pd.read_csv('../input/headbrain.csv')
headbrain = headbrain.values
X = headbrain[:, 2]
Y = headbrain[:, 3]
(X.shape, Y.shape) | code |
18127692/cell_5 | [
"image_output_1.png"
] | import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
headbrain = pd.read_csv('../input/headbrain.csv')
print(headbrain.shape) | code |
128012739/cell_4 | [
"text_html_output_1.png"
] | from datetime import datetime
import datetime
import yfinance as yf
ticker = 'SPY'
start_time = datetime(2020, 5, 1)
end_time = datetime(2023, 5, 1)
spy = yf.download(ticker, start=start_time, end=end_time) | code |
128012739/cell_6 | [
"text_plain_output_1.png"
] | from datetime import datetime
import datetime
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas_datareader.data as web
import time
import yfinance as yf
ticker = 'SPY'
start_time = datetime(2020, 5, 1)
end_time = datetime(2023, 5, 1)
spy = yf.download(ticker, start=start_time, end=end_time)
connected = False
while not connected:
try:
ticker_df = web.get_data_yahoo(ticker, start=start_time, end=end_time)
connected = True
except Exception as e:
time.sleep(5)
pass
ticker_df = ticker_df.reset_index()
df = pd.DataFrame(ticker_df)
df = df[df['Volume'] != 0]
df.reset_index(drop=True, inplace=True)
df.isna().sum()
df.tail() | code |
128012739/cell_2 | [
"text_plain_output_1.png"
] | !pip install yfinance --upgrade --no-cache-dir
!pip install empyrical
import yfinance as yf
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import plotly.graph_objects as go
import pandas_datareader.data as web
import datetime
import matplotlib.dates as mdates
import time
import empyrical as em
import plotly.express as px
#from IPython.core.display import display, HTML
#display(HTML("<style>.container { width:100% !important; }</style>"))
from plotly import offline
from datetime import datetime
pd.core.common.is_list_like = pd.api.types.is_list_like
offline.init_notebook_mode(connected=True)
yf.pdr_override()
print('✔️ Libraries Loaded!') | code |
128012739/cell_11 | [
"text_plain_output_1.png"
] | from datetime import datetime
import datetime
import numpy as np
import numpy as np
import numpy as np # linear algebra
import yfinance as yf
import yfinance as yf
ticker = 'SPY'
start_time = datetime(2020, 5, 1)
end_time = datetime(2023, 5, 1)
spy = yf.download(ticker, start=start_time, end=end_time)
spy = yf.download(ticker, start=start_time, end=end_time)
start_time = datetime(2020, 5, 1)
end_time = datetime(2023, 5, 1)
spy['5_day_MA'] = spy['Close'].rolling(window=5).mean()
spy['30_day_MA'] = spy['Close'].rolling(window=30).mean()
initial_capital = 100000
capital = initial_capital
position = 0
stop_loss = None
for i in range(30, len(spy) - 1):
pct_change = spy['Close'][i] / spy['Close'][i - 1] - 1
if spy['5_day_MA'][i] > spy['30_day_MA'][i] and position <= 0:
position = capital * 0.3
capital -= position
stop_loss = spy['Close'][i] * 0.9
elif spy['5_day_MA'][i] < spy['30_day_MA'][i] and position >= 0:
position = -capital * 0.2
capital -= position
stop_loss = spy['Close'][i] * 1.1
if position > 0:
if pct_change >= 0.1:
additional_position = capital * 0.3
position += additional_position
capital -= additional_position
stop_loss = spy['Close'][i] * 0.9
elif pct_change >= 0.2:
additional_position = capital * 0.4
position += additional_position
capital -= additional_position
stop_loss = spy['Close'][i] * 0.9
elif pct_change <= -0.1:
capital += position
position = 0
stop_loss = None
elif position < 0:
if pct_change <= -0.1:
additional_position = capital * 0.3
position -= additional_position
capital += additional_position
stop_loss = spy['Close'][i] * 1.1
elif pct_change <= -0.2:
additional_position = capital * 0.4
position -= additional_position
capital += additional_position
stop_loss = spy['Close'][i] * 1.1
elif pct_change >= 0.1:
capital -= position
position = 0
stop_loss = None
portfolio_value = capital + position
import numpy as np
import yfinance as yf
import pandas as pd
def strategy(data, short_window, long_window, long_allocation, short_allocation):
capital = 100000
position = 0
stop_loss = None
for i in range(long_window, len(data) - 1):
pct_change = data['Close'][i] / data['Close'][i - 1] - 1
if data[f'{short_window}_day_MA'][i] > data[f'{long_window}_day_MA'][i] and position <= 0:
position = capital * long_allocation
capital -= position
stop_loss = data['Close'][i] * 0.9
elif data[f'{short_window}_day_MA'][i] < data[f'{long_window}_day_MA'][i] and position >= 0:
position = -capital * short_allocation
capital -= position
stop_loss = data['Close'][i] * 1.1
return capital + position
symbol = 'SPY'
start_date = '2020-05-01'
end_date = '2023-05-01'
data = yf.download(symbol, start=start_date, end=end_date)
best_result = 0
best_params = None
short_windows = np.arange(3, 21, 2)
long_windows = np.arange(20, 61, 5)
long_allocations = np.arange(0.1, 0.51, 0.1)
short_allocations = np.arange(0.1, 0.51, 0.1)
for short_window in short_windows:
for long_window in long_windows:
if short_window >= long_window:
continue
data[f'{short_window}_day_MA'] = data['Close'].rolling(window=short_window).mean()
data[f'{long_window}_day_MA'] = data['Close'].rolling(window=long_window).mean()
for long_allocation in long_allocations:
for short_allocation in short_allocations:
result = strategy(data, short_window, long_window, long_allocation, short_allocation)
if result > best_result:
best_result = result
best_params = (short_window, long_window, long_allocation, short_allocation)
print(f'Best result: {best_result}')
print(f'Best parameters: {best_params}') | code |
128012739/cell_8 | [
"text_plain_output_1.png"
] | from datetime import datetime
from plotly import offline
import datetime
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas_datareader.data as web
import plotly.graph_objects as go
import time
import yfinance as yf
ticker = 'SPY'
start_time = datetime(2020, 5, 1)
end_time = datetime(2023, 5, 1)
spy = yf.download(ticker, start=start_time, end=end_time)
connected = False
while not connected:
try:
ticker_df = web.get_data_yahoo(ticker, start=start_time, end=end_time)
connected = True
except Exception as e:
time.sleep(5)
pass
ticker_df = ticker_df.reset_index()
df = pd.DataFrame(ticker_df)
df = df[df['Volume'] != 0]
df.reset_index(drop=True, inplace=True)
df.isna().sum()
ma5 = df['Close'].rolling(window=5).mean()
trace_ma5 = go.Scatter(x=df.Date, y=ma5, name='5-day Moving Average', line=dict(color='blue'))
ma30 = df['Close'].rolling(window=30).mean()
trace_ma30 = go.Scatter(x=df.Date, y=ma30, name='30-day Moving Average', line=dict(color='orange'))
trace_ma5 = go.Scatter(x=df.Date, y=ma5, name='5-day Moving Average', line=dict(color='blue'))
trace_ma30 = go.Scatter(x=df.Date, y=ma30, name='30-day Moving Average', line=dict(color='orange'))
trace = go.Candlestick(x=df.Date, open=df.Open, high=df.High, low=df.Low, close=df.Close)
chart_data = [trace, trace_ma30, trace_ma5]
layout = go.Layout(title='SPDR S&P 500 ETF Trust (SPY) with Moving Average', width=1800, height=900)
figure = go.Figure(data=chart_data, layout=layout)
offline.iplot(figure) | code |
128012739/cell_10 | [
"text_plain_output_2.png",
"text_plain_output_1.png"
] | from datetime import datetime
import datetime
import yfinance as yf
ticker = 'SPY'
start_time = datetime(2020, 5, 1)
end_time = datetime(2023, 5, 1)
spy = yf.download(ticker, start=start_time, end=end_time)
spy = yf.download(ticker, start=start_time, end=end_time)
start_time = datetime(2020, 5, 1)
end_time = datetime(2023, 5, 1)
spy['5_day_MA'] = spy['Close'].rolling(window=5).mean()
spy['30_day_MA'] = spy['Close'].rolling(window=30).mean()
initial_capital = 100000
capital = initial_capital
position = 0
stop_loss = None
for i in range(30, len(spy) - 1):
pct_change = spy['Close'][i] / spy['Close'][i - 1] - 1
if spy['5_day_MA'][i] > spy['30_day_MA'][i] and position <= 0:
position = capital * 0.3
capital -= position
stop_loss = spy['Close'][i] * 0.9
elif spy['5_day_MA'][i] < spy['30_day_MA'][i] and position >= 0:
position = -capital * 0.2
capital -= position
stop_loss = spy['Close'][i] * 1.1
if position > 0:
if pct_change >= 0.1:
additional_position = capital * 0.3
position += additional_position
capital -= additional_position
stop_loss = spy['Close'][i] * 0.9
elif pct_change >= 0.2:
additional_position = capital * 0.4
position += additional_position
capital -= additional_position
stop_loss = spy['Close'][i] * 0.9
elif pct_change <= -0.1:
capital += position
position = 0
stop_loss = None
elif position < 0:
if pct_change <= -0.1:
additional_position = capital * 0.3
position -= additional_position
capital += additional_position
stop_loss = spy['Close'][i] * 1.1
elif pct_change <= -0.2:
additional_position = capital * 0.4
position -= additional_position
capital += additional_position
stop_loss = spy['Close'][i] * 1.1
elif pct_change >= 0.1:
capital -= position
position = 0
stop_loss = None
portfolio_value = capital + position
print(f'Initial capital: {initial_capital}')
print(f'Final portfolio value: {portfolio_value}') | code |
128012739/cell_5 | [
"text_html_output_1.png"
] | from datetime import datetime
import datetime
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import pandas_datareader.data as web
import time
import yfinance as yf
ticker = 'SPY'
start_time = datetime(2020, 5, 1)
end_time = datetime(2023, 5, 1)
spy = yf.download(ticker, start=start_time, end=end_time)
connected = False
while not connected:
try:
ticker_df = web.get_data_yahoo(ticker, start=start_time, end=end_time)
connected = True
print('connected to yahoo')
except Exception as e:
print('type error: ' + str(e))
time.sleep(5)
pass
ticker_df = ticker_df.reset_index()
df = pd.DataFrame(ticker_df)
print(ticker_df) | code |
74072152/cell_21 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import matplotlib.pyplot as plt
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
import seaborn as sns
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.describe
df.isna().sum()
plt.figure(figsize=(10, 7))
sns.heatmap(df.corr(), annot=True, linewidths=0.2, fmt='.1f', cmap='coolwarm')
plt.show() | code |
74072152/cell_9 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
print(df.Outcome.value_counts()) | code |
74072152/cell_34 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LogisticRegression
model = LogisticRegression()
model = model.fit(X_train, y_train)
score = model.predict(X_train)
pred = model.predict(X_test)
model.coef_ | code |
74072152/cell_23 | [
"application_vnd.jupyter.stderr_output_1.png",
"image_output_1.png"
] | import matplotlib.pyplot as plt
import matplotlib.pyplot as plt
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
import seaborn as sns
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.describe
df.isna().sum()
outcome_column = df['Outcome']
outcome_column.head() | code |
74072152/cell_33 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LogisticRegression
model = LogisticRegression()
model = model.fit(X_train, y_train)
score = model.predict(X_train)
pred = model.predict(X_test)
print('Model Accuracy is : ', pred) | code |
74072152/cell_20 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import matplotlib.pyplot as plt
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
import seaborn as sns
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.describe
df.isna().sum()
plt.figure(figsize=(10, 5))
sns.barplot('Outcome', 'Pregnancies', data=df)
plt.title('Bars of Outcome and Pregnancies')
plt.xlabel('Outcome')
plt.ylabel('Pregnancies')
plt.show() | code |
74072152/cell_6 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.head() | code |
74072152/cell_1 | [
"text_plain_output_1.png"
] | import os
import numpy as np
import pandas as pd
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
74072152/cell_7 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.info() | code |
74072152/cell_32 | [
"text_html_output_1.png"
] | from sklearn.linear_model import LogisticRegression
model = LogisticRegression()
model = model.fit(X_train, y_train)
score = model.predict(X_train)
print('Training Score: ', model.score(X_train, y_train))
print('Testing Score: ', model.score(X_test, y_test)) | code |
74072152/cell_17 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.describe
df.isna().sum()
featureList = ['Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI']
featureList = ['Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI']
print(df[featureList].isin({0}).sum()) | code |
74072152/cell_35 | [
"text_plain_output_1.png"
] | from sklearn.linear_model import LogisticRegression
from sklearn.metrics import accuracy_score, confusion_matrix
model = LogisticRegression()
model = model.fit(X_train, y_train)
score = model.predict(X_train)
pred = model.predict(X_test)
accuracy_score(y_test, pred) | code |
74072152/cell_14 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.describe
df.isna().sum()
featureList = ['Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI']
print(df[featureList].isin({0}).sum()) | code |
74072152/cell_22 | [
"text_plain_output_1.png"
] | import matplotlib.pyplot as plt
import matplotlib.pyplot as plt
import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns
import seaborn as sns
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.describe
df.isna().sum()
feature_columns = df[['Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI', 'DiabetesPedigreeFunction', 'Age']]
feature_columns.head() | code |
74072152/cell_10 | [
"text_html_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.describe | code |
74072152/cell_27 | [
"image_output_1.png"
] | print(X_train.shape)
print(X_test.shape)
print(y_train.shape)
print(y_test.shape) | code |
74072152/cell_12 | [
"text_plain_output_1.png"
] | import pandas as pd
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
df = pd.read_csv('../input/diabetes-prediction-using-logistic-regression/diabetes-dataset.csv')
df.describe
df.isna().sum() | code |
106205967/cell_21 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
dls = TabularPandas(train, splits=splits, cat_names=['food_category', 'food_department', 'food_family', 'promotion_name', 'sales_country', 'marital_status', 'gender', 'education', 'member_card', 'occupation', 'houseowner', 'avg. yearly_income', 'brand_name', 'store_type', 'store_city', 'store_state', 'media_type'], cont_names=['store_sales(in millions)', 'store_cost(in millions)', 'unit_sales(in millions)', 'total_children', 'avg_cars_at home(approx)', 'num_children_at_home', 'avg_cars_at home(approx).1', 'SRP', 'gross_weight', 'net_weight', 'recyclable_package', 'low_fat', 'units_per_case', 'store_sqft', 'grocery_sqft', 'frozen_sqft', 'meat_sqft', 'coffee_bar', 'video_store', 'salad_bar', 'prepared_food', 'florist'], y_names='class', y_block=CategoryBlock(), procs=[Categorify, FillMissing, Normalize]).dataloaders(path='.')
dls
learn = tabular_learner(dls, layers=[200, 100], metrics=accuracy)
learn
learn.lr_find(suggest_funcs=(valley, slide))
learn.fit_one_cycle(10, 0.01)
preds, targs = learn.get_preds() | code |
106205967/cell_23 | [
"text_html_output_2.png"
] | import numpy as np
targs = targs.numpy()
preds = np.argmax(preds.numpy(), axis=-1)
print(preds[0:3])
print(targs[0:3]) | code |
106205967/cell_33 | [
"text_html_output_3.png"
] | import numpy as np
import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
targs = targs.numpy()
preds = np.argmax(preds.numpy(), axis=-1)
tpreds2 = np.argmax(tpreds, axis=-1)
print(testY)
print(tpreds2) | code |
106205967/cell_6 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
display(data[0:3].T)
display(data.info())
display(data.select_dtypes(include='object').columns.tolist())
display(data.select_dtypes(include='float64').columns.tolist()) | code |
106205967/cell_29 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
dls = TabularPandas(train, splits=splits, cat_names=['food_category', 'food_department', 'food_family', 'promotion_name', 'sales_country', 'marital_status', 'gender', 'education', 'member_card', 'occupation', 'houseowner', 'avg. yearly_income', 'brand_name', 'store_type', 'store_city', 'store_state', 'media_type'], cont_names=['store_sales(in millions)', 'store_cost(in millions)', 'unit_sales(in millions)', 'total_children', 'avg_cars_at home(approx)', 'num_children_at_home', 'avg_cars_at home(approx).1', 'SRP', 'gross_weight', 'net_weight', 'recyclable_package', 'low_fat', 'units_per_case', 'store_sqft', 'grocery_sqft', 'frozen_sqft', 'meat_sqft', 'coffee_bar', 'video_store', 'salad_bar', 'prepared_food', 'florist'], y_names='class', y_block=CategoryBlock(), procs=[Categorify, FillMissing, Normalize]).dataloaders(path='.')
dls
learn = tabular_learner(dls, layers=[200, 100], metrics=accuracy)
learn
learn.lr_find(suggest_funcs=(valley, slide))
learn.fit_one_cycle(10, 0.01)
preds, targs = learn.get_preds()
learn.show_results() | code |
106205967/cell_26 | [
"text_plain_output_2.png",
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
dls = TabularPandas(train, splits=splits, cat_names=['food_category', 'food_department', 'food_family', 'promotion_name', 'sales_country', 'marital_status', 'gender', 'education', 'member_card', 'occupation', 'houseowner', 'avg. yearly_income', 'brand_name', 'store_type', 'store_city', 'store_state', 'media_type'], cont_names=['store_sales(in millions)', 'store_cost(in millions)', 'unit_sales(in millions)', 'total_children', 'avg_cars_at home(approx)', 'num_children_at_home', 'avg_cars_at home(approx).1', 'SRP', 'gross_weight', 'net_weight', 'recyclable_package', 'low_fat', 'units_per_case', 'store_sqft', 'grocery_sqft', 'frozen_sqft', 'meat_sqft', 'coffee_bar', 'video_store', 'salad_bar', 'prepared_food', 'florist'], y_names='class', y_block=CategoryBlock(), procs=[Categorify, FillMissing, Normalize]).dataloaders(path='.')
dls
learn = tabular_learner(dls, layers=[200, 100], metrics=accuracy)
learn
learn.lr_find(suggest_funcs=(valley, slide))
learn.fit_one_cycle(10, 0.01)
preds, targs = learn.get_preds()
learn.recorder.plot_sched() | code |
106205967/cell_11 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
print(len(train), len(test))
testY = test['class'].map(normal_mapping) | code |
106205967/cell_19 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
dls = TabularPandas(train, splits=splits, cat_names=['food_category', 'food_department', 'food_family', 'promotion_name', 'sales_country', 'marital_status', 'gender', 'education', 'member_card', 'occupation', 'houseowner', 'avg. yearly_income', 'brand_name', 'store_type', 'store_city', 'store_state', 'media_type'], cont_names=['store_sales(in millions)', 'store_cost(in millions)', 'unit_sales(in millions)', 'total_children', 'avg_cars_at home(approx)', 'num_children_at_home', 'avg_cars_at home(approx).1', 'SRP', 'gross_weight', 'net_weight', 'recyclable_package', 'low_fat', 'units_per_case', 'store_sqft', 'grocery_sqft', 'frozen_sqft', 'meat_sqft', 'coffee_bar', 'video_store', 'salad_bar', 'prepared_food', 'florist'], y_names='class', y_block=CategoryBlock(), procs=[Categorify, FillMissing, Normalize]).dataloaders(path='.')
dls
learn = tabular_learner(dls, layers=[200, 100], metrics=accuracy)
learn
learn.lr_find(suggest_funcs=(valley, slide))
learn.fit_one_cycle(10, 0.01) | code |
106205967/cell_7 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
data['class'] = data['cost'].apply(lambda x: ('00' + str(int(x // 10)))[-2:])
display(data['class']) | code |
106205967/cell_18 | [
"text_plain_output_2.png",
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
dls = TabularPandas(train, splits=splits, cat_names=['food_category', 'food_department', 'food_family', 'promotion_name', 'sales_country', 'marital_status', 'gender', 'education', 'member_card', 'occupation', 'houseowner', 'avg. yearly_income', 'brand_name', 'store_type', 'store_city', 'store_state', 'media_type'], cont_names=['store_sales(in millions)', 'store_cost(in millions)', 'unit_sales(in millions)', 'total_children', 'avg_cars_at home(approx)', 'num_children_at_home', 'avg_cars_at home(approx).1', 'SRP', 'gross_weight', 'net_weight', 'recyclable_package', 'low_fat', 'units_per_case', 'store_sqft', 'grocery_sqft', 'frozen_sqft', 'meat_sqft', 'coffee_bar', 'video_store', 'salad_bar', 'prepared_food', 'florist'], y_names='class', y_block=CategoryBlock(), procs=[Categorify, FillMissing, Normalize]).dataloaders(path='.')
dls
learn = tabular_learner(dls, layers=[200, 100], metrics=accuracy)
learn
learn.lr_find(suggest_funcs=(valley, slide)) | code |
106205967/cell_32 | [
"image_output_1.png"
] | import numpy as np
targs = targs.numpy()
preds = np.argmax(preds.numpy(), axis=-1)
print(tpreds[0:3])
tpreds2 = np.argmax(tpreds, axis=-1)
print(tpreds2[0:3]) | code |
106205967/cell_8 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
print(Name)
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
print(normal_mapping) | code |
106205967/cell_16 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
dls = TabularPandas(train, splits=splits, cat_names=['food_category', 'food_department', 'food_family', 'promotion_name', 'sales_country', 'marital_status', 'gender', 'education', 'member_card', 'occupation', 'houseowner', 'avg. yearly_income', 'brand_name', 'store_type', 'store_city', 'store_state', 'media_type'], cont_names=['store_sales(in millions)', 'store_cost(in millions)', 'unit_sales(in millions)', 'total_children', 'avg_cars_at home(approx)', 'num_children_at_home', 'avg_cars_at home(approx).1', 'SRP', 'gross_weight', 'net_weight', 'recyclable_package', 'low_fat', 'units_per_case', 'store_sqft', 'grocery_sqft', 'frozen_sqft', 'meat_sqft', 'coffee_bar', 'video_store', 'salad_bar', 'prepared_food', 'florist'], y_names='class', y_block=CategoryBlock(), procs=[Categorify, FillMissing, Normalize]).dataloaders(path='.')
dls
learn = tabular_learner(dls, layers=[200, 100], metrics=accuracy)
learn | code |
106205967/cell_31 | [
"image_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
dls = TabularPandas(train, splits=splits, cat_names=['food_category', 'food_department', 'food_family', 'promotion_name', 'sales_country', 'marital_status', 'gender', 'education', 'member_card', 'occupation', 'houseowner', 'avg. yearly_income', 'brand_name', 'store_type', 'store_city', 'store_state', 'media_type'], cont_names=['store_sales(in millions)', 'store_cost(in millions)', 'unit_sales(in millions)', 'total_children', 'avg_cars_at home(approx)', 'num_children_at_home', 'avg_cars_at home(approx).1', 'SRP', 'gross_weight', 'net_weight', 'recyclable_package', 'low_fat', 'units_per_case', 'store_sqft', 'grocery_sqft', 'frozen_sqft', 'meat_sqft', 'coffee_bar', 'video_store', 'salad_bar', 'prepared_food', 'florist'], y_names='class', y_block=CategoryBlock(), procs=[Categorify, FillMissing, Normalize]).dataloaders(path='.')
dls
learn = tabular_learner(dls, layers=[200, 100], metrics=accuracy)
learn
learn.lr_find(suggest_funcs=(valley, slide))
learn.fit_one_cycle(10, 0.01)
preds, targs = learn.get_preds()
tst_dl = learn.dls.test_dl(test)
tpreds, _ = learn.get_preds(dl=tst_dl) | code |
106205967/cell_14 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
dls = TabularPandas(train, splits=splits, cat_names=['food_category', 'food_department', 'food_family', 'promotion_name', 'sales_country', 'marital_status', 'gender', 'education', 'member_card', 'occupation', 'houseowner', 'avg. yearly_income', 'brand_name', 'store_type', 'store_city', 'store_state', 'media_type'], cont_names=['store_sales(in millions)', 'store_cost(in millions)', 'unit_sales(in millions)', 'total_children', 'avg_cars_at home(approx)', 'num_children_at_home', 'avg_cars_at home(approx).1', 'SRP', 'gross_weight', 'net_weight', 'recyclable_package', 'low_fat', 'units_per_case', 'store_sqft', 'grocery_sqft', 'frozen_sqft', 'meat_sqft', 'coffee_bar', 'video_store', 'salad_bar', 'prepared_food', 'florist'], y_names='class', y_block=CategoryBlock(), procs=[Categorify, FillMissing, Normalize]).dataloaders(path='.')
dls | code |
106205967/cell_22 | [
"text_plain_output_1.png",
"image_output_1.png"
] | display(preds)
display(targs) | code |
106205967/cell_10 | [
"text_plain_output_4.png",
"text_plain_output_3.png",
"text_html_output_1.png",
"text_plain_output_2.png",
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
data = data.drop('cost', axis=1)
m = len(data)
print(m)
M = list(range(m))
random.seed(2021)
random.shuffle(M) | code |
106205967/cell_27 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
dls = TabularPandas(train, splits=splits, cat_names=['food_category', 'food_department', 'food_family', 'promotion_name', 'sales_country', 'marital_status', 'gender', 'education', 'member_card', 'occupation', 'houseowner', 'avg. yearly_income', 'brand_name', 'store_type', 'store_city', 'store_state', 'media_type'], cont_names=['store_sales(in millions)', 'store_cost(in millions)', 'unit_sales(in millions)', 'total_children', 'avg_cars_at home(approx)', 'num_children_at_home', 'avg_cars_at home(approx).1', 'SRP', 'gross_weight', 'net_weight', 'recyclable_package', 'low_fat', 'units_per_case', 'store_sqft', 'grocery_sqft', 'frozen_sqft', 'meat_sqft', 'coffee_bar', 'video_store', 'salad_bar', 'prepared_food', 'florist'], y_names='class', y_block=CategoryBlock(), procs=[Categorify, FillMissing, Normalize]).dataloaders(path='.')
dls
learn = tabular_learner(dls, layers=[200, 100], metrics=accuracy)
learn
learn.lr_find(suggest_funcs=(valley, slide))
learn.fit_one_cycle(10, 0.01)
preds, targs = learn.get_preds()
interp = ClassificationInterpretation.from_learner(learn)
interp.plot_confusion_matrix() | code |
106205967/cell_12 | [
"text_plain_output_1.png"
] | import pandas as pd
from fastai.tabular.all import *
pd.options.display.float_format = '{:.2f}'.format
set_seed(42)
data = pd.read_csv('../input/medias-cost-prediction-in-foodmart/media prediction and its cost.csv')
Name = sorted(data['class'].unique().tolist())
N = list(range(len(Name)))
normal_mapping = dict(zip(Name, N))
reverse_mapping = dict(zip(N, Name))
data = data.drop('cost', axis=1)
m = len(data)
M = list(range(m))
random.seed(2021)
random.shuffle(M)
train = data.iloc[M[0:m // 4 * 3]]
test = data.iloc[M[m // 4 * 3:]]
testY = test['class'].map(normal_mapping)
splits = RandomSplitter(seed=42)(train)
display(splits)
print(len(splits[0]), len(splits[1])) | code |
33104580/cell_4 | [
"text_html_output_1.png"
] | import matplotlib.pyplot as plt
import os
import warnings
import numpy as np
import pandas as pd
import os
from urllib.request import urlopen
import json
import plotly.express as px
import plotly.offline as py
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import seaborn as sns
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
from datetime import datetime, timedelta
import warnings
warnings.filterwarnings('ignore')
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
33104580/cell_6 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from datetime import datetime, timedelta
import numpy as np
import pandas as pd
latest_date = datetime.today() - timedelta(days=1)
latest_date = latest_date.strftime('%m/%d/%y')[1:]
df_cases = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_confirmed_usafacts.csv')[['countyFIPS', 'County Name', 'State', latest_date]]
df_cases = df_cases.rename(columns={'countyFIPS': 'county_fips', latest_date: 'confirmed'}).set_index('county_fips')
df_deaths = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_deaths_usafacts.csv')[['countyFIPS', latest_date]]
df_deaths = df_deaths.rename(columns={'countyFIPS': 'county_fips', latest_date: 'deaths'}).set_index('county_fips')
df_pop = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_county_population_usafacts.csv')[['countyFIPS', 'population']]
df_pop = df_pop.rename(columns={'countyFIPS': 'county_fips'}).set_index('county_fips')
df = df_cases.join(df_deaths)
df = df.join(df_pop)
df = df[df.index > 999]
df = df[df.population > 0]
del df_cases, df_deaths, df_pop
df['mortality'] = df['deaths'] / df['confirmed']
df['mortality'] = df['mortality'].fillna(0)
df['deaths_per_million'] = df['deaths'] * 1000000 / df['population']
df['cases_per_million'] = df['confirmed'] * 1000000 / df['population']
df['likely_infected_high'] = np.round(df['confirmed'] * 80 / df['population'], 2)
df['likely_infected_high'] = np.clip(df['likely_infected_high'], 0, 1)
df['likely_infected_low'] = np.round(df['confirmed'] * 28 / df['population'], 2)
df['likely_infected_low'] = np.clip(df['likely_infected_low'], 0, 1)
df['county_state'] = df['County Name'] + ', ' + df['State']
print('Number of counties: ' + str(df.index.nunique())) | code |
33104580/cell_11 | [
"text_plain_output_1.png"
] | from datetime import datetime, timedelta
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import seaborn as sns
import warnings
import numpy as np
import pandas as pd
import os
from urllib.request import urlopen
import json
import plotly.express as px
import plotly.offline as py
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import seaborn as sns
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
from datetime import datetime, timedelta
import warnings
warnings.filterwarnings('ignore')
latest_date = datetime.today() - timedelta(days=1)
latest_date = latest_date.strftime('%m/%d/%y')[1:]
df_cases = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_confirmed_usafacts.csv')[['countyFIPS', 'County Name', 'State', latest_date]]
df_cases = df_cases.rename(columns={'countyFIPS': 'county_fips', latest_date: 'confirmed'}).set_index('county_fips')
df_deaths = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_deaths_usafacts.csv')[['countyFIPS', latest_date]]
df_deaths = df_deaths.rename(columns={'countyFIPS': 'county_fips', latest_date: 'deaths'}).set_index('county_fips')
df_pop = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_county_population_usafacts.csv')[['countyFIPS', 'population']]
df_pop = df_pop.rename(columns={'countyFIPS': 'county_fips'}).set_index('county_fips')
df = df_cases.join(df_deaths)
df = df.join(df_pop)
df = df[df.index > 999]
df = df[df.population > 0]
del df_cases, df_deaths, df_pop
df['mortality'] = df['deaths'] / df['confirmed']
df['mortality'] = df['mortality'].fillna(0)
df['deaths_per_million'] = df['deaths'] * 1000000 / df['population']
df['cases_per_million'] = df['confirmed'] * 1000000 / df['population']
df['likely_infected_high'] = np.round(df['confirmed'] * 80 / df['population'], 2)
df['likely_infected_high'] = np.clip(df['likely_infected_high'], 0, 1)
df['likely_infected_low'] = np.round(df['confirmed'] * 28 / df['population'], 2)
df['likely_infected_low'] = np.clip(df['likely_infected_low'], 0, 1)
df['county_state'] = df['County Name'] + ', ' + df['State']
df_county_stats = pd.read_csv('/kaggle/input/uncover/county_health_rankings/county_health_rankings/us-county-health-rankings-2020.csv')[['fips', 'segregation_index', 'percent_black', 'median_household_income', 'percent_adults_with_obesity', 'percent_smokers', 'percent_with_access_to_exercise_opportunities', 'percent_some_college', 'percent_unemployed', 'percent_children_in_poverty']]
df_county_stats = df_county_stats.rename(columns={'fips': 'county_fips', 'segregation_index': 'segregation_level'}).set_index('county_fips')
df = df.join(df_county_stats)
df_county_stats = pd.read_csv('/kaggle/input/county-ranking-data/county_ranking.csv')[['fipscode', 'v052_rawvalue', 'v053_rawvalue', 'v044_rawvalue', 'v147_rawvalue', 'v002_cilow', 'v136_other_data_2']]
df_county_stats = df_county_stats.rename(columns={'fipscode': 'county_fips', 'v052_rawvalue': 'percent_below_18', 'v053_rawvalue': 'percent_above_65', 'v044_rawvalue': 'income_inequality', 'v147_rawvalue': 'life_expectancy', 'v002_cilow': 'poor_fair_health', 'v136_other_data_2': 'over_crowding'}).set_index('county_fips')
df = df.join(df_county_stats)
df = df.reset_index()
df['county_fips'] = df['county_fips'].astype(str).str.rjust(5, '0')
plt.figure(figsize=(10, 5))
sns.distplot(df.likely_infected_high, hist=True, kde=False, color='red', hist_kws={'edgecolor': 'black', 'linewidth': 1}, kde_kws={'linewidth': 2})
print('Summary Statistic of Percetnage of Population Likely Infected across counties: \n')
print(df.likely_infected_high.describe())
plt.xlim(0, 1)
plt.title('Distribution of county population likely infected')
plt.xlabel('Percentage of population likely infected')
plt.show() | code |
33104580/cell_19 | [
"text_plain_output_1.png"
] | from datetime import datetime, timedelta
from urllib.request import urlopen
import json
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import plotly.express as px
import plotly.express as px
import seaborn as sns
import warnings
import numpy as np
import pandas as pd
import os
from urllib.request import urlopen
import json
import plotly.express as px
import plotly.offline as py
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import seaborn as sns
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
from datetime import datetime, timedelta
import warnings
warnings.filterwarnings('ignore')
latest_date = datetime.today() - timedelta(days=1)
latest_date = latest_date.strftime('%m/%d/%y')[1:]
df_cases = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_confirmed_usafacts.csv')[['countyFIPS', 'County Name', 'State', latest_date]]
df_cases = df_cases.rename(columns={'countyFIPS': 'county_fips', latest_date: 'confirmed'}).set_index('county_fips')
df_deaths = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_deaths_usafacts.csv')[['countyFIPS', latest_date]]
df_deaths = df_deaths.rename(columns={'countyFIPS': 'county_fips', latest_date: 'deaths'}).set_index('county_fips')
df_pop = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_county_population_usafacts.csv')[['countyFIPS', 'population']]
df_pop = df_pop.rename(columns={'countyFIPS': 'county_fips'}).set_index('county_fips')
df = df_cases.join(df_deaths)
df = df.join(df_pop)
df = df[df.index > 999]
df = df[df.population > 0]
del df_cases, df_deaths, df_pop
df['mortality'] = df['deaths'] / df['confirmed']
df['mortality'] = df['mortality'].fillna(0)
df['deaths_per_million'] = df['deaths'] * 1000000 / df['population']
df['cases_per_million'] = df['confirmed'] * 1000000 / df['population']
df['likely_infected_high'] = np.round(df['confirmed'] * 80 / df['population'], 2)
df['likely_infected_high'] = np.clip(df['likely_infected_high'], 0, 1)
df['likely_infected_low'] = np.round(df['confirmed'] * 28 / df['population'], 2)
df['likely_infected_low'] = np.clip(df['likely_infected_low'], 0, 1)
df['county_state'] = df['County Name'] + ', ' + df['State']
df_county_stats = pd.read_csv('/kaggle/input/uncover/county_health_rankings/county_health_rankings/us-county-health-rankings-2020.csv')[['fips', 'segregation_index', 'percent_black', 'median_household_income', 'percent_adults_with_obesity', 'percent_smokers', 'percent_with_access_to_exercise_opportunities', 'percent_some_college', 'percent_unemployed', 'percent_children_in_poverty']]
df_county_stats = df_county_stats.rename(columns={'fips': 'county_fips', 'segregation_index': 'segregation_level'}).set_index('county_fips')
df = df.join(df_county_stats)
df_county_stats = pd.read_csv('/kaggle/input/county-ranking-data/county_ranking.csv')[['fipscode', 'v052_rawvalue', 'v053_rawvalue', 'v044_rawvalue', 'v147_rawvalue', 'v002_cilow', 'v136_other_data_2']]
df_county_stats = df_county_stats.rename(columns={'fipscode': 'county_fips', 'v052_rawvalue': 'percent_below_18', 'v053_rawvalue': 'percent_above_65', 'v044_rawvalue': 'income_inequality', 'v147_rawvalue': 'life_expectancy', 'v002_cilow': 'poor_fair_health', 'v136_other_data_2': 'over_crowding'}).set_index('county_fips')
df = df.join(df_county_stats)
df = df.reset_index()
df['county_fips'] = df['county_fips'].astype(str).str.rjust(5, '0')
plt.xlim(0, 1)
def show_values_on_bars(axs, h_v="v", space=0.4, text_size=10):
def _show_on_single_plot(ax):
if h_v == "v":
for p in ax.patches:
_x = p.get_x() + p.get_width() / 2
_y = p.get_y() + p.get_height()
value = p.get_height()
ax.text(_x, _y, value, ha="center", size=text_size)
elif h_v == "h":
for p in ax.patches:
_x = p.get_x() + p.get_width() + float(space)
_y = p.get_y() + p.get_height()- float(0.2)
value = p.get_width()
value = "{:.1%}".format(value)
ax.text(_x, _y, value, ha="left", size=text_size)
if isinstance(axs, np.ndarray):
for idx, ax in np.ndenumerate(axs):
_show_on_single_plot(ax)
else:
_show_on_single_plot(axs)
plt.figure(figsize=(10,10))
g=sns.barplot(x='likely_infected_low', y='county_state',data=df.sort_values(['likely_infected_low'], ascending=False).head(20), color="lightgreen")
show_values_on_bars(g, "h", space=0.01, text_size=10)
plt.xlim(0, 1.1)
plt.xlabel("Percentage of population infected")
plt.ylabel("County, State")
plt.title("Likely spread of Virus if spread is 28 fold")
import plotly.express as px
with urlopen('https://raw.githubusercontent.com/plotly/datasets/master/geojson-counties-fips.json') as response:
counties = json.load(response)
fig = px.choropleth(df, geojson=counties, locations='county_fips', color='likely_infected_low', color_continuous_scale='Reds', range_color=(0, 0.2), scope='usa', title='Percentage of population likely already infected if spread is 28 folds', hover_name='county_state', hover_data=['confirmed', 'deaths'], labels={'likely_infected_low': '% Likely Infected', 'confirmed': 'Confirmed Cases ', 'deaths': 'Deaths '})
fig.update_layout(margin={'r': 0, 't': 30, 'l': 0, 'b': 0})
fig.layout.template = None
config = dict({'scrollZoom': False})
fig.write_html('1.html')
fig = px.choropleth(df, geojson=counties, locations='county_fips', color='likely_infected_high', color_continuous_scale='Reds', range_color=(0, 0.2), scope='usa', title='Percentage of population likely already infected if spread is 80 folds', hover_name='county_state', hover_data=['confirmed', 'deaths'], labels={'likely_infected_high': '% Likely Infected', 'confirmed': 'Confirmed Cases ', 'deaths': 'Deaths '})
fig.update_layout(margin={'r': 0, 't': 30, 'l': 0, 'b': 0})
fig.layout.template = None
config = dict({'scrollZoom': False})
fig.show(config=config)
fig.write_html('2.html') | code |
33104580/cell_7 | [
"image_output_1.png"
] | from datetime import datetime, timedelta
import numpy as np
import pandas as pd
latest_date = datetime.today() - timedelta(days=1)
latest_date = latest_date.strftime('%m/%d/%y')[1:]
df_cases = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_confirmed_usafacts.csv')[['countyFIPS', 'County Name', 'State', latest_date]]
df_cases = df_cases.rename(columns={'countyFIPS': 'county_fips', latest_date: 'confirmed'}).set_index('county_fips')
df_deaths = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_deaths_usafacts.csv')[['countyFIPS', latest_date]]
df_deaths = df_deaths.rename(columns={'countyFIPS': 'county_fips', latest_date: 'deaths'}).set_index('county_fips')
df_pop = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_county_population_usafacts.csv')[['countyFIPS', 'population']]
df_pop = df_pop.rename(columns={'countyFIPS': 'county_fips'}).set_index('county_fips')
df = df_cases.join(df_deaths)
df = df.join(df_pop)
df = df[df.index > 999]
df = df[df.population > 0]
del df_cases, df_deaths, df_pop
df['mortality'] = df['deaths'] / df['confirmed']
df['mortality'] = df['mortality'].fillna(0)
df['deaths_per_million'] = df['deaths'] * 1000000 / df['population']
df['cases_per_million'] = df['confirmed'] * 1000000 / df['population']
df['likely_infected_high'] = np.round(df['confirmed'] * 80 / df['population'], 2)
df['likely_infected_high'] = np.clip(df['likely_infected_high'], 0, 1)
df['likely_infected_low'] = np.round(df['confirmed'] * 28 / df['population'], 2)
df['likely_infected_low'] = np.clip(df['likely_infected_low'], 0, 1)
df['county_state'] = df['County Name'] + ', ' + df['State']
df_county_stats = pd.read_csv('/kaggle/input/uncover/county_health_rankings/county_health_rankings/us-county-health-rankings-2020.csv')[['fips', 'segregation_index', 'percent_black', 'median_household_income', 'percent_adults_with_obesity', 'percent_smokers', 'percent_with_access_to_exercise_opportunities', 'percent_some_college', 'percent_unemployed', 'percent_children_in_poverty']]
df_county_stats = df_county_stats.rename(columns={'fips': 'county_fips', 'segregation_index': 'segregation_level'}).set_index('county_fips')
df = df.join(df_county_stats)
df_county_stats = pd.read_csv('/kaggle/input/county-ranking-data/county_ranking.csv')[['fipscode', 'v052_rawvalue', 'v053_rawvalue', 'v044_rawvalue', 'v147_rawvalue', 'v002_cilow', 'v136_other_data_2']]
df_county_stats = df_county_stats.rename(columns={'fipscode': 'county_fips', 'v052_rawvalue': 'percent_below_18', 'v053_rawvalue': 'percent_above_65', 'v044_rawvalue': 'income_inequality', 'v147_rawvalue': 'life_expectancy', 'v002_cilow': 'poor_fair_health', 'v136_other_data_2': 'over_crowding'}).set_index('county_fips')
df = df.join(df_county_stats)
df = df.reset_index()
df['county_fips'] = df['county_fips'].astype(str).str.rjust(5, '0')
df.head() | code |
33104580/cell_8 | [
"text_html_output_1.png"
] | !pip install chart_studio
!pip install plotly-geo | code |
33104580/cell_17 | [
"text_html_output_1.png"
] | from datetime import datetime, timedelta
from urllib.request import urlopen
import json
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import plotly.express as px
import plotly.express as px
import seaborn as sns
import warnings
import numpy as np
import pandas as pd
import os
from urllib.request import urlopen
import json
import plotly.express as px
import plotly.offline as py
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import seaborn as sns
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
from datetime import datetime, timedelta
import warnings
warnings.filterwarnings('ignore')
latest_date = datetime.today() - timedelta(days=1)
latest_date = latest_date.strftime('%m/%d/%y')[1:]
df_cases = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_confirmed_usafacts.csv')[['countyFIPS', 'County Name', 'State', latest_date]]
df_cases = df_cases.rename(columns={'countyFIPS': 'county_fips', latest_date: 'confirmed'}).set_index('county_fips')
df_deaths = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_deaths_usafacts.csv')[['countyFIPS', latest_date]]
df_deaths = df_deaths.rename(columns={'countyFIPS': 'county_fips', latest_date: 'deaths'}).set_index('county_fips')
df_pop = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_county_population_usafacts.csv')[['countyFIPS', 'population']]
df_pop = df_pop.rename(columns={'countyFIPS': 'county_fips'}).set_index('county_fips')
df = df_cases.join(df_deaths)
df = df.join(df_pop)
df = df[df.index > 999]
df = df[df.population > 0]
del df_cases, df_deaths, df_pop
df['mortality'] = df['deaths'] / df['confirmed']
df['mortality'] = df['mortality'].fillna(0)
df['deaths_per_million'] = df['deaths'] * 1000000 / df['population']
df['cases_per_million'] = df['confirmed'] * 1000000 / df['population']
df['likely_infected_high'] = np.round(df['confirmed'] * 80 / df['population'], 2)
df['likely_infected_high'] = np.clip(df['likely_infected_high'], 0, 1)
df['likely_infected_low'] = np.round(df['confirmed'] * 28 / df['population'], 2)
df['likely_infected_low'] = np.clip(df['likely_infected_low'], 0, 1)
df['county_state'] = df['County Name'] + ', ' + df['State']
df_county_stats = pd.read_csv('/kaggle/input/uncover/county_health_rankings/county_health_rankings/us-county-health-rankings-2020.csv')[['fips', 'segregation_index', 'percent_black', 'median_household_income', 'percent_adults_with_obesity', 'percent_smokers', 'percent_with_access_to_exercise_opportunities', 'percent_some_college', 'percent_unemployed', 'percent_children_in_poverty']]
df_county_stats = df_county_stats.rename(columns={'fips': 'county_fips', 'segregation_index': 'segregation_level'}).set_index('county_fips')
df = df.join(df_county_stats)
df_county_stats = pd.read_csv('/kaggle/input/county-ranking-data/county_ranking.csv')[['fipscode', 'v052_rawvalue', 'v053_rawvalue', 'v044_rawvalue', 'v147_rawvalue', 'v002_cilow', 'v136_other_data_2']]
df_county_stats = df_county_stats.rename(columns={'fipscode': 'county_fips', 'v052_rawvalue': 'percent_below_18', 'v053_rawvalue': 'percent_above_65', 'v044_rawvalue': 'income_inequality', 'v147_rawvalue': 'life_expectancy', 'v002_cilow': 'poor_fair_health', 'v136_other_data_2': 'over_crowding'}).set_index('county_fips')
df = df.join(df_county_stats)
df = df.reset_index()
df['county_fips'] = df['county_fips'].astype(str).str.rjust(5, '0')
plt.xlim(0, 1)
def show_values_on_bars(axs, h_v="v", space=0.4, text_size=10):
def _show_on_single_plot(ax):
if h_v == "v":
for p in ax.patches:
_x = p.get_x() + p.get_width() / 2
_y = p.get_y() + p.get_height()
value = p.get_height()
ax.text(_x, _y, value, ha="center", size=text_size)
elif h_v == "h":
for p in ax.patches:
_x = p.get_x() + p.get_width() + float(space)
_y = p.get_y() + p.get_height()- float(0.2)
value = p.get_width()
value = "{:.1%}".format(value)
ax.text(_x, _y, value, ha="left", size=text_size)
if isinstance(axs, np.ndarray):
for idx, ax in np.ndenumerate(axs):
_show_on_single_plot(ax)
else:
_show_on_single_plot(axs)
plt.figure(figsize=(10,10))
g=sns.barplot(x='likely_infected_low', y='county_state',data=df.sort_values(['likely_infected_low'], ascending=False).head(20), color="lightgreen")
show_values_on_bars(g, "h", space=0.01, text_size=10)
plt.xlim(0, 1.1)
plt.xlabel("Percentage of population infected")
plt.ylabel("County, State")
plt.title("Likely spread of Virus if spread is 28 fold")
import plotly.express as px
with urlopen('https://raw.githubusercontent.com/plotly/datasets/master/geojson-counties-fips.json') as response:
counties = json.load(response)
fig = px.choropleth(df, geojson=counties, locations='county_fips', color='likely_infected_low', color_continuous_scale='Reds', range_color=(0, 0.2), scope='usa', title='Percentage of population likely already infected if spread is 28 folds', hover_name='county_state', hover_data=['confirmed', 'deaths'], labels={'likely_infected_low': '% Likely Infected', 'confirmed': 'Confirmed Cases ', 'deaths': 'Deaths '})
fig.update_layout(margin={'r': 0, 't': 30, 'l': 0, 'b': 0})
fig.layout.template = None
config = dict({'scrollZoom': False})
fig.show(config=config)
fig.write_html('1.html') | code |
33104580/cell_24 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from datetime import datetime, timedelta
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import seaborn as sns
import warnings
import numpy as np
import pandas as pd
import os
from urllib.request import urlopen
import json
import plotly.express as px
import plotly.offline as py
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import seaborn as sns
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
from datetime import datetime, timedelta
import warnings
warnings.filterwarnings('ignore')
latest_date = datetime.today() - timedelta(days=1)
latest_date = latest_date.strftime('%m/%d/%y')[1:]
df_cases = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_confirmed_usafacts.csv')[['countyFIPS', 'County Name', 'State', latest_date]]
df_cases = df_cases.rename(columns={'countyFIPS': 'county_fips', latest_date: 'confirmed'}).set_index('county_fips')
df_deaths = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_deaths_usafacts.csv')[['countyFIPS', latest_date]]
df_deaths = df_deaths.rename(columns={'countyFIPS': 'county_fips', latest_date: 'deaths'}).set_index('county_fips')
df_pop = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_county_population_usafacts.csv')[['countyFIPS', 'population']]
df_pop = df_pop.rename(columns={'countyFIPS': 'county_fips'}).set_index('county_fips')
df = df_cases.join(df_deaths)
df = df.join(df_pop)
df = df[df.index > 999]
df = df[df.population > 0]
del df_cases, df_deaths, df_pop
df['mortality'] = df['deaths'] / df['confirmed']
df['mortality'] = df['mortality'].fillna(0)
df['deaths_per_million'] = df['deaths'] * 1000000 / df['population']
df['cases_per_million'] = df['confirmed'] * 1000000 / df['population']
df['likely_infected_high'] = np.round(df['confirmed'] * 80 / df['population'], 2)
df['likely_infected_high'] = np.clip(df['likely_infected_high'], 0, 1)
df['likely_infected_low'] = np.round(df['confirmed'] * 28 / df['population'], 2)
df['likely_infected_low'] = np.clip(df['likely_infected_low'], 0, 1)
df['county_state'] = df['County Name'] + ', ' + df['State']
df_county_stats = pd.read_csv('/kaggle/input/uncover/county_health_rankings/county_health_rankings/us-county-health-rankings-2020.csv')[['fips', 'segregation_index', 'percent_black', 'median_household_income', 'percent_adults_with_obesity', 'percent_smokers', 'percent_with_access_to_exercise_opportunities', 'percent_some_college', 'percent_unemployed', 'percent_children_in_poverty']]
df_county_stats = df_county_stats.rename(columns={'fips': 'county_fips', 'segregation_index': 'segregation_level'}).set_index('county_fips')
df = df.join(df_county_stats)
df_county_stats = pd.read_csv('/kaggle/input/county-ranking-data/county_ranking.csv')[['fipscode', 'v052_rawvalue', 'v053_rawvalue', 'v044_rawvalue', 'v147_rawvalue', 'v002_cilow', 'v136_other_data_2']]
df_county_stats = df_county_stats.rename(columns={'fipscode': 'county_fips', 'v052_rawvalue': 'percent_below_18', 'v053_rawvalue': 'percent_above_65', 'v044_rawvalue': 'income_inequality', 'v147_rawvalue': 'life_expectancy', 'v002_cilow': 'poor_fair_health', 'v136_other_data_2': 'over_crowding'}).set_index('county_fips')
df = df.join(df_county_stats)
df = df.reset_index()
df['county_fips'] = df['county_fips'].astype(str).str.rjust(5, '0')
plt.xlim(0, 1)
def show_values_on_bars(axs, h_v="v", space=0.4, text_size=10):
def _show_on_single_plot(ax):
if h_v == "v":
for p in ax.patches:
_x = p.get_x() + p.get_width() / 2
_y = p.get_y() + p.get_height()
value = p.get_height()
ax.text(_x, _y, value, ha="center", size=text_size)
elif h_v == "h":
for p in ax.patches:
_x = p.get_x() + p.get_width() + float(space)
_y = p.get_y() + p.get_height()- float(0.2)
value = p.get_width()
value = "{:.1%}".format(value)
ax.text(_x, _y, value, ha="left", size=text_size)
if isinstance(axs, np.ndarray):
for idx, ax in np.ndenumerate(axs):
_show_on_single_plot(ax)
else:
_show_on_single_plot(axs)
plt.figure(figsize=(10,10))
g=sns.barplot(x='likely_infected_low', y='county_state',data=df.sort_values(['likely_infected_low'], ascending=False).head(20), color="lightgreen")
show_values_on_bars(g, "h", space=0.01, text_size=10)
plt.xlim(0, 1.1)
plt.xlabel("Percentage of population infected")
plt.ylabel("County, State")
plt.title("Likely spread of Virus if spread is 28 fold")
df_temp = df[df.confirmed > 500]
plt.figure(figsize=(20, 8))
plt.subplot(1, 2, 1)
g = sns.barplot(x='mortality', y='county_state', data=df[df.confirmed > 500].sort_values(['mortality'], ascending=False).head(10), color='red')
show_values_on_bars(g, 'h', space=0.002, text_size=20)
plt.xlim(0, 0.2)
plt.xlabel('Covid Mortality Rate', size=20)
plt.ylabel(' ', size=20)
plt.yticks(size=15)
plt.title('Counties with highest Covid Mortality', size=25)
plt.subplot(1, 2, 2)
g = sns.barplot(x='mortality', y='county_state', data=df[df.confirmed > 500].sort_values(['mortality'], ascending=True).head(10), color='blue')
show_values_on_bars(g, 'h', space=0.002, text_size=20)
plt.xlim(0, 0.05)
plt.xlabel('Covid Mortality Rate', size=20)
plt.ylabel(' ')
plt.yticks(size=15)
plt.title('Counties with lowest Covid Mortality', size=25)
plt.tight_layout() | code |
33104580/cell_14 | [
"text_plain_output_1.png"
] | from datetime import datetime, timedelta
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import seaborn as sns
import warnings
import numpy as np
import pandas as pd
import os
from urllib.request import urlopen
import json
import plotly.express as px
import plotly.offline as py
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import seaborn as sns
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
from datetime import datetime, timedelta
import warnings
warnings.filterwarnings('ignore')
latest_date = datetime.today() - timedelta(days=1)
latest_date = latest_date.strftime('%m/%d/%y')[1:]
df_cases = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_confirmed_usafacts.csv')[['countyFIPS', 'County Name', 'State', latest_date]]
df_cases = df_cases.rename(columns={'countyFIPS': 'county_fips', latest_date: 'confirmed'}).set_index('county_fips')
df_deaths = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_deaths_usafacts.csv')[['countyFIPS', latest_date]]
df_deaths = df_deaths.rename(columns={'countyFIPS': 'county_fips', latest_date: 'deaths'}).set_index('county_fips')
df_pop = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_county_population_usafacts.csv')[['countyFIPS', 'population']]
df_pop = df_pop.rename(columns={'countyFIPS': 'county_fips'}).set_index('county_fips')
df = df_cases.join(df_deaths)
df = df.join(df_pop)
df = df[df.index > 999]
df = df[df.population > 0]
del df_cases, df_deaths, df_pop
df['mortality'] = df['deaths'] / df['confirmed']
df['mortality'] = df['mortality'].fillna(0)
df['deaths_per_million'] = df['deaths'] * 1000000 / df['population']
df['cases_per_million'] = df['confirmed'] * 1000000 / df['population']
df['likely_infected_high'] = np.round(df['confirmed'] * 80 / df['population'], 2)
df['likely_infected_high'] = np.clip(df['likely_infected_high'], 0, 1)
df['likely_infected_low'] = np.round(df['confirmed'] * 28 / df['population'], 2)
df['likely_infected_low'] = np.clip(df['likely_infected_low'], 0, 1)
df['county_state'] = df['County Name'] + ', ' + df['State']
df_county_stats = pd.read_csv('/kaggle/input/uncover/county_health_rankings/county_health_rankings/us-county-health-rankings-2020.csv')[['fips', 'segregation_index', 'percent_black', 'median_household_income', 'percent_adults_with_obesity', 'percent_smokers', 'percent_with_access_to_exercise_opportunities', 'percent_some_college', 'percent_unemployed', 'percent_children_in_poverty']]
df_county_stats = df_county_stats.rename(columns={'fips': 'county_fips', 'segregation_index': 'segregation_level'}).set_index('county_fips')
df = df.join(df_county_stats)
df_county_stats = pd.read_csv('/kaggle/input/county-ranking-data/county_ranking.csv')[['fipscode', 'v052_rawvalue', 'v053_rawvalue', 'v044_rawvalue', 'v147_rawvalue', 'v002_cilow', 'v136_other_data_2']]
df_county_stats = df_county_stats.rename(columns={'fipscode': 'county_fips', 'v052_rawvalue': 'percent_below_18', 'v053_rawvalue': 'percent_above_65', 'v044_rawvalue': 'income_inequality', 'v147_rawvalue': 'life_expectancy', 'v002_cilow': 'poor_fair_health', 'v136_other_data_2': 'over_crowding'}).set_index('county_fips')
df = df.join(df_county_stats)
df = df.reset_index()
df['county_fips'] = df['county_fips'].astype(str).str.rjust(5, '0')
plt.xlim(0, 1)
def show_values_on_bars(axs, h_v='v', space=0.4, text_size=10):
def _show_on_single_plot(ax):
if h_v == 'v':
for p in ax.patches:
_x = p.get_x() + p.get_width() / 2
_y = p.get_y() + p.get_height()
value = p.get_height()
ax.text(_x, _y, value, ha='center', size=text_size)
elif h_v == 'h':
for p in ax.patches:
_x = p.get_x() + p.get_width() + float(space)
_y = p.get_y() + p.get_height() - float(0.2)
value = p.get_width()
value = '{:.1%}'.format(value)
ax.text(_x, _y, value, ha='left', size=text_size)
if isinstance(axs, np.ndarray):
for idx, ax in np.ndenumerate(axs):
_show_on_single_plot(ax)
else:
_show_on_single_plot(axs)
plt.figure(figsize=(10, 10))
g = sns.barplot(x='likely_infected_low', y='county_state', data=df.sort_values(['likely_infected_low'], ascending=False).head(20), color='lightgreen')
show_values_on_bars(g, 'h', space=0.01, text_size=10)
plt.xlim(0, 1.1)
plt.xlabel('Percentage of population infected')
plt.ylabel('County, State')
plt.title('Likely spread of Virus if spread is 28 fold') | code |
33104580/cell_22 | [
"text_plain_output_1.png",
"image_output_1.png"
] | from datetime import datetime, timedelta
import matplotlib.pyplot as plt
import numpy as np
import os
import pandas as pd
import seaborn as sns
import warnings
import numpy as np
import pandas as pd
import os
from urllib.request import urlopen
import json
import plotly.express as px
import plotly.offline as py
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import seaborn as sns
import matplotlib.pyplot as plt
plt.style.use('fivethirtyeight')
from datetime import datetime, timedelta
import warnings
warnings.filterwarnings('ignore')
latest_date = datetime.today() - timedelta(days=1)
latest_date = latest_date.strftime('%m/%d/%y')[1:]
df_cases = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_confirmed_usafacts.csv')[['countyFIPS', 'County Name', 'State', latest_date]]
df_cases = df_cases.rename(columns={'countyFIPS': 'county_fips', latest_date: 'confirmed'}).set_index('county_fips')
df_deaths = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_deaths_usafacts.csv')[['countyFIPS', latest_date]]
df_deaths = df_deaths.rename(columns={'countyFIPS': 'county_fips', latest_date: 'deaths'}).set_index('county_fips')
df_pop = pd.read_csv('https://usafactsstatic.blob.core.windows.net/public/data/covid-19/covid_county_population_usafacts.csv')[['countyFIPS', 'population']]
df_pop = df_pop.rename(columns={'countyFIPS': 'county_fips'}).set_index('county_fips')
df = df_cases.join(df_deaths)
df = df.join(df_pop)
df = df[df.index > 999]
df = df[df.population > 0]
del df_cases, df_deaths, df_pop
df['mortality'] = df['deaths'] / df['confirmed']
df['mortality'] = df['mortality'].fillna(0)
df['deaths_per_million'] = df['deaths'] * 1000000 / df['population']
df['cases_per_million'] = df['confirmed'] * 1000000 / df['population']
df['likely_infected_high'] = np.round(df['confirmed'] * 80 / df['population'], 2)
df['likely_infected_high'] = np.clip(df['likely_infected_high'], 0, 1)
df['likely_infected_low'] = np.round(df['confirmed'] * 28 / df['population'], 2)
df['likely_infected_low'] = np.clip(df['likely_infected_low'], 0, 1)
df['county_state'] = df['County Name'] + ', ' + df['State']
df_county_stats = pd.read_csv('/kaggle/input/uncover/county_health_rankings/county_health_rankings/us-county-health-rankings-2020.csv')[['fips', 'segregation_index', 'percent_black', 'median_household_income', 'percent_adults_with_obesity', 'percent_smokers', 'percent_with_access_to_exercise_opportunities', 'percent_some_college', 'percent_unemployed', 'percent_children_in_poverty']]
df_county_stats = df_county_stats.rename(columns={'fips': 'county_fips', 'segregation_index': 'segregation_level'}).set_index('county_fips')
df = df.join(df_county_stats)
df_county_stats = pd.read_csv('/kaggle/input/county-ranking-data/county_ranking.csv')[['fipscode', 'v052_rawvalue', 'v053_rawvalue', 'v044_rawvalue', 'v147_rawvalue', 'v002_cilow', 'v136_other_data_2']]
df_county_stats = df_county_stats.rename(columns={'fipscode': 'county_fips', 'v052_rawvalue': 'percent_below_18', 'v053_rawvalue': 'percent_above_65', 'v044_rawvalue': 'income_inequality', 'v147_rawvalue': 'life_expectancy', 'v002_cilow': 'poor_fair_health', 'v136_other_data_2': 'over_crowding'}).set_index('county_fips')
df = df.join(df_county_stats)
df = df.reset_index()
df['county_fips'] = df['county_fips'].astype(str).str.rjust(5, '0')
plt.xlim(0, 1)
def show_values_on_bars(axs, h_v="v", space=0.4, text_size=10):
def _show_on_single_plot(ax):
if h_v == "v":
for p in ax.patches:
_x = p.get_x() + p.get_width() / 2
_y = p.get_y() + p.get_height()
value = p.get_height()
ax.text(_x, _y, value, ha="center", size=text_size)
elif h_v == "h":
for p in ax.patches:
_x = p.get_x() + p.get_width() + float(space)
_y = p.get_y() + p.get_height()- float(0.2)
value = p.get_width()
value = "{:.1%}".format(value)
ax.text(_x, _y, value, ha="left", size=text_size)
if isinstance(axs, np.ndarray):
for idx, ax in np.ndenumerate(axs):
_show_on_single_plot(ax)
else:
_show_on_single_plot(axs)
plt.figure(figsize=(10,10))
g=sns.barplot(x='likely_infected_low', y='county_state',data=df.sort_values(['likely_infected_low'], ascending=False).head(20), color="lightgreen")
show_values_on_bars(g, "h", space=0.01, text_size=10)
plt.xlim(0, 1.1)
plt.xlabel("Percentage of population infected")
plt.ylabel("County, State")
plt.title("Likely spread of Virus if spread is 28 fold")
plt.figure(figsize=(10, 5))
df_temp = df[df.confirmed > 500]
sns.distplot(df_temp.mortality, hist=True, kde=False, color='green', hist_kws={'edgecolor': 'black', 'linewidth': 1}, kde_kws={'linewidth': 2})
plt.title('Distribution of Mortality Rate')
plt.xlabel('Covid Mortality Rate') | code |
130007697/cell_1 | [
"text_plain_output_1.png"
] | import os
import numpy as np
import pandas as pd
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename)) | code |
130007697/cell_18 | [
"application_vnd.jupyter.stderr_output_3.png",
"text_plain_output_2.png",
"application_vnd.jupyter.stderr_output_1.png"
] | import cv2 as cv
import matplotlib.pyplot as plt
import torch
import cv2 as cv
import matplotlib.pyplot as plt
img = cv.imread('/kaggle/input/car-plate-detection/images/Cars0.png')
img = cv.cvtColor(img, cv.COLOR_BGR2RGB)
rec = cv.rectangle(img, (226, 125), (419, 173), (0, 250, 0), 2)
rec = cv.circle(rec, ((226 + 419) // 2, (125 + 173) // 2), 2, (255, 0, 0), 2)
import torch
yolo = torch.hub.load('ultralytics/yolov5', 'custom', path='/kaggle/working/yolov5/runs/train/exp/weights/best.pt')
img = '/kaggle/input/number-plate-detection/TEST/TEST.jpeg'
results = yolo(img)
cordinates = results.xyxy[0][:, :-1]
results.pandas().xyxy[0] | code |
130007697/cell_16 | [
"text_plain_output_1.png"
] | import torch
import torch
yolo = torch.hub.load('ultralytics/yolov5', 'custom', path='/kaggle/working/yolov5/runs/train/exp/weights/best.pt') | code |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.