Datasets:

---

suriyagunasekar

/

stackoverflow-python-with-meta-data

like 12

[ "Yes. What you're looking for is the module variable lsb0:\nbitstring.lsb0 = True\ndataBits = bitstring.BitArray('0b10100000')\nprint(dataBits[0])\n\nPrints out False.\n" ]

[ 0 ]

[]

[]

[ "bitstring", "python" ]

[ "It should be difference, because in csv all data are saved like strings, so if use index_col=0 here is correctly create FloatIndex, but columns names are strings, also data in columns should be parsed differently (e.g. if mixed strings and numeric):\nf = 'file.csv'\ndf.to_csv(f)\n\ndf = pd.read_csv(f, index_col=0)\nprint (df)\n 0.25 0.5 0.75 1 1.25\n0.1 0.876156 0.888282 0.904731 0.910351 0.906285\n0.2 0.810528 0.797068 0.806520 0.805698 0.806592\n0.3 0.696281 0.704307 0.703541 0.705199 0.708672\n0.4 0.601264 0.605194 0.607882 0.611617 0.612409\n0.5 0.502995 0.509210 0.513652 0.516065 0.519950\n\n\nprint (df.columns)\nIndex(['0.25', '0.5', '0.75', '1.0', '1.25'], dtype='object')\n\nAnother idea is use pickle, read_pickle and DataFrame.to_pickle for correct save DataFrames with columns and index:\nprint (df.columns)\nFloat64Index([0.25, 0.5, 0.75, 1.0, 1.25], dtype='float64')\n\nf = 'file'\ndf.to_pickle(f)\n\ndf1 = pd.read_pickle(f)\nprint (df1)\n 0.25 0.50 0.75 1.00 1.25\n0.1 0.876156 0.888282 0.904731 0.910351 0.906285\n0.2 0.810528 0.797068 0.806520 0.805698 0.806592\n0.3 0.696281 0.704307 0.703541 0.705199 0.708672\n0.4 0.601264 0.605194 0.607882 0.611617 0.612409\n0.5 0.502995 0.509210 0.513652 0.516065 0.519950\n\nprint (df1.columns)\nFloat64Index([0.25, 0.5, 0.75, 1.0, 1.25], dtype='float64')\n\nprint (df1.equals(df))\nTrue\n\n" ]

[ 1 ]

[]

[]

[ "dataframe", "numpy", "pandas", "python" ]

[ "You could create a fixture in the root of the tests folder (tests/conftest.py) to help you read any file relative to the tests folder:\nfrom pathlib import Path\nimport pytest\n\[email protected]()\ndef get_file():\n def _(file_path: str):\n return (Path(__file__).parent / file_path).read_text()\n\n return _\n\nIn any test you could then use the fixture to read the file for you:\ndef test_a(get_file):\n content = get_file('input/data.txt')\n ...\n\nIn this example the fixture returns a function that would read the content of any file within tests/ directory, but there are multiple other options:\n\nThe fixture can return a Path object instead of the content.\nIf you would always use the same file, it can return file object/content directly so within your test function you wouldn't need to call it.\n\n" ]

[ 1 ]

[]

[]

[ "pytest", "python" ]

[ "I guess you could multiple processes.\none main which define the future time stamp at which both action will be performed, and which passes this timestamp via a queue to the second.\nThen at the given time stamp both process execute their own tasks~\nseems like it's better to edit than to add another answer so I edit~\nYou can make 2 functions (A & B), one for each of your two actions.\ndef A(*args, **kwargs):\n get ICU data and save it to whatever format you need\ndef B(*args, **kwargs):\n get HTTP request data and save it to whatever format you need\n\nThen, you define when you want to call them and use for instance threading.Timer to call them. Timer takes as input how long from call it should wait before running a specific function.\nSo you do your math to define how long you want to wait before calling both functions together. For instance, something like :\nwaitTime = target_timeStamp - current_time\nand you call your functions\nthreading.Timer(waitTime, A).start()\nthreading.Timer(waitTime, B).start()\n\ndoc there for the details of Timer (at the bottom)\nhttps://docs.python.org/3/library/threading.html\nThe thing is that the two thread will be \"simultaneous\" modulo your interpreter and machine. If you get limited and need the time between the two functions to be reduced, then you need to do something similar, but before you must call your Timer from 2 different processes.\nAnd if that it not doing it for you, then I'd suggest to move to C/CPP if you really need to go as fast as the OS can do, but I am not sure it will be really relevant in your case, since anyway your http request will have some delay due to the ping.\nAlso you should save each data with its corresponding timestamp, so if you cannot match exactly the timings you want, you can also interpolate the values~\nHope this explanation helps\n" ]

[ 0 ]

[]

[]

[ "csv", "python", "windows_11" ]

[ "If your input is sorted by path (i.e. the same paths are sequential), you can use itertools.groupby:\nfrom itertools import groupby\nfrom os.path import dirname\n\nout = [list(g) for _,g in groupby(paths, dirname)]\n\nIf the paths are not sorted, you can use a dictionary as intermediate:\nout = {}\nfor p in paths:\n (out.setdefault(dirname(p), [])\n .append(p)\n )\n \nout = list(out.values())\n\nOutput:\n[['/var/lib/cons/states/tent1/tops-ok_2022_11_28',\n '/var/lib/cons/states/tent1/tops-ok_2022_11_27'],\n ['/var/lib/cons/states/tent2/tops-ok_2022_11_28',\n '/var/lib/cons/states/tent2/tops-ok_2022_11_27'],\n ['/var/lib/cons/states/tent3/tops-ok_2022_11_28',\n '/var/lib/cons/states/tent3/tops-ok_2022_11_27'],\n ['/var/lib/cons/states/tent4/tops-ok_2022_11_28',\n '/var/lib/cons/states/tent4/tops-ok_2022_11_27']]\n\nalternative with pathlib:\nfrom itertools import groupby\nfrom pathlib import Path\n\nout = [list(g) for _,g in groupby(paths, lambda x: Path(x).parent)]\n\n" ]

[ 6 ]

[]

[]

[ "path", "python" ]

[ "So the problem is that you added map as a static variable.\nYour patch works correctly as you can see here:\npatch actually works\nThe problem is that when it runs it's already too late because the map variable was already calculated (before the patch).\nIf you want you can move it to the init function and it will function correctly:\nclass WorkingDirectory:\n\ndef __init__(self, env: str):\n self.map = {\n \"foo\": \"path/to/foo\",\n \"bar\": \"path/to/bar\",\n \"baz\": os.environ.get(\"BAZ_PATH\")\n }\n self.env = env\n self.path = self.map[self.env]\n\nIf for some reason you wish to keep it static, you have to also patch the object itself.\nwriting something like this will do the trick:\nclass TestWorkingDirectory(unittest.TestCase):\[email protected](os.environ, {\"BAZ_PATH\": DUMMY_BAZ_PATH})\ndef test_controlled_baz(self):\n with mock.patch.object(WorkingDirectory, \"map\", {\n \"foo\": \"path/to/foo\",\n \"bar\": \"path/to/bar\",\n \"baz\": os.environ.get(\"BAZ_PATH\")\n }):\n wd = WorkingDirectory(\"baz\")\n self.assertEqual(wd.path, DUMMY_BAZ_PATH)\n\n", "That's not directly answer to your question but a valid answer either way imo:\nDon't try to patch that (it's possible, but harder and cumbersome).\nUse config file for your project.\ne.g. use pyproject.toml and inside configure the pytest extension:\n[tool.pytest.ini_options]\nenv=[\n\"SOME_VAR_FOR_TESTS=some_value_for_that_var\"\n]\n\n" ]

[ 5, 0 ]

[]

[]

[ "environment_variables", "python", "python_unittest", "unit_testing" ]

[ "If I understood you question right, then the df.applymap(mapping) function is what you are searching for. It applies the mapping (for example a lambda) to all cells of you data frame. Here is an example:\nimport pandas as pd\n\ndata = {\"c1\": [\"Test\", \"Test[ (Test)]\"],\n \"c2\": [\"Test\", \"Test[ (Test)]\"],\n \"c3\": [\"Test\", \"Test[ (Test)]\"]}\n\ndf = pd.DataFrame.from_dict(data)\n\nfiltered = df.applymap(lambda s : s.replace(\"[ (\", \"[(\"))\n\nprint(filtered)\n\nAnd this is the result:\n c1 c2 c3\n0 Test Test Test\n1 Test[(Test)] Test[(Test)] Test[(Test)]\n\n" ]

[ 0 ]

[]

[]

[ "dataframe", "pandas", "python" ]

[ "Did you try to convert your Json into a dictionary?\nfor instance using these resources:\nhttps://www.geeksforgeeks.org/convert-json-to-dictionary-in-python/\nYou convert to a dictionary, you edit the contents of the key you want, then save it back to Json if you need\n" ]

[ 1 ]

[]

[]

[ "json", "python", "replace" ]

[ "The hint at the end was really the issue. The error in actual was ValueError: setting an array element with a sequence. I used a function to pad all the rows with 0 with were less in size than row with max length.\n# Find length of max row\nmax = max(map(len, X_arr))\n\n# Pad the rows with 0\nX_arr_same = np.array([np.pad(v, (0, m - len(v)), 'constant') for v in X_arr])\n\nResult was:\nlen(X_arr_same[0]): 152576\nlen(X_arr_same[1]): 152576\nlen(X_arr_same[2]): 152576\n\nThis solved the issue and the block ran successfully,\nfrom sklearn import svm\nclf = svm.SVC()\nclf.fit(X_arr_same, y)\n\n" ]

[ 0 ]

[]

[]

[ "arrays", "machine_learning", "numpy", "pandas", "python" ]

[ "We are going to inspect how small sets uses 216 bytes.\nFirst of all a set object in python is represented by following C structure.\ntypedef struct {\n PyObject_HEAD\n\n Py_ssize_t fill; /* Number active and dummy entries*/\n Py_ssize_t used; /* Number active entries */\n\n /* The table contains mask + 1 slots, and that's a power of 2.\n * We store the mask instead of the size because the mask is more\n * frequently needed.\n */\n Py_ssize_t mask;\n\n /* The table points to a fixed-size smalltable for small tables\n * or to additional malloc'ed memory for bigger tables.\n * The table pointer is never NULL which saves us from repeated\n * runtime null-tests.\n */\n setentry *table;\n Py_hash_t hash; /* Only used by frozenset objects */\n Py_ssize_t finger; /* Search finger for pop() */\n\n setentry smalltable[PySet_MINSIZE];\n PyObject *weakreflist; /* List of weak references */\n} PySetObject;\n\nNow remember, getsizeof() calls the object’s __sizeof__ method and adds an additional garbage collector overhead if the object is managed by the garbage collector.\nOk, set implements the __sizeof__.\nstatic PyObject *\nset_sizeof(PySetObject *so, PyObject *Py_UNUSED(ignored))\n{\n Py_ssize_t res;\n\n res = _PyObject_SIZE(Py_TYPE(so));\n if (so->table != so->smalltable)\n res = res + (so->mask + 1) * sizeof(setentry);\n return PyLong_FromSsize_t(res);\n}\n\nNow lets inspect the line\nres = _PyObject_SIZE(Py_TYPE(so));\n\n_PyObject_SIZE is just a macro which expands to (typeobj)->tp_basicsize.\n#define _PyObject_SIZE(typeobj) ( (typeobj)->tp_basicsize )\n\nThis code is essentially trying to access the tp_basicsize slot to get the size in bytes of instances of the type which is just sizeof(PySetObject) in case of set.\nPyTypeObject PySet_Type = {\n PyVarObject_HEAD_INIT(&PyType_Type, 0)\n \"set\", /* tp_name */\n sizeof(PySetObject), /* tp_basicsize */\n 0, /* tp_itemsize */\n # Skipped rest of the code for brevity.\n\nI have modified the set_sizeof C function with the following changes.\nstatic PyObject *\nset_sizeof(PySetObject *so, PyObject *Py_UNUSED(ignored))\n{\n Py_ssize_t res;\n\n unsigned long py_object_head_size = sizeof(so->ob_base); // Because PyObject_HEAD expands to PyObject ob_base;\n unsigned long fill_size = sizeof(so->fill);\n unsigned long used_size = sizeof(so->used);\n unsigned long mask_size = sizeof(so->mask);\n unsigned long table_size = sizeof(so->table);\n unsigned long hash_size = sizeof(so->hash);\n unsigned long finger_size = sizeof(so->finger);\n unsigned long smalltable_size = sizeof(so->smalltable);\n unsigned long weakreflist_size = sizeof(so->weakreflist);\n int is_using_fixed_size_smalltables = so->table == so->smalltable;\n\n printf(\"| PySetObject Fields | Size(bytes) |\\n\");\n printf(\"|------------------------------------|\\n\");\n printf(\"| PyObject_HEAD | '%zu' |\\n\", py_object_head_size);\n printf(\"| fill | '%zu' |\\n\", fill_size);\n printf(\"| used | '%zu' |\\n\", used_size);\n printf(\"| mask | '%zu' |\\n\", mask_size);\n printf(\"| table | '%zu' |\\n\", table_size);\n printf(\"| hash | '%zu' |\\n\", hash_size);\n printf(\"| finger | '%zu' |\\n\", finger_size);\n printf(\"| smalltable | '%zu' |\\n\", smalltable_size); \n printf(\"| weakreflist | '%zu' |\\n\", weakreflist_size);\n printf(\"-------------------------------------|\\n\");\n printf(\"| Total | '%zu' |\\n\", py_object_head_size+fill_size+used_size+mask_size+table_size+hash_size+finger_size+smalltable_size+weakreflist_size);\n printf(\"\\n\");\n printf(\"Total size of PySetObject '%zu' bytes\\n\", sizeof(PySetObject));\n printf(\"Has set resized: '%s'\\n\", is_using_fixed_size_smalltables ? \"No\": \"Yes\");\n if(!is_using_fixed_size_smalltables) {\n printf(\"Size of malloc'ed table: '%zu' bytes\\n\", (so->mask + 1) * sizeof(setentry));\n }\n\n res = _PyObject_SIZE(Py_TYPE(so));\n if (so->table != so->smalltable)\n res = res + (so->mask + 1) * sizeof(setentry);\n return PyLong_FromSsize_t(res);\n}\n\nand compiling and running these changes gives me\n>>> import sys\n>>> \n>>> set_ = set()\n>>> sys.getsizeof(set_)\n| PySetObject Fields | Size(bytes) |\n|------------------------------------|\n| PyObject_HEAD | '16' |\n| fill | '8' |\n| used | '8' |\n| mask | '8' |\n| table | '8' |\n| hash | '8' |\n| finger | '8' |\n| smalltable | '128' |\n| weakreflist | '8' |\n-------------------------------------|\n| Total | '200' |\n\nTotal size of PySetObject '200' bytes\nHas set resized: 'No'\n216\n>>> set_.add(1)\n>>> set_.add(2)\n>>> set_.add(3)\n>>> set_.add(4)\n>>> set_.add(5)\n>>> sys.getsizeof(set_)\n| PySetObject Fields | Size(bytes) |\n|------------------------------------|\n| PyObject_HEAD | '16' |\n| fill | '8' |\n| used | '8' |\n| mask | '8' |\n| table | '8' |\n| hash | '8' |\n| finger | '8' |\n| smalltable | '128' |\n| weakreflist | '8' |\n-------------------------------------|\n| Total | '200' |\n\nTotal size of PySetObject '200' bytes\nHas set resized: 'Yes'\nSize of malloc'ed table: '512' bytes\n728\n\nThe return value is 216/728 bytes because sys.getsize add 16 bytes of GC overhead.\nBut the important thing to note here is this line.\n| smalltable | '128' |\n\nBecause for small tables(before the first resize) so->table is just a reference to fixed size(8) so->smalltable(No malloc'ed memory) so sizeof(PySetObject) is sufficient enough to get the size because it also includes the storage size( 128(16(size of setentry) * 8)).\nNow what happens when the resize occurs. It constructs entirely new table(malloc'ed) and uses that table instead of so->smalltables,this means that the sets which have resized also carry out a dead-weight of 128 bytes(Size of fixed size small table) along with the size of malloc'ed so->table.\nelse {\n newtable = PyMem_NEW(setentry, newsize);\n if (newtable == NULL) {\n PyErr_NoMemory();\n return -1;\n }\n }\n\n /* Make the set empty, using the new table. */\n assert(newtable != oldtable);\n memset(newtable, 0, sizeof(setentry) * newsize);\n so->mask = newsize - 1;\n so->table = newtable;\n\n" ]

[ 7 ]

[]

[]

[ "cpython", "memory", "python", "python_internals", "set" ]

[ "@mozway is right you better give an example input and expected result.\nAnyway if you're dealing with a variable number of columns in the input please refer to Handling Variable Number of Columns with Pandas - Python\nBest\n" ]

[ 0 ]

[]

[]

[ "database", "dataframe", "python", "raw_data" ]

[ "\nIs Numpy doing something special when printing the single-precision float or there is something I missed?\n\n0.2581568658351898 is not exactly encodable as a 32-bit float.\nThe closest is 0.2581568658351898193359375 or 0x1.085a46p-2\nWhen 0.2581568658351898193359375 is printed with reduced precision, the result is 0.25815687\n\n0.2581568 658351898 Source code\n0.2581568 658351898193359375 True value\n0.2581568 7 Output \n\n", "You can force the number of digits to be printed using f-strings\nprint(f\"{np.float32(0.2581568658351898):.25f}\")\n\nwhere :.25f tells the interpreter to print the value as a floating point number with 25 decimal digits.\n" ]

[ 1, 0 ]

[]

[]

[ "floating_point", "ieee_754", "numpy", "python" ]

[ "If\ncert = crypto.dump_certificate(crypto.FILETYPE_PEM, k)\n\ndoesn't do what you want, then it doesn't look like pyOpenSSL supports public key dumping. There is an unmerged branch here that adds that functionality but I can't claim that it does what is purports.\n", "Updated:\nNow it has the method to get public key directly.\nkey = crypto.PKey()\nkey.generate_key(crypto.TYPE_RSA, 2048)\npublickey_contents = crypto.dump_publickey(crypto.FILETYPE_PEM, key)\n\nwith the method dump_publickey you can get what you want.\n" ]

[ 3, 1 ]

[]

[]

[ "openssl", "pyopenssl", "python" ]

[ "You can make an iteration loop by yourself. You could easily add a loop index independent from next calls, so that you could even use a skip condition that uses the current index.\nskip_iteration = True\nit = iter(range(10))\niterating = True\nvalue = next(it)\nwhile iterating:\n try:\n print(value, end=' ')\n if skip_iteration:\n # iteration skip code\n skip_iteration = False\n else:\n # normal iteration code\n value = next(it)\n except StopIteration:\n iterating = False\n\nHere only the first iteration is skipped, which outputs :\n0 0 1 2 3 4 5 6 7 8 9\n\nThis code relies on the fact that the next function raises StopIteration if it reaches the end of the iterator.\n" ]

[ 0 ]

[ "If the condition in the if statement is somehow related to the iterator i itself then the loop might not end but if the condition is not depended on i then there shall not be any problem\nyou can also try skipping that particular iteration by using continue.\n" ]

[ -1 ]

[ "loops", "python" ]

[ "You have wrong record_path, which should be ['Report Details', 'report Accessible'].\njs_obj = json.loads(json_str.replace('r2', 'r1')) # keep columns consistent\ndf = pd.json_normalize(js_obj, ['Report Details', 'report Accessible'], \n ['TileName', ['Report Details', 'r1name'], ['Report Details', 'r1link']])\ndf.columns = ['ReportAccssible', 'TileName', 'ReportName', 'ReportLink']\n\nYou will get what you want.\n ReportAccssible TileName ReportName ReportLink\n0 operations Master Primary link1\n1 Sales Master Primary link1\n2 operations Master Secondry link2\n3 Sales Master Secondry link2\n\n" ]

[ 1 ]

[]

[]

[ "pandas", "python" ]

[ "Selenium in Python has By.TAG_NAME attribute.\nBy.tagName is Selenium Java syntax.\n", "it's self.driver.find_element(By.TAG_NAME,\"iframe\")\n" ]

[ 0, 0 ]

[]

[]

[ "automation", "python", "selenium", "web_scraping" ]

[ "The problem is that you use the same col_stats twice.\nIt becomes clear if you have a bit of tracing :\n+ print(f\"comparing {row[col]=!r} > {max_int}\")\n if row[col] > max_int:\n max_int = row[col]\n+ print(f\"{max_int=!r}\")\n\n+ print(f\"comparing {row[col]=!r} < {min_int}\")\n if row[col] < min_int:\n min_int = row[col]\n+ print(f\"{min_int=!r}\")\n\ncomparing row[col]=20 > -1000000\nmax_int=20\ncomparing row[col]=20 < 1000000\nmin_int=20\ncomparing row[col]=21 > 20\nmax_int=21\ncomparing row[col]=21 < 20\ncomparing row[col]=20 > 21\ncomparing row[col]=20 < 20\ncomparing row[col]=20 > 21\ncomparing row[col]=20 < 20\n{'Age': {'max': 21, 'min': 20}}\ncomparing row[col]=20000 > -1000000\nmax_int=20000\ncomparing row[col]=20000 < 1000000\nmin_int=20000\ncomparing row[col]=25000 > 20000\nmax_int=25000\ncomparing row[col]=25000 < 20000\ncomparing row[col]=10000 > 25000\ncomparing row[col]=10000 < 20000\nmin_int=10000\ncomparing row[col]=15000 > 25000\ncomparing row[col]=15000 < 10000\n{'Age': {'max': 25000, 'min': 10000}, 'Salary': {'max': 25000, 'min': 10000}}\n\nThe calculations are indeed correct.\nBut what happens after is the error :\n col_stats[\"max\"] = max_int\n col_stats[\"min\"] = min_int\n stats_dict[header[col]] = col_stats\n\nThe first time it will add two entries (max: 21 and min: 21) to the col_stats dict, and add one entry to stats_dict[\"Age\"]. Ok.\nThe second time it will change the two entries of the same col_stats dictionary, thus changing what is accessible through stats_dict[\"Age\"].\nYour problem is you have mutable data, and you mutated it unintentionally.\nOne easy solution is to copy the col_stats :\n stats_dict[header[col]] = col_stats.copy()\n\nBut you can also simply do not share the same dict :\n# remove the declaration earlier\n# [...]\n\n+ col_stats = {} # A NEW DICT !\n col_stats[\"max\"] = max_int\n col_stats[\"min\"] = min_int\n\nBeware of sharing mutable data !\n" ]

[ 0 ]

[]

[]

[ "dictionary", "python" ]

[ "Your pattern matches at least 2 characters, being a digit and 1 or more times one of \\d , . \\s\nYou can match either a space or ( and then capture a single digit followed by optionally repeating the chars in the character class.\n[\\s(](\\d[\\d,.\\s]*)\n\nSee a regex demo.\nIf you don't want trailing spaces, dots or comma's:\n[\\s(](\\d+(?:[\\d,.\\s]*\\d)?)\\b\n\nExplanation\n\n[\\s(] Match either a whitespace char or (\n( Capture group 1\n\n\\d+ Match 1+ digits\n(?:[\\d,.\\s]*\\d)? Optionally match one of the chars in the character class followed by a digit\n\n\n) Close group 1\n\\b A word boundary to prevent a partial word match\n\nRegex demo\n" ]

[ 2 ]

[]

[]

[ "python", "regex" ]

[ "You need to properly use subqueries (google it!). Try something like this, it might work (and please fix your variable names, qwertyu is not good, should be descriptive like unique_id, uname, dname, first_name, etc.)\ndef insrtDirector(q,w,e,r,t,y,u):\n # Assume w and e are always subqueries with one result\n # Also assume no outside source can supply w and e (must be sanitized to not be subject to sql injection)\n insrt = \"INSERT INTO Director VALUES (%s,{uname_subquery},{dname_subquery},%s,%s,%s,%s)\".format(\n uname_subquery=w,\n dmane_subquery=e,\n )\n # This print command is just temporary so you can see what it looks like\n print(\"insert command:\", insrt)\n sql1 = (q,r,t,y,u)\n cur.execute(insrt, sql1)\n\ninsrtDirector(\n uuid.uuid4(),\n \"(SELECT UName from University WHERE UName = 'University College London')\",\n \"(SELECT DName from Department WHERE DName = 'English')\",\n 'Christopher (3)',\n 'Nolan (3)',\n 1970,\n 'Westminster, London, United Kingdom',\n)\n\n" ]

[ 0 ]

[]

[]

[ "postgresql", "psycopg2", "python" ]

[ "The solution I figured out\n# read data from a dataframe into a dictionary\nwords_data_dict = {row.French: row.English for (index, row) in data.iterrows()}\nprint(words_data_dict)\n\n# convert the dictionary to a list\nlist_of_entry = list(words_data_dict.items())\nprint(list_of_entry)\n\n# generate a random row from the dictionary\nrandom_data = random.choice(list_of_entry)\n\n# specify Key using Index\nprint(random_data[0])\n\n" ]

[ 0 ]

[]

[]

[ "dictionary", "dictionary_comprehension", "nested", "python" ]

[ "I use such an environment to develop python app inside a container.\nimage_create.sh # script to create image to use it local and on the server\n\nimage_dockerfile # dockerfile with script how to create an image\n\ncontainer_create.sh # create named container from image\n\ncontainer_restart.sh # restart existing container\n\ncontainer_stop.sh # stop existing container\n\nExamples:\nimage_dockerfile :\nFROM python:3.9.15-slim-bullseye\nUSER root\nRUN pip3 install requests telethon\nRUN apt-get update\nRUN apt-get --assume-yes install git\n\nimage_create.sh :\ndocker rmi python_find_a_job:lts\ndocker build . -f python_find_a_job -t python_find_a_job:lts\n\ncontainer_create.sh :\ndocker rm -f python_find_a_job\ndocker run -t -d --name python_find_a_job -i python_find_a_job:lts\ndocker ps -aq\n\ncontainer_restart.sh :\ndocker container restart python_find_a_job\ndocker ps -aq\n\ncontainer_stop.sh :\ndocker stop python_find_a_job\ndocker ps -aq\n\nFor VSCcode:\na) Prepare files (see above).\nb) Run:\nimage_create.sh\ncontainer_create.sh\nc) Open project folder in VSCode\nd) Click on left bottom green / Attach to running container / select container name (python_find_a_job).\ne) Clone repository.\nf) Install extension 'Python'.\nNow you can run and debug inside the container.\nAfter work:\ngit push\ncontainer_stop.sh\nBefore work:\ncontainer_restart.sh\ngit pull\n", "Visual Studio Code, and Docker Desktop each offer a feature called \"Dev Containers\" (VSCode) or \"Dev Environments\" (DD), or CodeSpaces (GitHub)\nIn this approach, a Docker container is created by scanning the source, and generating a container that contains the development toolchain. Visual Studio then attaches to the container, and allows you to develop even though you do not have node/python3/dotnet/etc. installed on your development PC.\nThe xhr error indicates something went wrong downloading a scanning image or something else is strange about your project.\nThere is an optional Dockerfile that can be created if scanning fails to find an image, that is normally kept in a .devcontainers / .devenvironments folder depending on which of Docker / VSCode / GitHub / other you are using.\nYour project might also have one (or more) Dockerfile's that are used to package the running app up as a docker image, so don't be confused if you end up with 2. Thats not a problem and is expected really.\n", "I have got good results with VSCode using docker compose with compose.yml. But in this case, docker compose will create a new container from image every time.\ncompose.yml example:\nversion: '3.3'\nservices:\n my_container:\n image: python_find_a_job:lts\n stdin_open: true # docker run -i\n tty: true # docker run -t\n container_name: find_a_job\n network_mode: \"host\"\n volumes:\n - ~/.PYTHON/find_a_job:/opt/find_a_job\n mongodb_container:\n image: mongo:latest\n environment:\n MONGO_INITDB_ROOT_USERNAME: root\n MONGO_INITDB_ROOT_PASSWORD: pass\n network_mode: \"host\"\n ports:\n - 27017:27017\n volumes:\n - mongodb_data_container:/data/db\nvolumes:\n mongodb_data_container:\n\nBefore work:\ndocker compose up\n\nVSCode / Attach to Running Container ...\nFile / Open Folder (/opt/find_a_job from above example of compose.yml).\nAfter work:\ndocker compose down\n\nP.S. I have tried PyCharm community ed. But it can't add Docker interpreter. Only PyCharm Professional ed. can do it.\nSo I use VSCode to run & debug (F5) inside a docker container. There was some minor problems with setting VSCode tasks, extensions etc. But main part about run & debug inside a container works perfect. Changes in a project can be saved locally or pushed to git by VSCode or manually.\n" ]

[ 1, 0, 0 ]

[]

[]

[ "docker", "python", "visual_studio_code" ]

[ "Intel has release their MKL under a community license, which is free, with limited technical support. Currently MKL under the Community License is available for Linux and Windows, and it is expected they will provide a version for Mac OS X soon.\nhttps://software.intel.com/en-us/comment/1839012\nIn one of their recent webinars, I asked for their plans for a Mac OS X MKL under the community license. They say it is coming soon.\nUpdate 2:\nContinuum provides Anaconda Python with Intel MKL included for all platforms.\nhttps://www.anaconda.com/blog/developer-blog/anaconda-25-release-now-mkl-optimizations/\nIntel even makes it easy to compile and link against the MKL from the Anaconda Python distribution. \nhttps://software.intel.com/en-us/articles/using-intel-distribution-for-python-with-anaconda\nUpdate:\nIt now appears that Intel has their own version of Python that they are providing to beta testers.\nhttps://software.intel.com/en-us/forums/intel-distribution-for-python/topic/581593\n", "I know that this is an older question, but in case it comes up for someone who is searching: I would recommend trying out anaconda. For $29.00 they have an add-on that includes mkl optimized numpy + scipy.\n", "MacPorts seems to have recently added an MKL variant to their NumPy port (as well as to SciPy and PyTorch). Tested on my 16” MacBook Pro 2019 with 2.4GHz 8-core Intel Core i9 and macOS Ventura 13.0.1, Numpy with MKL is significantly faster than Numpy with the Accelerate framework, which is another fast replacement for OpenBLAS that is built into macOS. I tested using this code which I got from Puget Systems:\nimport numpy as np\nimport time\nn = 20000\nA = np.random.randn(n,n).astype('float64')\nB = np.random.randn(n,n).astype('float64')\nstart_time = time.time()\nnrm = np.linalg.norm(A@B)\nprint(\" took {} seconds \".format(time.time() - start_time))\nprint(\" norm = \",nrm)\n\nThe result of my testing is that Numpy with mkl took ~47 seconds while Numpy with accelerate took ~66 seconds. Accelerate also used more threads.\nTo install this with MacPorts you first have to install MacPorts, then run sudo port install py310-numpy -openblas +mkl in the terminal.\n" ]

[ 4, 3, 1 ]

[]

[]

[ "intel_mkl", "macos", "numpy", "python", "python_3.3" ]

[ "CAUTION: NEVER publish any API key\nI was able to run your code without error.\nimport requests\nfrom urllib.parse import urlencode\n...\noperationName = \"highrollerSportBets\"\nvariables = {\"limit\":10}\n\nurl = 'https://stake.com/_api/graphql'\nsa_key = 'xxxx'\nsa_api = 'https://api.scrapingant.com/v2/general'\nqParams = {'url': url, 'x-api-key': sa_key}\nreqUrl = f'{sa_api}?{urlencode(qParams)}' # just change this\n\nprint(reqUrl)\nr = requests.post(url=reqUrl, json={\"query\": body, \"operationName\": operationName, \"variables\": variables}, headers=headers)\nprint(r.text)\n\nRun\n$ python3 --version\nPython 3.10.6\n\n$ python3 so.py \nhttps://api.scrapingant.com/v2/general?url=https%3A%2F%2Fstake.com%2F_api%2Fgraphql&x-api-key=XXX\n{\"detail\":\"Our browser was detected by target site. Try again later or contact us [email protected]\"}\n\n", "I am a Stake.com player too, and i use the API (and a VPN), every day to automate some bets.\nTo bypass cloudflare, you just need to get one cookie, named 'cf_clearance', and then use it requests. That's all!\nrequest_data = {\n \"query\": body,\n \"variables\": {\"limit\": 20,\"offset\": 0},\n}\n\ncookies = {\"cf_clearance\": cf_clearance_value} \nsession = requests.Session()\nsession.headers[\"x-access-token\"] = \"xxxx\"\nsession.headers[\"user-agent\"] = \"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.113 Safari/537.36\"\ntime.sleep(1)\nresponse = session.post(\"https://stake.com/_api/graphql\", cookies=cookies,json=request_data)\n\n" ]

[ 0, 0 ]

[]

[]

[ "graphql", "python", "python_3.x" ]

[ "If you want a set-like thing for which items can appear with multiplicity greater than 1, then you could use a multiset. These can be represented by Counter objects. There is no built-in intersection method for those, but you could write a function which computes it by taking the min of two counts:\nfrom collections import Counter\n\ndef intersection(s1,s2):\n '''intersection of multisets s1,s2'''\n d = {}\n for i in s1:\n c = min(s1[i],s2[i])\n if c > 0:\n d[i] = c\n return Counter(d)\n\n#test:\n\ns1 = Counter([1, 1, 2, 3])\ns2 = Counter([4, 5, 1, 1])\n\nprint(s1)\nprint(s2)\nprint(intersection(s1,s2))\n\nOutput:\nCounter({1: 2, 2: 1, 3: 1})\nCounter({1: 2, 4: 1, 5: 1})\nCounter({1: 2})\n\n" ]

[ 2 ]

[]

[]

[ "intersection", "python", "set" ]

[ "I think this post is probably what you need. With np.reshape() you can just have your list filled with all the values you need and do the reshaping after in a single step.\n" ]

[ 0 ]

[]

[]

[ "list", "python" ]

[ "There is an issue with the indentation here:\nfor date in ['20210801','20210807','20210814']:\n addingstep = addSteps(date)\n steps.append(addingstep)\n steps =json.dumps(steps)\n\nFix:\nfor date in ['20210801','20210807','20210814']:\n addingstep = addSteps(date)\n steps.append(addingstep)\n\nsteps = json.dumps(steps)\n\n" ]

[ 2 ]

[]

[]

[ "amazon_emr", "aws_lambda", "list", "python", "python_3.x" ]

[ "What about sampling the combinations of indices of the True values:\nfrom itertools import combinations\nimport numpy as np\n\na = np.arange(n1+n2)\n\nout = [np.isin(a, x).tolist() for x in combinations(range(n1+n2), r=n1)]\n\nOutput:\n[[True, True, False, False, False],\n [True, False, True, False, False],\n [True, False, False, True, False],\n [True, False, False, False, True],\n [False, True, True, False, False],\n [False, True, False, True, False],\n [False, True, False, False, True],\n [False, False, True, True, False],\n [False, False, True, False, True],\n [False, False, False, True, True]]\n\n" ]

[ 3 ]

[]

[]

[ "numpy", "python", "python_itertools" ]

[ "This is probably a late answer, but I finally got it working. All credit goes to Sturla Molden at this link.\nThe key is, note that double** is an array of type np.uintp. Therefore, we have\nxpp = (x.ctypes.data + np.arange(x.shape[0]) * x.strides[0]).astype(np.uintp)\ndoublepp = np.ctypeslib.ndpointer(dtype=np.uintp)\n\nAnd then use doublepp as the type, pass xpp in. See full code attached.\nThe C code:\n// dummy.c \n#include <stdlib.h> \n\n__declspec(dllexport) void foobar(const int m, const int n, const \ndouble **x, double **y) \n{ \n size_t i, j; \n for(i=0; i<m; i++) \n for(j=0; j<n; j++) \n y[i][j] = x[i][j]; \n} \n\nThe Python code:\n# test.py \nimport numpy as np \nfrom numpy.ctypeslib import ndpointer \nimport ctypes \n\n_doublepp = ndpointer(dtype=np.uintp, ndim=1, flags='C') \n\n_dll = ctypes.CDLL('dummy.dll') \n\n_foobar = _dll.foobar \n_foobar.argtypes = [ctypes.c_int, ctypes.c_int, _doublepp, _doublepp] \n_foobar.restype = None \n\ndef foobar(x): \n y = np.zeros_like(x) \n xpp = (x.__array_interface__['data'][0] \n + np.arange(x.shape[0])*x.strides[0]).astype(np.uintp) \n ypp = (y.__array_interface__['data'][0] \n + np.arange(y.shape[0])*y.strides[0]).astype(np.uintp) \n m = ctypes.c_int(x.shape[0]) \n n = ctypes.c_int(x.shape[1]) \n _foobar(m, n, xpp, ypp) \n return y \n\nif __name__ == '__main__': \n x = np.arange(9.).reshape((3, 3)) \n y = foobar(x) \n\nHope it helps,\nShawn\n", "#include <stdio.h>\n\nvoid test(double (*in_array)[3], int N){\n int i, j;\n\n for(i = 0; i < N; i++){\n for(j = 0; j < N; j++){\n printf(\"%e \\t\", in_array[i][j]);\n }\n printf(\"\\n\");\n }\n}\n\nint main(void)\n{\n double a[][3] = {\n {1., 2., 3.},\n {4., 5., 6.},\n {7., 8., 9.},\n };\n\n test(a, 3);\n return 0;\n}\n\nif you want to use a double ** in your function, you must pass an array of pointer to double (not a 2d array):\n#include <stdio.h>\n\nvoid test(double **in_array, int N){\n int i, j;\n\n for(i = 0; i < N; i++){\n for(j = 0; j< N; j++){\n printf(\"%e \\t\", in_array[i][j]);\n }\n printf(\"\\n\");\n }\n}\n\nint main(void)\n{\n double a[][3] = {\n {1., 2., 3.},\n {4., 5., 6.},\n {7., 8., 9.},\n };\n double *p[] = {a[0], a[1], a[2]};\n\n test(p, 3);\n return 0;\n}\n\nAnother (as suggested by @eryksun): pass a single pointer and do some arithmetic to get the index:\n#include <stdio.h>\n\nvoid test(double *in_array, int N){\n int i, j;\n\n for(i = 0; i < N; i++){\n for(j = 0; j< N; j++){\n printf(\"%e \\t\", in_array[i * N + j]);\n }\n printf(\"\\n\");\n }\n}\n\nint main(void)\n{\n double a[][3] = {\n {1., 2., 3.},\n {4., 5., 6.},\n {7., 8., 9.},\n };\n\n test(a[0], 3);\n return 0;\n}\n\n", "While the reply might be rather late, I hope it could help other people with the same problem.\nAs numpy arrays are internally saved as 1d arrays, one can simply rebuild 2d shape in C. Here is a small MWE:\n// libtest2d.c\n#include <stdlib.h> // for malloc and free\n#include <stdio.h> // for printf\n\n// create a 2d array from the 1d one\ndouble ** convert2d(unsigned long len1, unsigned long len2, double * arr) {\n double ** ret_arr;\n\n // allocate the additional memory for the additional pointers\n ret_arr = (double **)malloc(sizeof(double*)*len1);\n\n // set the pointers to the correct address within the array\n for (int i = 0; i < len1; i++) {\n ret_arr[i] = &arr[i*len2];\n }\n\n // return the 2d-array\n return ret_arr;\n}\n\n// print the 2d array\nvoid print_2d_list(unsigned long len1,\n unsigned long len2,\n double * list) {\n\n // call the 1d-to-2d-conversion function\n double ** list2d = convert2d(len1, len2, list);\n\n // print the array just to show it works\n for (unsigned long index1 = 0; index1 < len1; index1++) {\n for (unsigned long index2 = 0; index2 < len2; index2++) {\n printf(\"%1.1f \", list2d[index1][index2]);\n }\n printf(\"\\n\");\n }\n\n // free the pointers (only)\n free(list2d);\n}\n\nand \n# test2d.py\n\nimport ctypes as ct\nimport numpy as np\n\nlibtest2d = ct.cdll.LoadLibrary(\"./libtest2d.so\")\nlibtest2d.print_2d_list.argtypes = (ct.c_ulong, ct.c_ulong,\n np.ctypeslib.ndpointer(dtype=np.float64,\n ndim=2,\n flags='C_CONTIGUOUS'\n )\n )\nlibtest2d.print_2d_list.restype = None\n\narr2d = np.meshgrid(np.linspace(0, 1, 6), np.linspace(0, 1, 11))[0]\n\nlibtest2d.print_2d_list(arr2d.shape[0], arr2d.shape[1], arr2d)\n\nIf you compile the code with gcc -shared -fPIC libtest2d.c -o libtest2d.so and then run python test2d.py it should print the array.\nI hope the example is more or less self-explaining. The idea is, that the shape is also given to the C-Code which then creates a double ** pointer for which the space for the additional pointers is reserved. And these then are then set to point to the correct part of the original array.\nPS: I am rather a beginner in C so please comment if there are reasons not to do this.\n", "Here i hava pass two 2d numpy array and print value of one array for the reference\nyou can use and write your own logic in cpp\ncpp_function.cpp\ncompile it using : g++ -shared -fPIC cpp_function.cpp -o cpp_function.so\n#include <iostream>\nextern \"C\" {\nvoid mult_matrix(double *a1, double *a2, size_t a1_h, size_t a1_w, \n size_t a2_h, size_t a2_w, int size)\n{\n //std::cout << \"a1_h & a1_w\" << a1_h << a1_w << std::endl; \n //std::cout << \"a2_h & a2_w\" << a2_h << a2_w << std::endl; \n for (size_t i = 0; i < a1_h; i++) {\n for (size_t j = 0; j < a1_w; j++) {\n printf(\"%f \", a1[i * a1_h + j]);\n }\n printf(\"\\n\");\n }\n printf(\"\\n\");\n }\n\n}\n\nPython File\nmain.py\nimport ctypes\nimport numpy\nfrom time import time\n\nlibmatmult = ctypes.CDLL(\"./cpp_function.so\")\nND_POINTER_1 = numpy.ctypeslib.ndpointer(dtype=numpy.float64, \n ndim=2,\n flags=\"C\")\nND_POINTER_2 = numpy.ctypeslib.ndpointer(dtype=numpy.float64, \n ndim=2,\n flags=\"C\")\nlibmatmult.mult_matrix.argtypes = [ND_POINTER_1, ND_POINTER_2, ctypes.c_size_t, ctypes.c_size_t]\n# print(\"-->\", ctypes.c_size_t)\n\ndef mult_matrix_cpp(a,b):\n shape = a.shape[0] * a.shape[1]\n libmatmult.mult_matrix.restype = None\n libmatmult.mult_matrix(a, b, *a.shape, *b.shape , a.shape[0] * a.shape[1])\n\nsize_a = (300,300)\nsize_b = size_a\n\na = numpy.random.uniform(low=1, high=255, size=size_a)\nb = numpy.random.uniform(low=1, high=255, size=size_b)\n\nt2 = time()\nout_cpp = mult_matrix_cpp(a,b)\nprint(\"cpp time taken:{:.2f} ms\".format((time() - t2) * 1000))\nout_cpp = numpy.array(out_cpp).reshape(size_a[0], size_a[1])\n\n" ]

[ 28, 2, 0, 0 ]

[]

[]

[ "c", "ctypes", "numpy", "python" ]

[ "Your code is correct.\nPython tries to be helpful by displaying bytes that map to an ASCII character as that character. For example, \\x2c maps to ,.\n>>> b',' == b'\\x2c'\nTrue\n\n" ]

[ 0 ]

[]

[]

[ "byte", "hex", "python", "python_2.7", "python_3.x" ]

[ "OOM stands for \"out of memory\". Your GPU is running out of memory, so it can't allocate memory for this tensor. There are a few things you can do:\n\nDecrease the number of filters in your Dense, Conv2D layers\nUse a smaller batch_size (or increase steps_per_epoch and validation_steps)\nUse grayscale images (you can use tf.image.rgb_to_grayscale)\nReduce the number of layers\nUse MaxPooling2D layers after convolutional layers\nReduce the size of your images (you can use tf.image.resize for that)\nUse smaller float precision for your input, namely np.float32\nIf you're using a pre-trained model, freeze the first layers (like this)\n\nThere is more useful information about this error:\nOOM when allocating tensor with shape[800000,32,30,62]\n\nThis is a weird shape. If you're working with images, you should normally have 3 or 1 channel. On top of that, it seems like you are passing your entire dataset at once; you should instead pass it in batches.\n", "From [800000,32,30,62] it seems your model put all the data in one batch.\nTry specified batch size like\nhistory = model.fit([trainimage, train_product_embd],train_label, validation_data=([validimage,valid_product_embd],valid_label), epochs=10, steps_per_epoch=100, validation_steps=10, batch_size=32)\n\nIf it still OOM then try reduce the batch_size\n", "Happened to me as well. \nYou can try reducing trainable parameters by using some form of Transfer Learning - try freezing the initial few layers and use lower batch sizes. \n", "I think the most common reason for this case to arise would be the absence of MaxPooling layers.\nUse the same architecture, but add atleast one MaxPool layer after Conv2D layers. This might even improve the overall performance of the model.\nYou can even try reducing the depth of the model, i.e., remove the unnecessary layers and optimize.\n" ]

[ 56, 2, 0, 0 ]

[]

[]

[ "deep_learning", "keras", "machine_learning", "python", "tensorflow" ]

[ "There are two errors that prevented this. First, you need to assign the edge attributes to the graph when defining it. Second, you need to use nx.draw_networkx_edge_labels() and not nx.draw_networkx_labels(). The latter is for node labels, not edge labels.\nimport pandas as pd\nimport networkx as nx\n\ndf = pd.DataFrame([[1, 2, 4], [3, 4, 2], [1, 3, 1]], columns=['A', 'B', 'Weight'])\n\nG = nx.from_pandas_edgelist(df, source='A',\n target='B',\n create_using=nx.DiGraph(),\n edge_attr='Weight')\nweight = nx.get_edge_attributes(G, 'Weight')\npos = nx.circular_layout(G, scale=1)\nnx.draw(G, pos, with_labels=True)\nnx.draw_networkx_nodes(G, pos, node_size=300)\nnx.draw_networkx_edges(G, pos=pos, edgelist=G.edges(), edge_color='black')\nnx.draw_networkx_edge_labels(G, pos=pos, edge_labels=weight)\nplt.show()\n\nplots\n\n" ]

[ 1 ]

[]

[]

[ "networkx", "pandas", "python" ]

[ "Okay, so I solved the issue after a non-zero amount of googling and messing up my pulseaudio installation enough times to warrant a reinstall more often than I would like to admit.\nrpi_ws281x requires the PWM pin (GPIO 18 on the Pi Zero 2w) for which it needs sudo privileges. So the solution is to run pulseaudio as a system service, instead of a user service.\nNow, the Pulseaudio creators warn against doing this, but in this case it is alright, since the Pi (in this case) is not really meant to have a user logged in anyway.\nSetting up pulseaudio as a system service, is not very complicated (see this and this):\nFirst, disable user mode (if it was enabled):\nsudo systemctl --global disable pulseaudio.service pulseaudio.socket\nCreate a systemd file (/etc/systemd/system/pulseaudio.service):\n[Unit]\nDescription=PulseAudio Daemon\n \n[Install]\nWantedBy=multi-user.target\n \n[Service]\nType=simple\nPrivateTmp=true\nExecStart=/usr/bin/pulseaudio --system --realtime --disallow-exit --no-cpu-limit \n\nAdd the pulse and pi user to the required groups:\nsudo usermod -a -G audio pulse #add pulse to audio group\nsudo usermod -a -G pulse-access pi #add pi to pulse-access group\nsudo usermod -a -G pulse-access root #add root to pulse-access group\n\nAnd finally, start the service:\nsudo systemctl enable pulseaudio.service\nsudo systemctl start pulseaudio.service\n\nNow, scripts run with sudo should be able to interact with pulseaudio. I have only verified this with PyAudio, but it should work for others too.\n" ]

[ 0 ]

[]

[]

[ "pulseaudio", "pyaudio", "python", "raspberry_pi", "sudo" ]

[ "Here's an example of how you can use a multidimensional array with Python and ctypes. \nI wrote the following C code, and used gcc in MinGW to compile this to slib.dll:\n#include <stdio.h>\n\ntypedef struct TestStruct {\n int a;\n float array[30][4];\n} TestStruct;\n\nextern void print_struct(TestStruct *ts) {\n int i,j;\n for (j = 0; j < 30; ++j) {\n for (i = 0; i < 4; ++i) {\n printf(\"%g \", ts->array[j][i]);\n }\n printf(\"\\n\");\n }\n}\n\nNote that the struct contains a 'two-dimensional' array.\nI then wrote the following Python script:\nfrom ctypes import *\n\nclass TestStruct(Structure):\n _fields_ = [(\"a\", c_int),\n (\"array\", (c_float * 4) * 30)]\n\nslib = CDLL(\"slib.dll\")\nslib.print_struct.argtypes = [POINTER(TestStruct)]\nslib.print_struct.restype = None\n\nt = TestStruct()\n\nfor i in range(30):\n for j in range(4):\n t.array[i][j] = i + 0.1*j\n\nslib.print_struct(byref(t))\n\nWhen I ran the Python script, it called the C function, which printed out the contents of the multidimensional array: \nC:\\>slib.py\n0.1 0.2 0.3 0.4\n1.1 1.2 1.3 1.4\n2.1 2.2 2.3 2.4\n3.1 3.2 3.3 3.4\n4.1 4.2 4.3 4.4\n5.1 5.2 5.3 5.4\n... rest of output omitted\n\nI've used Python 2, whereas the tags on your question indicate that you're using Python 3. However, I don't believe this should make a difference.\n", "Here i hava pass two 2d numpy array and print value of one array for the reference\nyou can use this and write multidimensional array\ncpp_function.cpp\ncompile it using : g++ -shared -fPIC cpp_function.cpp -o cpp_function.so\n#include <iostream>\nextern \"C\" {\nvoid mult_matrix(double *a1, double *a2, size_t a1_h, size_t a1_w, \n size_t a2_h, size_t a2_w, int size)\n{\n //std::cout << \"a1_h & a1_w\" << a1_h << a1_w << std::endl; \n //std::cout << \"a2_h & a2_w\" << a2_h << a2_w << std::endl; \n for (size_t i = 0; i < a1_h; i++) {\n for (size_t j = 0; j < a1_w; j++) {\n printf(\"%f \", a1[i * a1_h + j]);\n }\n printf(\"\\n\");\n }\n printf(\"\\n\");\n }\n\n}\n\nPython File\nmain.py\nimport ctypes\nimport numpy\nfrom time import time\n\nlibmatmult = ctypes.CDLL(\"./cpp_function.so\")\nND_POINTER_1 = numpy.ctypeslib.ndpointer(dtype=numpy.float64, \n ndim=2,\n flags=\"C\")\nND_POINTER_2 = numpy.ctypeslib.ndpointer(dtype=numpy.float64, \n ndim=2,\n flags=\"C\")\nlibmatmult.mult_matrix.argtypes = [ND_POINTER_1, ND_POINTER_2, ctypes.c_size_t, ctypes.c_size_t]\n# print(\"-->\", ctypes.c_size_t)\n\ndef mult_matrix_cpp(a,b):\n shape = a.shape[0] * a.shape[1]\n libmatmult.mult_matrix.restype = None\n libmatmult.mult_matrix(a, b, *a.shape, *b.shape , a.shape[0] * a.shape[1])\n\nsize_a = (300,300)\nsize_b = size_a\n\na = numpy.random.uniform(low=1, high=255, size=size_a)\nb = numpy.random.uniform(low=1, high=255, size=size_b)\n\nt2 = time()\nout_cpp = mult_matrix_cpp(a,b)\nprint(\"cpp time taken:{:.2f} ms\".format((time() - t2) * 1000))\nout_cpp = numpy.array(out_cpp).reshape(size_a[0], size_a[1])\n\n" ]

[ 16, 0 ]

[]

[]

[ "ctypes", "multidimensional_array", "python", "python_3.x", "structure" ]

[ "It's my answer:\nWin32API Mouse vs Real Mouse Click\nBut I don't know how to write a custom driver :)\n" ]

[ 0 ]

[]

[]

[ "events", "keyboard", "python" ]

[ "Let's create some data:\nimport matplotlib.pyplot as plt\nimport matplotlib.dates as mdates\nimport numpy as np\n\ndays = np.array([\"2022-01-01\", \"2022-07-01\", \"2023-02-15\", \"2023-11-15\", \"2024-03-03\"],\n dtype = \"datetime64\")\nval = np.array([20, 20, -10, -10, 10])\n\nFor the date in the x-axis, we import matplotlib.dates, which provides the month locator and the date formater. The locator sets the ticks each 3 months, and the formater sets the way the labels are displayed (month-00).\nFor the y-axis data, you require changing the sign of the data (hence the negative sign in ax2.plot(), but you want the curve in the same position, so afterwards you need to invert the axis. And so, the curves in both plots are identical, but the y-axis values have different signs and directions.\nfig, (ax1, ax2) = plt.subplots(figsize = (10,5), nrows = 2)\n\nax1.plot(days, val, marker = \"x\")\n# set the locator to Jan, Apr, Jul, Oct\nax1.xaxis.set_major_locator(mdates.MonthLocator( bymonth = (1, 4, 7, 10) ))\n# set the formater for month-year, with lower y to show only two digits\nax1.xaxis.set_major_formatter(mdates.DateFormatter(\"%b-%y\"))\n\n# change the sign of the y data plotted\nax2.plot(days, -val, marker = \"x\")\n#invert the y axis\nax2.invert_yaxis()\n# set the locator to Jan, Apr, Jul, Oct\nax2.xaxis.set_major_locator(mdates.MonthLocator( bymonth = (1, 4, 7, 10) ))\n# set the formater for month-year, with lower y to show only two digits\nax2.xaxis.set_major_formatter(mdates.DateFormatter(\"%b-%y\"))\n\nplt.show()\n\n\n" ]

[ 0 ]

[]

[]

[ "matplotlib", "pandas", "python", "yticks" ]

[ "Everything just worked once I switched to MySQL. It was not working on Sqlite. I don't know why it doesn't work with sqlite maybe it is a bug. But changing the database solved my problem.\n" ]

[ 0 ]

[]

[]

[ "django", "django_4.1", "django_modeladmin", "django_models", "python" ]

[ "Use return instead of sys.exit\nCopied from bigbounty's comment\n", "If you are calling a function from another function and needs to directly exit from the inner function without return you can use abort.\nabort will directly return the response and exit the program from the place you are calling it.\nimport sys \nfrom flask import abort\n\nif str(strEnabled) == 'True':\n printOperation = \"Operation: Enabling of user\" \nelse:\n # abort(200) #return 200\n abort(500, description=\"Exit\") #return 500\n\n\nNote: cloud functions use flask module internally, so you need not\ninstall it separately.\n\n" ]

[ 3, 0 ]

[]

[]

[ "google_cloud_functions", "google_cloud_platform", "python" ]

[ "Use numpy.argsort for positions of second largest values:\ndf['new'] = df['new'] = df.columns.to_numpy()[np.argsort(df.to_numpy())[:, -2]]\nprint(df)\n a b c new\n0 4 20 25 b\n1 5 10 20 b\n2 3 40 5 c\n3 1 50 15 c\n4 2 30 10 c\n\nYour solution should working, but is slow:\ndef second_largest(l = []): \n return (l.nlargest(2).idxmin())\n\nprint(df.apply(second_largest, axis = 1))\n\n", "If efficiency is important, numpy.argpartition is quite efficient:\nN = 2\ncols = df.columns.to_numpy()\npd.Series(cols[np.argpartition(df.to_numpy().T, -N, axis=0)[-N]], index=df.index)\n\nIf you want a pure pandas (less efficient):\nout = df.stack().groupby(level=0).apply(lambda s: s.nlargest(2).index[-1][1])\n\nOutput:\n0 b\n1 b\n2 c\n3 c\n4 c\ndtype: object\n\n" ]

[ 4, 2 ]

[]

[]

[ "dataframe", "numpy", "pandas", "python" ]

[ "There are a few issues in your attempt. Starting from the top:\n\nAssuming that ans[i] represents the maximum number of partitions of the substring s[0:i] into valid words, you'll need to make this list one entry longer, so that there is an ans[n], which will eventually contain the answer, i.e. the maximum number of partitions in s[0:n] which is s. So change:\nans = [0]*n\n\nto\nans = [0]*(n+1)\n\n\nBecause of the reasoning given in the first bullet point, the final return should be return ans[n] instead of return ans[n-2].\n\nAgain because of the reasoning given in the first billet point, the outer loop should make one more iteration, so change:\nfor i in range(1, len(s)):\n\nto\nfor i in range(1, len(s)+1):\n\n\nAssuming that j represents the size of the substring to test, the inner loop should allow i-j to eventually reach back to 0, so j should be able to reach the value i. That means the inner loop should make two more iterations. So change:\nfor j in range(1, i-1):\n\nto\nfor j in range(1, i+1):\n\n\nIn wordcheck, the slice from s should start at i-j, as j represents the size of the slice. So change:\nstart = i-j+1\n\nto\nstart = i-j\n\n\nIn wordcheck, the loop will always exit in its first iteration. It cannot loop more as it always executes a return. You should not exit the loop when there is no success. The return float('-inf') should only be made when all words have been tried. So remove:\n else:\n return float('-inf')\n\nand instead add at after the loop:\nreturn float('-inf')\n\n\n\nThose are the issues. The code with all those corrections:\ndef word_split_dp(s):\n n = len(s)\n ans = [0]*(n+1) #correction\n # base case\n ans[0] = 0\n for i in range(1, len(s) + 1): # correction\n ans[i] = float('-inf')\n for j in range(1, i + 1): # correction\n ans[i] = max(ans[i], ans[i-j] + wordcheck(s,i,j))\n \n print(ans)\n return ans[n] # correction\n \ndef wordcheck(s,i,j):\n words=(\"war\",\"month\",\"on\",\"the\",\"heat\",\"eat\",\"he\",\"arm\",\"at\")\n for word in words:\n start = i-j # correction\n if s[start:i] == word:\n return 1\n # don't interrupt loop otherwise\n return float('-inf')\n \ns=\"warmontheheat\"\nprint(word_split_dp(s)) # -inf\ns=\"warontheheat\"\nprint(word_split_dp(s)) # 5\n\n" ]

[ 0 ]

[]

[]

[ "algorithm", "python", "word_break" ]

[ "I have made small test on your code, changing the last line of your onpick function to figR.canvas.draw() solve the issue for me, the function should look like:\ndef onpick(event):\n if event.artist != line:\n return\n n = len(event.ind)\n if not n:\n return\n print(\"Index \",event.ind)\n \n axsR.plot(X[event.ind])\n axsR.set_ylim(-0.5,1.5)\n \n figR.canvas.draw()\n return True\n\ndraw() will re-draw the figure. This allows you to work in interactive mode an in case you have changed your data or formatted the figure, allow the graph itself to change.\n" ]

[ 1 ]

[]

[]

[ "events", "matplotlib", "python" ]

[ "It is partially unclear what you are trying to do since you didn't provide a minimal reproducible example. I've filled in as best I can and I think this should still help you solve your problem. You may need to adjust the type of aggregation to fit your needs.\nYour for loop is of course executing as many times as it says it is.\nI believe you forgot to set the value of boot_rois to boot, otherwise your code just regenerates the ROI list.\n boot_rois[roi] = boot\n\nA secondary problem with your code is that everytime you do your inner for loop, you are just overwriting the same key in your dictionary. You probably want to do something like this instead. It isn't completely clear what type of math you are trying to do, but assuming you want to calculate the average between your 599 random arrays in a rolling fashion you could do this:\nimport numpy as np\nN=10\ndata_rois={\"interoception\" : [1,2], \"extero\" : [2,3] }\nrois = data_rois.keys() # A long list of rois\nboot_rois = {}\nfor roi in rois:\n last_boot = None\n for i in range(7):\n boot = np.random.choice(data_rois[roi], size=N)\n print(boot)\n boot = np.mean(boot)\n # During first iteration aggregator last_boot is None\n if(last_boot is not None):\n # Average with the last iteration and repeat\n # This logic may need to be replaced with whatever math you are trying to do\n boot = np.mean([boot,last_boot])\n \n boot_rois[roi] = boot\n last_boot = boot_rois[roi]\n \nprint(boot_rois)\n\nNote: Doing it this way means that the average is not the average of each of the 7, if you want to do that you can store them in a sum variable and divide by the number of iterations you performed in the inner for loop. Mathematically, doing the mean multiple times is different than summing everything and dividing by the number of sums. Make sure your math is correct.\n", "What about using comprehension list instead of nested loops:\nrois = [\"interoception\", \"extero\", ...] # A long list of rois\nboot_rois = {}\nfor roi in rois:\n # will execute np.random.choice 599 times and store these results in a list\n rand_choices = [np.mean(np.random.choice(data_rois[roi], size=N)) for _ in range(599)]\n # will calculate the standard deviation of those 599 results\n boot_rois[roi] = np.std(rand_choices)\n\nThis will run np.random.choice 599 times and store these results in a list (I added np.mean(...) so you can calculate the stdev on those 599 mean values), so you can run np.std on that list, and store this final result in boot_rois[roi]\n\nBelow is runnable code I used for tests purpose. It generates 20 random numbers between 0 and 50, and calculates the stdev:\nimport random\nimport numpy as np\n\nrand_ints = [random.randint(0, 50) for _ in range(20)]\nprint(rand_ints)\nstdev = np.std(rand_ints)\nprint(stdev)\n\nFirst execution:\n\n[9, 44, 13, 0, 43, 12, 4, 40, 35, 38, 43, 0, 3, 38, 39, 45, 37, 14, 4, 21]\n16.908281994336384\n\n\nSecond execution:\n\n[2, 20, 17, 32, 0, 39, 23, 27, 24, 41, 8, 21, 2, 7, 21, 3, 27, 7, 15, 36]\n12.531560158256433\n\n\n\nIn order to emulate calculating the stdev of means of samples, I tried this:\nimport random\nimport numpy as np\n\nrand_ints = [np.mean([random.randint(0, 50) for _ in range(10)]) for _ in range(20)]\nprint(rand_ints)\nstdev = np.std(rand_ints)\nprint(stdev)\n\nWhich gave me:\n\n[25.8, 16.9, 27.6, 21.8, 20.6, 30.5, 19.4, 32.9, 27.8, 18.5, 24.5, 18.7, 23.1, 26.9, 30.6, 25.1, 24.9, 26.5, 21.8, 25.8]\n4.2607833786758045\n\n\n" ]

[ 2, 1 ]

[]

[]

[ "dictionary", "for_loop", "python" ]

[ "In Azure Functions, HTTP Error 503 Service Unavailable can be caused due to few reasons like:\n\nThe backend server returned 503 due to a memory leak/issue in the code.\nPlatform issue due to the backend server not running/ allocated\nFunction host is down/restarting.\n\n\nHave a look into the \"Diagnose and Solve problems\" blade in the Azure Function App and select the \"Function app down or reporting\" detector.\n\nAlso, Microsoft Azure Support will do the analysis to find the root cause on your function app if we send an email at [email protected] with your subscription ID as well as your function app name using function logs.\n", "There might be a few issues causing this so a solid 1 truth solutions seems not be fully possible. But I will list a few possible solutions.\n\nIt seems to mostly be an issue when running code in the portal with Code+test. Try run the function without using the portal. (example on others issue Azure Function dies after 30 seconds and cannot run multiple invocations concurrently )\n\nSome report there might be an issue with region. (Not working in one region but after move to other region it works)\n\nAlso been reports of solving this with premium plan (But looks like you tried this)\n\nCould be some other resources like a loadbalancer with lower timeouts. (Example Why does my POST request throw a 503 after 30 seconds? ). Don't know your full setup. Might be when calling the function or a request the function does.\n\nIf none of the above helps it might be some python timeout like python async timeout. I am no python expert though so not fully sure here.\n\n\n" ]

[ 0, 0 ]

[]

[]

[ "azure", "azure_functions", "python", "serverless" ]

[ "You could use os.listdir() to list all files of a given folder.\nThen you could either delete all files or delete files based on their file name or file type/suffix.\nE. g.:\nimport os\nfrom pathlib import Path\n\nfiles = os.listdir(\"your/path\")\nfor file in files:\n os.remove(file) # delete all files\n\n if Path(file).name == \"Test\" or Path(file).suffix == \".txt\":\n os.remove(file) # delete file based on their name or suffix\n\n" ]

[ 1 ]

[]

[]

[ "delete_file", "python" ]

[ "After reproducing from my end, this was working fine. I believe that the there is no value that is getting returned from process_data. Make sure there is no null value that is getting returned from process_data. Below is a sample code that I used in process_data which gave me expected results.\ndef process_data(df_base,df_new):\n final_dataframe = pd.concat([df_base, df_new])\n return final_dataframe\n\ninit.py\nimport logging\nimport pandas as pd\n\nimport azure.functions as func\nfrom io import BytesIO\n\ndef main(myblob: func.InputStream, base: func.InputStream, blobout: func.Out[bytes]):\n df_base = pd.read_csv(BytesIO(base.read()))\n df_new = pd.read_csv(BytesIO(myblob.read()))\n df_final = process_data(df_base, df_new)\n df_final = df_final.to_csv(index=False)\n blobout.set(df_final.encode())\n \ndef process_data(df_base,df_new):\n final_dataframe = pd.concat([df_base, df_new])\n return final_dataframe\n\nfunction.json\n{\n \"scriptFile\": \"__init__.py\",\n \"bindings\": [\n {\n \"name\": \"myblob\",\n \"type\": \"blobTrigger\",\n \"direction\": \"in\",\n \"path\": \"new-container/{name}\",\n \"connection\": \"<STORAGE_CONNECTION>\"\n },\n {\n \"name\": \"base\",\n \"type\": \"blob\",\n \"path\": \"base-container/base.csv\",\n \"connection\": \"<STORAGE_CONNECTION>\",\n \"direction\": \"in\"\n },\n {\n \"name\": \"blobout\",\n \"type\": \"blob\",\n \"path\": \"base-container/base.csv\",\n \"connection\": \"<STORAGE_CONNECTION>\",\n \"direction\": \"out\"\n }\n ]\n}\n\nRESULT\n\n" ]

[ 0 ]

[]

[]

[ "azure", "azure_blob_storage", "azure_functions", "csv", "python" ]

[ "As @ddejohn correctly said it in the comments, the behavior you face is actually coming from Excel, automatically converting the data to date. Thus pandas will have to deal with that data AS date, and treat it later to get the correct format back as you expect, as like you say you cannot modify the input Excel file.\nHere is a short script to make it work as you expect:\nimport pandas as pd\n\ndef rev(x: str) -> str:\n '''\n converts '2003-02-14 00:00:00' to '14.02.03'\n '''\n\n hours = '00:00:00'\n if not hours in x:\n return x\n y = x.split()[0]\n y = y.split('-')\n return '.'.join([i[-2:] for i in y[::-1]])\n\nfile = r'C:\\your\\folder\\path\\Classeur1.xlsx'\ndf = pd.read_excel(file, dtype=str)\n\ndf['column'] = df['column'].apply(rev)\n\n\nReplace df['column'] by your actual column name.\nYou then get the desired format in your dataframe.\n" ]

[ 0 ]

[]

[]

[ "excel", "pandas", "python" ]

[ "Firat aggreagte sum, then sorting values and convert to DataFrame:\ndf1 = (df.groupby('country')['days']\n .sum()\n .sort_values(ascending=False)\n .reset_index(name='count_days'))\nprint (df1)\n country count_days\n0 Spain 8\n1 Italy 4\n2 UK 3\n3 France 2\n4 Germany 1\n5 Portugal 1\n\nLast convert column to list:\ncountries = df1['country'].tolist()\n\nSolution without DataFrame df1:\ncountries = df.groupby('country')['days'].sum().sort_values(ascending=False).index.tolist()\n\n" ]

[ 2 ]

[]

[]

[ "dataframe", "pandas", "python" ]

[ "The page loads data dynamically using an API, so you don't see that data when u try to get it. U need to figure out which query provides the information and what data is required to get it. I made a small example where we get the necessary tokens and information via api. Next change page varible to get new ids.\nimport requests\nimport pandas as pd\nimport json\n\n\ndef get_token():\n url = 'https://www.kaggle.com/code?sortBy=voteCount&page=1'\n response = requests.get(url)\n return response.cookies.get_dict()['CSRF-TOKEN'], response.cookies.get_dict()['XSRF-TOKEN']\n\n\ndef get_kernel_ids(csrf_token: str, xsrf_token: str, page: int):\n url = \"https://www.kaggle.com/api/i/kernels.KernelsService/ListKernelIds\"\n payload = json.dumps({\n \"sortBy\": \"VOTE_COUNT\",\n \"pageSize\": 100,\n \"group\": \"EVERYONE\",\n \"page\": page,\n \"tagIds\": \"\",\n \"excludeResultsFilesOutputs\": False,\n \"wantOutputFiles\": False,\n \"excludeKernelIds\": []\n })\n headers = {\n 'accept': 'application/json',\n 'content-type': 'application/json',\n 'cookie': f'CSRF-TOKEN={csrf_token}',\n 'x-xsrf-token': xsrf_token\n }\n response = requests.post(url, headers=headers, data=payload)\n return response.json()['kernelIds']\n\n\ndef get_info(csrf_token: str, xsrf_token: str):\n url = \"https://www.kaggle.com/api/i/kernels.KernelsService/GetKernelListDetails\"\n payload = json.dumps({\n \"deletedAccessBehavior\": \"RETURN_NOTHING\",\n \"unauthorizedAccessBehavior\": \"RETURN_NOTHING\",\n \"excludeResultsFilesOutputs\": False,\n \"wantOutputFiles\": False,\n \"kernelIds\": get_kernel_ids(csrf_token, xsrf_token, 1),\n \"outputFileTypes\": [],\n \"includeInvalidDataSources\": False\n })\n headers = {\n 'accept': 'application/json',\n 'content-type': 'application/json',\n 'cookie': f'CSRF-TOKEN={csrf_token}',\n 'x-xsrf-token': xsrf_token\n }\n response = requests.post(url, headers=headers, data=payload)\n data = []\n for kernel in response.json()['kernels']:\n data.append({\n 'title': kernel['title'],\n 'stars': kernel['totalVotes'],\n 'resume': kernel['dataSources'][0]['name'] if 'dataSources' in kernel else 'No attached data sources',\n 'comments': kernel['totalComments'],\n 'url': f'https://www.kaggle.com{kernel[\"scriptUrl\"]}'\n })\n return data\n\n\ncsrf, xsrf = get_token()\ndf = pd.DataFrame(get_info(csrf, xsrf))\nprint(df.to_string())\n\nOUTPUT:\n title stars resume comments url\n0 Comprehensive data exploration with Python 12772 House Prices - Advanced Regression Techniques 1819 https://www.kaggle.com/code/pmarcelino/comprehensive-data-exploration-with-python\n1 Titanic Data Science Solutions 9374 Titanic - Machine Learning from Disaster 2316 https://www.kaggle.com/code/startupsci/titanic-data-science-solutions\n2 Titanic Tutorial 8161 Titanic - Machine Learning from Disaster 26348 https://www.kaggle.com/code/alexisbcook/titanic-tutorial\n3 Stacked Regressions : Top 4% on LeaderBoard 6749 House Prices - Advanced Regression Techniques 1075 https://www.kaggle.com/code/serigne/stacked-regressions-top-4-on-leaderboard\n4 Introduction to CNN Keras - 0.997 (top 6%) 6438 Digit Recognizer 1002 https://www.kaggle.com/code/yassineghouzam/introduction-to-cnn-keras-0-997-top-6\n5 Data ScienceTutorial for Beginners 6213 Pokemon- Weedle's Cave 1160 https://www.kaggle.com/code/kanncaa1/data-sciencetutorial-for-beginners\n6 Hello, Python 5952 No attached data sources 329 https://www.kaggle.com/code/colinmorris/hello-python\n7 Introduction to Ensembling/Stacking in Python 5657 Titanic - Machine Learning from Disaster 1031 https://www.kaggle.com/code/arthurtok/introduction-to-ensembling-stacking-in-python\n8 How Models Work 5308 Mobile Price Classification 2 https://www.kaggle.com/code/dansbecker/how-models-work\n9 A Data Science Framework: To Achieve 99% Accuracy 5266 Titanic - Machine Learning from Disaster 657 https://www.kaggle.com/code/ldfreeman3/a-data-science-framework-to-achieve-99-accuracy\n10 Credit Fraud || Dealing with Imbalanced Datasets 4274 Credit Card Fraud Detection 629 https://www.kaggle.com/code/janiobachmann/credit-fraud-dealing-with-imbalanced-datasets\n11 Exploring Survival on the Titanic 3735 Titanic - Machine Learning from Disaster 1041 https://www.kaggle.com/code/mrisdal/exploring-survival-on-the-titanic\n12 Start Here: A Gentle Introduction 3372 Home Credit Default Risk 543 https://www.kaggle.com/code/willkoehrsen/start-here-a-gentle-introduction\n13 Functions and Getting Help 2974 No attached data sources 145 https://www.kaggle.com/code/colinmorris/functions-and-getting-help\n14 Your First Machine Learning Model 2913 Mobile Price Classification 386 https://www.kaggle.com/code/dansbecker/your-first-machine-learning-model\n15 Exercise: Your First Machine Learning Model 2840 Melbourne Housing Snapshot 381 https://www.kaggle.com/code/yogeshtak/exercise-your-first-machine-learning-model\n16 Titanic Top 4% with ensemble modeling 2724 Titanic - Machine Learning from Disaster 408 https://www.kaggle.com/code/yassineghouzam/titanic-top-4-with-ensemble-modeling\n17 Model Validation 2655 Mobile Price Classification 5 https://www.kaggle.com/code/dansbecker/model-validation\n18 EDA To Prediction(DieTanic) 2544 Titanic - Machine Learning from Disaster 342 https://www.kaggle.com/code/ash316/eda-to-prediction-dietanic\n19 Winning solutions of kaggle competitions 2527 [Private Datasource] 207 https://www.kaggle.com/code/sudalairajkumar/winning-solutions-of-kaggle-competitions\n20 Booleans and Conditionals 2501 No attached data sources 75 https://www.kaggle.com/code/colinmorris/booleans-and-conditionals\n21 Underfitting and Overfitting 2500 Mobile Price Classification 7 https://www.kaggle.com/code/dansbecker/underfitting-and-overfitting\n22 Getting Started with Kaggle 2420 No attached data sources 4265 https://www.kaggle.com/code/alexisbcook/getting-started-with-kaggle\n23 Full Preprocessing Tutorial 2366 Data Science Bowl 2017 490 https://www.kaggle.com/code/gzuidhof/full-preprocessing-tutorial\n24 Machine Learning Tutorial for Beginners 2333 Biomechanical features of orthopedic patients 292 https://www.kaggle.com/code/kanncaa1/machine-learning-tutorial-for-beginners\n25 Everything you can do with a time series 2330 DJIA 30 Stock Time Series 171 https://www.kaggle.com/code/thebrownviking20/everything-you-can-do-with-a-time-series\n26 Deep Learning Tutorial for Beginners 2296 Sign Language Digits Dataset 248 https://www.kaggle.com/code/kanncaa1/deep-learning-tutorial-for-beginners\n27 Getting staRted in R: First Steps 2216 Chocolate Bar Ratings 102 https://www.kaggle.com/code/rtatman/getting-started-in-r-first-steps\n28 Approaching (Almost) Any NLP Problem on Kaggle 2169 glove.840B.300d.txt 231 https://www.kaggle.com/code/abhishek/approaching-almost-any-nlp-problem-on-kaggle\n29 Data Science Glossary on Kaggle 2125 [Private Datasource] 229 https://www.kaggle.com/code/shivamb/data-science-glossary-on-kaggle\n30 Coronavirus (COVID-19) Visualization & Prediction 2069 No attached data sources 691 https://www.kaggle.com/code/therealcyberlord/coronavirus-covid-19-visualization-prediction\n31 Creating, Reading and Writing 2063 18,393 Pitchfork Reviews 95 https://www.kaggle.com/code/residentmario/creating-reading-and-writing\n32 Dive into dplyr (tutorial #1) 1828 Palmer Archipelago (Antarctica) penguin data 134 https://www.kaggle.com/code/jessemostipak/dive-into-dplyr-tutorial-1\n33 Back to (predict) the future - Interactive M5 EDA 1820 US Natural Disaster Declarations 322 https://www.kaggle.com/code/headsortails/back-to-predict-the-future-interactive-m5-eda\n34 COVID-19 - Analysis, Visualization & Comparisons 1773 World Happiness Report 359 https://www.kaggle.com/code/imdevskp/covid-19-analysis-visualization-comparisons\n35 Time series Basics : Exploring traditional TS 1767 Predict Future Sales 174 https://www.kaggle.com/code/jagangupta/time-series-basics-exploring-traditional-ts\n36 Titanic - Advanced Feature Engineering Tutorial 1753 Titanic - Machine Learning from Disaster 322 https://www.kaggle.com/code/gunesevitan/titanic-advanced-feature-engineering-tutorial\n37 Regularized Linear Models 1721 House Prices - Advanced Regression Techniques 336 https://www.kaggle.com/code/apapiu/regularized-linear-models\n38 Random Forests 1711 Mobile Price Classification 2 https://www.kaggle.com/code/dansbecker/random-forests\n39 Deep Learning For NLP: Zero To Transformers & BERT 1682 glove.840B.300d.txt 89 https://www.kaggle.com/code/tanulsingh077/deep-learning-for-nlp-zero-to-transformers-bert\n40 Feature Selection and Data Visualization 1681 Breast Cancer Wisconsin (Diagnostic) Data Set 304 https://www.kaggle.com/code/kanncaa1/feature-selection-and-data-visualization\n41 Is it a bird? Creating a model from your own data 1670 No attached data sources 32 https://www.kaggle.com/code/jhoward/is-it-a-bird-creating-a-model-from-your-own-data\n42 Lists 1651 No attached data sources 60 https://www.kaggle.com/code/colinmorris/lists\n43 Submitting From A Kernel 1650 House Prices - Advanced Regression Techniques 491 https://www.kaggle.com/code/dansbecker/submitting-from-a-kernel\n44 Basic Data Exploration 1552 Mobile Price Classification 8 https://www.kaggle.com/code/dansbecker/basic-data-exploration\n45 Indexing, Selecting & Assigning 1494 18,393 Pitchfork Reviews 121 https://www.kaggle.com/code/residentmario/indexing-selecting-assigning\n46 Getting Started with a Movie Recommendation System 1482 TMDB 5000 Movie Dataset 181 https://www.kaggle.com/code/ibtesama/getting-started-with-a-movie-recommendation-system\n47 House prices: Lasso, XGBoost, and a detailed EDA 1470 House Prices - Advanced Regression Techniques 255 https://www.kaggle.com/code/erikbruin/house-prices-lasso-xgboost-and-a-detailed-eda\n48 Twitter sentiment Extaction-Analysis,EDA and Model 1440 [Private Datasource] 122 https://www.kaggle.com/code/tanulsingh077/twitter-sentiment-extaction-analysis-eda-and-model\n49 Loops and List Comprehensions 1436 No attached data sources 75 https://www.kaggle.com/code/colinmorris/loops-and-list-comprehensions\n50 Exploratory Analysis - Zillow 1428 Zillow Prize: Zillow’s Home Value Prediction (Zestimate) 169 https://www.kaggle.com/code/philippsp/exploratory-analysis-zillow\n51 Data Analysis & XGBoost Starter (0.35460 LB) 1414 Quora Question Pairs 168 https://www.kaggle.com/code/anokas/data-analysis-xgboost-starter-0-35460-lb\n52 A Statistical Analysis & ML workflow of Titanic 1411 Titanic - Machine Learning from Disaster 321 https://www.kaggle.com/code/masumrumi/a-statistical-analysis-ml-workflow-of-titanic\n53 A Journey through Titanic 1399 Titanic - Machine Learning from Disaster 417 https://www.kaggle.com/code/omarelgabry/a-journey-through-titanic\n54 Plotly Tutorial for Beginners 1370 World University Rankings 143 https://www.kaggle.com/code/kanncaa1/plotly-tutorial-for-beginners\n55 Tutorial on reading large datasets 1365 Riiid train data (multiple formats) 111 https://www.kaggle.com/code/rohanrao/tutorial-on-reading-large-datasets\n56 Python Data Visualizations 1365 Iris Species 162 https://www.kaggle.com/code/benhamner/python-data-visualizations\n57 Working with External Libraries 1353 No attached data sources 53 https://www.kaggle.com/code/colinmorris/working-with-external-libraries\n58 Strings and Dictionaries 1346 No attached data sources 52 https://www.kaggle.com/code/colinmorris/strings-and-dictionaries\n59 Explore Your Data 1329 home data for ml course 237 https://www.kaggle.com/code/dansbecker/explore-your-data\n60 Handling Missing Values 1324 Melbourne Housing Market 440 https://www.kaggle.com/code/dansbecker/handling-missing-values\n61 Basic EDA,Cleaning and GloVe 1311 GloVe: Global Vectors for Word Representation 162 https://www.kaggle.com/code/shahules/basic-eda-cleaning-and-glove\n62 Pytorch Tutorial for Deep Learning Lovers 1310 Digit Recognizer 122 https://www.kaggle.com/code/kanncaa1/pytorch-tutorial-for-deep-learning-lovers\n63 EDA is Fun! 1298 PUBG Finish Placement Prediction (Kernels Only) 186 https://www.kaggle.com/code/deffro/eda-is-fun\n64 NLP with Disaster Tweets - EDA, Cleaning and BERT 1278 Pickled glove.840B.300d 209 https://www.kaggle.com/code/gunesevitan/nlp-with-disaster-tweets-eda-cleaning-and-bert\n65 Data Cleaning Challenge: Handling missing values 1260 Detailed NFL Play-by-Play Data 2009-2018 377 https://www.kaggle.com/code/rtatman/data-cleaning-challenge-handling-missing-values\n66 Titanic Survival Predictions (Beginner) 1252 Titanic - Machine Learning from Disaster 272 https://www.kaggle.com/code/nadintamer/titanic-survival-predictions-beginner\n67 Santander EDA and Prediction 1215 Santander Customer Transaction Prediction 189 https://www.kaggle.com/code/gpreda/santander-eda-and-prediction\n68 Titanic best working Classifier 1181 Titanic - Machine Learning from Disaster 195 https://www.kaggle.com/code/sinakhorami/titanic-best-working-classifier\n69 Data Science for tabular data: Advanced Techniques 1143 No Data Sources 101 https://www.kaggle.com/code/vbmokin/data-science-for-tabular-data-advanced-techniques\n70 Keras U-Net starter - LB 0.277 1120 2018 Data Science Bowl 163 https://www.kaggle.com/code/keegil/keras-u-net-starter-lb-0-277\n71 Simple Exploration Notebook - Zillow Prize 1104 Zillow Prize: Zillow’s Home Value Prediction (Zestimate) 131 https://www.kaggle.com/code/sudalairajkumar/simple-exploration-notebook-zillow-prize\n72 EDA and models 1096 IEEE-CIS Fraud Detection 218 https://www.kaggle.com/code/artgor/eda-and-models\n73 How to: Preprocessing when using embeddings 1094 Quora Insincere Questions Classification 102 https://www.kaggle.com/code/christofhenkel/how-to-preprocessing-when-using-embeddings\n74 COVID-19 Literature Clustering 1087 COVID-19 Open Research Dataset Challenge (CORD-19) 232 https://www.kaggle.com/code/maksimeren/covid-19-literature-clustering\n75 Head Start for Data Scientist 1086 Titanic - Machine Learning from Disaster 233 https://www.kaggle.com/code/hiteshp/head-start-for-data-scientist\n76 Be my guest - Recruit Restaurant EDA 1079 Weather Data for Recruit Restaurant Competition 237 https://www.kaggle.com/code/headsortails/be-my-guest-recruit-restaurant-eda\n77 NB-SVM strong linear baseline 1070 Toxic Comment Classification Challenge 152 https://www.kaggle.com/code/jhoward/nb-svm-strong-linear-baseline\n78 Seaborn Tutorial for Beginners 1070 Fatal Police Shootings in the US 182 https://www.kaggle.com/code/kanncaa1/seaborn-tutorial-for-beginners\n79 A look at different embeddings.! 1057 Quora Insincere Questions Classification 111 https://www.kaggle.com/code/sudalairajkumar/a-look-at-different-embeddings\n80 Deep Neural Network Keras way 1053 Digit Recognizer 201 https://www.kaggle.com/code/poonaml/deep-neural-network-keras-way\n81 Simple Exploration+Baseline - GA Customer Revenue 1044 Google Analytics Customer Revenue Prediction 162 https://www.kaggle.com/code/sudalairajkumar/simple-exploration-baseline-ga-customer-revenue\n82 Simple Matplotlib & Visualization Tips 1040 Students Performance in Exams 148 https://www.kaggle.com/code/subinium/simple-matplotlib-visualization-tips\n83 Speech representation and data exploration 1031 TensorFlow Speech Recognition Challenge 110 https://www.kaggle.com/code/davids1992/speech-representation-and-data-exploration\n84 Feature engineering, xgboost 1023 Predict Future Sales 240 https://www.kaggle.com/code/dlarionov/feature-engineering-xgboost\n85 XGBoost 1018 House Prices - Advanced Regression Techniques 12 https://www.kaggle.com/code/dansbecker/xgboost\n86 Stock Market Analysis + Prediction using LSTM 1015 Tesla Stock Price 190 https://www.kaggle.com/code/faressayah/stock-market-analysis-prediction-using-lstm\n87 Two Sigma News Official Getting Started Kernel 994 Two Sigma: Using News to Predict Stock Movements 142 https://www.kaggle.com/code/dster/two-sigma-news-official-getting-started-kernel\n88 EDA, feature engineering and everything 984 Two Sigma: Using News to Predict Stock Movements 171 https://www.kaggle.com/code/artgor/eda-feature-engineering-and-everything\n89 A study on Regression applied to the Ames dataset 979 House Prices - Advanced Regression Techniques 139 https://www.kaggle.com/code/juliencs/a-study-on-regression-applied-to-the-ames-dataset\n90 Customer Segmentation 976 E-Commerce Data 72 https://www.kaggle.com/code/fabiendaniel/customer-segmentation\n91 How I made top 0.3% on a Kaggle competition 975 Kernel Files 124 https://www.kaggle.com/code/lavanyashukla01/how-i-made-top-0-3-on-a-kaggle-competition\n92 Extensive EDA and Modeling XGB Hyperopt 972 IEEE-CIS Fraud Detection 187 https://www.kaggle.com/code/kabure/extensive-eda-and-modeling-xgb-hyperopt\n93 Summary Functions and Maps 965 18,393 Pitchfork Reviews 132 https://www.kaggle.com/code/residentmario/summary-functions-and-maps\n94 NovelAI with webUI(Stable Diffusion version) 955 NovelAi-model-pruned 7 https://www.kaggle.com/code/inmine/novelai-with-webui-stable-diffusion-version\n95 Customer Segmentation: Clustering ️ 934 Customer Personality Analysis 256 https://www.kaggle.com/code/karnikakapoor/customer-segmentation-clustering\n96 Introduction to financial concepts and data 934 Optiver Realized Volatility Prediction 130 https://www.kaggle.com/code/jiashenliu/introduction-to-financial-concepts-and-data\n97 Resampling strategies for imbalanced datasets 934 Porto Seguro’s Safe Driver Prediction 80 https://www.kaggle.com/code/rafjaa/resampling-strategies-for-imbalanced-datasets\n98 Jane Street: EDA of day 0 and feature importance 932 Jane Street Market Prediction 151 https://www.kaggle.com/code/carlmcbrideellis/jane-street-eda-of-day-0-and-feature-importance\n99 10 Simple hacks to speed up your Data Analysis 929 Titanic - Machine Learning from Disaster 204 https://www.kaggle.com/code/parulpandey/10-simple-hacks-to-speed-up-your-data-analysis\n\n" ]

[ 3 ]

[]

[]

[ "beautifulsoup", "python", "python_3.x", "web_scraping" ]

[ "You need to actually start consuming messages. As it is right now you have a function called callback that never gets called (and is never referenced in your code).\nchannel.basic_consume(queue='main', on_message_callback=callback)\nchannel.start_consuming()\n\nAdd these lines at the end of your code. This will assign callback as the on message function and start_consuming will tell pika to start consuming the queue.\nMore detailed examples available here.\n", "Same problem. Working container with no output. I added PYTHONUNBUFFERED=1 for enable log.\n# Dockerfile\nENV PYTHONUNBUFFERED=1 \n\n\n" ]

[ 0, 0 ]

[]

[]

[ "cloudamqp", "docker", "docker_compose", "pika", "python" ]

[ "You can create a function that accepts any amount of arguments using the unpack operater *.\nHere is an example of a function that takes any amount of arguments:\n\ndef my_awesome_function(*args):\n # do awesome job!\n print(args)\n\n # depending on len of args do something...\n\n\nOr if you always have a one argument, but an unknown number of following arguments, you can formulate it like this:\n\ndef my_awesome_function(num, *all_other_args):\n # my number\n print(num)\n\n # do awesome job!\n print(all_other_args)\n\n # depending on len of args do something...\n\n\nHere is full explanation of how to deal with unknown number of args, and more usefull operators: https://www.scaler.com/topics/python/packing-and-unpacking-in-python/\nIf you are looking to do it the other way around, such that a function deeds all the items in a list as separate arguments. This could be done like this with the * operator as well.\ndef foo(a, b, c):\n print(a, b, c)\n\nvalues = ['adam', 'dave', 'elon']\n\n# a valid funciton call would be\nfoo(*values)\n\n# values -> ['adam', 'dave', 'elon']\n# *values -> 'adam', 'dave', 'elon'\n\nHopefully this helps. Please leave a comment if I misunderstood.\nGood Luck!\n" ]

[ 2 ]

[]

[]

[ "python" ]

[ "Please try the same code with steps in below\n\nFirst :\n\nall_descriptors.append(descriptors)\n\n\nthen :\n\na=np.array(a).astype('float32')\n\n" ]

[ 0 ]

[]

[]

[ "feature_extraction", "opencv", "python", "sift" ]

[ "Since re does not support unknown length lookbehind patterns, the plain regex solution is not possible. However, the PyPi regex library supports such lookbehind patterns.\nAfter installing PyPi regex, you can use\n(?<!fake.*)(man[a-z]?\\b|dog)(?!.*fake)\n\nSee the regex demo.\nDetails:\n\n(?<!fake.*) - a negative lookbehind that fails the match if there is fake string followed with any zero or more chars other than line break chars as many as possible immediately to the left of the current location\n(man[a-z]?\\b|dog) - man + a lowercase ASCII letter followed with a word boundary or dog string\n(?!.*fake) - a negative lookahead that fails the match if there are any zero or more chars other than line break chars as many as possible and then a fake string immediately to the left of the current location.\n\nIn Python:\nimport regex\nfor x in ['a man dogs', \"fake: too many dogs\", 'hi']:\n print(regex.findall(r\"(?<!fake.*)(man[a-z]?\\b|dog)(?!.*fake)\", x, flags=re.IGNORECASE))\n\n\n", "In your pattern (man[a-z]?\\b|dog)(?!^.*fake) the negative lookahead is after the match, but the word fake can still occur before one of the matches.\nWith Python re you can get out of the way what you don't want to keep, and capture what you want to keep using a capture group.\nWhat you could do is not use a negative lookahead, but match a whole line that contains the word fake\n^.*\\bfake\\b.*|(man[a-z]?\\b|dog)\n\nExplanation\n\n^.*\\bfake\\b.* Match a whole line that contains the word fake\n| Or\n(man[a-z]?\\b|dog) Capture group 1, match either man and optional char a-z or match dog\n\nRegex demo | Python demo\nimport re\n\npattern = r\"^.*\\bfake\\b.*|(man[a-z]?\\b|dog)\"\n\nfor x in ['a man dogs', \"fake: too many dogs\", 'hi']:\n res = [s for s in re.findall(pattern, x, re.IGNORECASE) if s]\n print(res)\n\nOutput\n['man', 'dog']\n[]\n[]\n\n" ]

[ 1, 1 ]

[]

[]

[ "python", "python_3.x", "python_re", "regex" ]

[ "No, there is not an exact equivalent in the python ibm_db for the db2look tool.\nAlternative approaches exist.\nNothing (except suitable authorities/permissions) prevents you from running a stored procedure that exececutes (i.e. shells out to) db2look on the database-server and return its output to the python script.\nIf the workstation running python ibm_db also has the Db2 fat client installed, then python can directly run db2look as long as your Db2 client has the relevant database(s) catalogued.\nYou can also use python to execute an undocumented DB2-LUW stored procedure (sysproc.DB2LK_GENERATE_DDL()) as described in this answer, subject to your account having relevant rights.\nYou can also write your own queries for the catalog views (i.e. re-invent the wheel), in order to generate the DDL, which lets you do whatever you want.\n" ]

[ 1 ]

[]

[]

[ "db2", "python" ]

[ "I'll try to help, but I'm not an expert!\nThe reason why http://127.0.0.1:8000/blog/static_files/blog/main.css is working is because you are accessing the main.css file directly. I would assume that your <link rel=\"stylesheet\" type=\"text/css\" href=\"{% static 'blog/main.css' % }\"> is not pointing to the correct directory. \nI looked at the documentation and it seems that your folder structure is not formed in the correct way. Here is a template for how everything should be structured:\n[projectname]/ <- project root\n├── [projectname]/ <- Django root\n│ ├── __init__.py\n│ ├── settings/\n│ │ ├── common.py\n│ │ ├── development.py\n│ │ ├── i18n.py\n│ │ ├── __init__.py\n│ │ └── production.py\n│ ├── urls.py\n│ └── wsgi.py\n├── apps/\n│ └── __init__.py\n├── configs/\n│ ├── apache2_vhost.sample\n│ └── README\n├── doc/\n│ ├── Makefile\n│ └── source/\n│ └── *snap*\n├── manage.py\n├── README.rst\n├── run/\n│ ├── media/\n│ │ └── README\n│ ├── README\n│ └── static/\n│ └── README\n├── static/\n│ └── README\n└── templates/\n ├── base.html\n ├── core\n │ └── login.html\n └── README\n\nOn the screenshot that you provided it looks as if your templates folder is inside the django root instead of the project root. That's important, because it will be searching for your .css in the wrong place. \nI encourage you to try fixing the directories and then instead of using href=\"{% static 'blog/main.css' % }\"> simply use href=\"css/style.css\". \nPlease notice that it's css/style.css instead of style.css. In your templates folder, you can create a folder called css and then store all of the .css files in there. Then you can simply point to the css folder with href=\"css/style.css\".\nThis should work as a temporary solution until someone more knowledgable can give you the correct answer. \nEDIT:\nYour base.html should be in the templates folder also! So: project_root/templates (insert base.html here) /css/ (insert your .css here(\n", "Note this out: Keeping Debug = False in your local might cause the error of css not loading. Change Debug = True to run your css in local with django.\n", "\"whitenoise\" can solve \"MIME type\" error then CSS is loaded:\nFirst, install \"whitenoise\":\npip install whitenoise\n\nThen, add it to \"MIDDLEWARE\" in \"settings.py\". That's it. Finally, CSS is loaded:\nMIDDLEWARE = [\n # ...\n \"django.middleware.security.SecurityMiddleware\",\n \"whitenoise.middleware.WhiteNoiseMiddleware\", # Here\n # ...\n]\n\n", "whitenoise can solve MIME type error then CSS is loaded:\nFirst, install whitenoise:\npip install whitenoise\n\nThen, add it to MIDDLEWARE in settings.py. That's it.\nFinally, CSS is loaded:\nMIDDLEWARE = [\n # ...\n \"django.middleware.security.SecurityMiddleware\",\n \"whitenoise.middleware.WhiteNoiseMiddleware\", # Here\n # ...\n ]\n\n" ]

[ 2, 2, 0, 0 ]

[]

[]

[ "css", "django", "django_staticfiles", "python", "python_3.x" ]

[ "I suspect that because the 2nd line in dibujarCirculo resets glColor3f to (0,0,0), you keep losing the change you made in keyPressed. Have you tried initializing glColor3f somewhere other than dibujarCirculo ?\n", "You can do something like this where you store the current shape color globally and update that when key presses are detected\nfrom OpenGL.GL import *\nfrom OpenGL.GLUT import *\nfrom math import pi \nfrom math import sin\nfrom math import cos\n\nimport random\n\nshapeColor = [0, 0, 0]\n\nnumAgents = 300\n\ndef initGL(width, height):\n glClearColor(0.529, 0.529, 0.529, 0.0)\n glMatrixMode(GL_PROJECTION)\n\ndef dibujarCirculo():\n glClear(GL_COLOR_BUFFER_BIT)\n glColor3f(shapeColor[0], shapeColor[1], shapeColor[2])\n\n glBegin(GL_POLYGON)\n for i in range(400):\n x = 0.25*sin(i) #Cordenadas polares x = r*sin(t) donde r = radio/2 (Circunferencia centrada en el origen)\n y = 0.25*cos(i) #Cordenadas polares y = r*cos(t)\n glVertex2f(x, y) \n glEnd()\n glFlush()\n\ndef keyPressed(*args):\n key = args[0]\n global shapeColor\n if key == b'r':\n shapeColor = [1,0,0]\n elif key == b'g':\n shapeColor = [0,1,0]\n elif key == b'b':\n shapeColor = [0,0,1] \n\ndef main():\n global window\n glutInit(sys.argv)\n glutInitDisplayMode(GLUT_SINGLE|GLUT_RGB)\n glutInitWindowSize(500,500)\n glutInitWindowPosition(200,200)\n\n #creando la ventana\n window = glutCreateWindow(\"Taller uno\")\n\n glutDisplayFunc(dibujarCirculo)\n glutIdleFunc(dibujarCirculo)\n glutKeyboardFunc(keyPressed)\n initGL(500,500)\n glutMainLoop()\n\nif __name__ == \"__main__\":\n main()\n\n" ]

[ 0, 0 ]

[]

[]

[ "opengl", "pyopengl", "python", "python_2.7" ]

[ "The offset is formatted differently depending on the useMathText argument. If True it will show the offset in a latex-like (MathText) format as x 10^6 instead of 1e6\nimport numpy as np\nimport matplotlib.pyplot as plt\nplt.figure()\nx = np.linspace(0,1000)\ny = x**2\nplt.plot(x,y)\nplt.ticklabel_format(style='sci', axis='y', scilimits=(0,0), useMathText=True)\nplt.show()\n\n\nNote that the above will not work for version 2.0.2 (possibly other older versions). In that case you need to set the formatter manually and specify the option:\nimport numpy as np\nimport matplotlib.pyplot as plt\nplt.figure()\nx = np.linspace(0,1000)\ny = x**2\nplt.plot(x,y)\nplt.gca().yaxis.set_major_formatter(plt.ScalarFormatter(useMathText=True))\nplt.ticklabel_format(style='sci', axis='y', scilimits=(0,0))\nplt.show()\n\n", "If you need this for all of your plots, you can edit the rcParams, for example as follows.\nimport matplotlib as mpl\n\nmpl.rc('axes.formatter', use_mathtext=True)\n\nPut at at the top of your script. If you want it for all of your scripts I recommend looking up matplotlib stylesheets.\n" ]

[ 4, 0 ]

[]

[]

[ "matplotlib", "python", "python_3.x" ]

[]

[]

[ "I don't see why your GPU RAM is exploding. Perhaps it helps if you debug your code following here:\nhttps://discuss.pytorch.org/t/how-to-debug-causes-of-gpu-memory-leaks/6741/13\nimport torch\nimport gc\nfor obj in gc.get_objects():\n try:\n if torch.is_tensor(obj) or (hasattr(obj, 'data') and torch.is_tensor(obj.data)):\n print(type(obj), obj.size())\n except:\n pass\n \n\nOr if you call torch.cuda.empty_cache() after del item_features according to here: https://discuss.pytorch.org/t/how-to-delete-a-tensor-in-gpu-to-free-up-memory/48879/3\n" ]

[ -1 ]

[ "clip", "gpu", "memory", "python" ]

[ "You can perform a merge operation and provide a list of columns that you want from Table_1.\nimport pandas as pd\n\nTable_1 = pd.DataFrame({\n \"Product_L1\":[\"Portland cement\", \"Portland cement\", \"Portland cement\", \"Portland cement\", \"Portland cement\", \"Portland cement\", \"Portland cement\"],\n \"Product_L2\":[\"Cement\", \"Cement\", \"Cement\", \"Cement\", \"Cement\", \"Cement\", \"Cement\"],\n \"Geography_L1\": [\"Peru\", \"Switzerland\", \"USA\", \"Brazil\", \"South Africa\", \"India\", \"Brazil\"],\n \"Geography_L2\": [\"South America\", \"Europe\", \"North America\", \"South America\", \"Africa\", \"Asia\", \"South America\"],\n \"Year\": [2021, 2021, 2021, 2021, 2021, 2021, 2020],\n \"obs_value\": [1, 2, 3, 4, 5, 6, 7]\n })\n\nTable_2 = pd.DataFrame({\n \"Product_L1\":[\"Portland cement\", \"Portland cement\"],\n \"Product_L2\":[\"Cement\", \"Cement\"],\n \"Geography_L1\": [\"Brazil\", \"Switzerland\"],\n \"Geography_L2\": [\"South America\", \"Europe\"],\n \"Year\": [2021, 2021]\n })\n\ncolumns_list = ['Product_L1','Product_L2','Geography_L1','Geography_L2','Year','obs_value']\n\nresult = pd.merge(Table_2, Table_1[columns_list], how='left')\n\nresult is a new dataframe:\n Product_L1 Product_L2 Geography_L1 Geography_L2 Year obs_value\n0 Portland cement Cement Brazil South America 2021 4\n1 Portland cement Cement Switzerland Europe 2021 2\n\nEDIT: Based upon the update to the question, I think what you are trying to do is achievable using set_index and unstack. This will create a new dataframe with the observed values listed in columns 'Year_2020', 'Year_2021' etc.\nindex_columns = ['Product_L1','Product_L2','Geography_L1','Geography_L2', 'Year']\nedit_df = Table_1.set_index(index_columns)['obs_value'].unstack().add_prefix('Year_').reset_index()\n\n" ]

[ 1 ]

[]

[]

[ "match", "python", "where_clause" ]

[ "Backoff means you go back to a n-1 gram level to calculate the probabilities when you encounter a word with prob=0. So in our case you will use a 3-gram model to calculate the probability of \"sunny\" in the context \"is a very\".\nThe most used scheme is called \"stupid backoff\" and whenever you go back 1 level you multiply the odds by 0.4. So if sunny exists in the 3-gram model the probability would be 0.4 * P(\"sunny\"|\"is a very\").\nYou can go back to the unigram model if needed multipliying by 0.4^n where n is the number of times you backed off.\n" ]

[ 2 ]

[]

[]

[ "nlp", "python" ]

[ "Interestingly, I found quite nice and thorough solutions in ROS community. http://wiki.ros.org/docker/Tutorials/GUI\nFor my problem, my final choice is the second way in the tutorial:\ndocker run --rm -it \\\n --user=$(id -u) \\\n --env=\"DISPLAY\" \\\n --workdir=/app \\\n --volume=\"$PWD\":/app \\\n --volume=\"/etc/group:/etc/group:ro\" \\\n --volume=\"/etc/passwd:/etc/passwd:ro\" \\\n --volume=\"/etc/shadow:/etc/shadow:ro\" \\\n --volume=\"/etc/sudoers.d:/etc/sudoers.d:ro\" \\\n --volume=\"/tmp/.X11-unix:/tmp/.X11-unix:rw\" \\\n deepaul python test.python\n\n", "As far as I know, there are two ways you can to this:\n\nYou can give Jupyter a try. Install Jupyter via Conda or pip, and then run Jupyter-notebook server. By exporting the server port of Jupyter, you can visit the Jupyter-notebook via a browser. You can then create a new python notebook and import the .py file you have, copy the code under your if __name__ == '__main__' to the new notebook if necessary. Finally, run the code in Jupyter, the image will show up below the code on the web page. matplotlib works smoothly with Jupyter. If you are willing to open a browser to run the code and view the result, this is the best way I can think of.\nYou can use the matplotlib headlessly. That means to remove all the code such as plt.show(). Use plt.savefig to save figures to filesystem instead of showing it in an opened window. Then you can check out these saved images using any image viewer.\n\nI tried mounting X11 to docker images some time ago, like YW P Kwon's answer. It will only work on systems that use X11, and you can do this only on a local machine (I am not sure if X11 forward works). It is also not recommended in docker. While with the Jupyter and Headless solution, you can run your code on any platform. But you do need to modify your code a little bit.\n", "I tried in my own way and it works fine..\n#To run \"matplotlib\" inside docker, you must install these inside containers:-#\n\n$yum install python3 -y //(must)\n$yum install python3-tkinter -y //(must)\n$pip3 install matplotlib // (must)\n$pip3 install pandas (if required)\n$pip3 install scikit-learn ( if required)\n\nNow, run your container.\n\n$docker run --rm -it --env=DISPLAY --workdir=/app -v=\"$PWD:/app <image_name> python3 model.py\n\n", "Bellow, I have described how to enable matplotlib plotting from running docker container and windows 10 machine.\nNote, I didn't have to change matplotlib backend using matplotlib.use('TkAgg') as maybe others suggested in other posts.\nFor test I used simple python program:\n# test.py\nimport matplotlib.pyplot as plt\nplt.plot([1,2], [3,4])\nplt.show()\n\nThe first thing is to add the following into the dockerfile, to install tkinter backend which matplotlib uses for graphical representation:\nRUN apt-get update -y\nRUN apt-get install -y libx11-dev\nRUN apt-get install -y python3-tk\n\nThe second thing is to install vcXsrv server onto windows machine, which enables and establishes communication from running container to windows machine. Install and run it - you may follow instructions described on this page.\nThe last thing to do before building and running the docker image is to set $DISPLAY variable in wsl terminal. I did this with:\nexport DISPLAY=IP:0.0\n\n, where IP is the network IP of windows machine.\nIn my case I ran:\nexport DISPLAY=172.16.31.103:0.0\n\nAfter successfully executed steps above the image can be built.\nBuild it with:\ndocker build -t image_name .\n\nThen run the container using command below where for display the before-set variable will be used.\ndocker run -ti --rm -e DISPLAY=$DISPLAY image_name\n\nIf everything was set correctly the matplotlib figure will pop-up.\n", "It looks like matplotlib is not meant to be used in docker. Use seaborn instead.\nI noteice, when encounter thet problem, that matplotlib was used for show and save figures. Seaborn can save files and it works fine in docker. Removing matplotlib and use seaborn might be solution.\n" ]

[ 25, 3, 3, 0, 0 ]

[]

[]

[ "docker", "matplotlib", "python" ]

[ "If need compare index values with both columns create Multiindex and use Index.isin:\ndf = df1[df1.set_index('A', append=True).index.isin(df2.set_index('B', append=True).index)]\nprint (df)\n A\n0 apple\n2 orange\n\n" ]

[ 2 ]

[]

[]

[ "pandas", "python" ]

[ "First we imported the Flask class. An instance of this class will be our WSGI application.\nNext we create an instance of this class. The first argument is the name of the application’s module or package. If you are using a single module (as in this example), you should use name because depending on if it’s started as application or imported as module the name will be different ('main' versus the actual import name). This is needed so that Flask knows where to look for templates, static files, and so on. For more information have a look at the Flask documentation.\nWe then use the route() decorator to tell Flask what URL should trigger our function.\nThe function is given a name which is also used to generate URLs for that particular function, and returns the message we want to display in the user’s browser.\n" ]

[ 0 ]

[]

[]

[ "flask", "python" ]

[ "You can use the built-in ast.literal_eval:\n>>> import ast\n>>> ast.literal_eval(\"{'muffin' : 'lolz', 'foo' : 'kitty'}\")\n{'muffin': 'lolz', 'foo': 'kitty'}\n\nThis is safer than using eval. As its own docs say:\n\n>>> help(ast.literal_eval)\nHelp on function literal_eval in module ast:\n\nliteral_eval(node_or_string)\n Safely evaluate an expression node or a string containing a Python\n expression. The string or node provided may only consist of the following\n Python literal structures: strings, numbers, tuples, lists, dicts, booleans,\n and None.\n\nFor example:\n>>> eval(\"shutil.rmtree('mongo')\")\nTraceback (most recent call last):\n File \"<stdin>\", line 1, in <module>\n File \"<string>\", line 1, in <module>\n File \"/opt/Python-2.6.1/lib/python2.6/shutil.py\", line 208, in rmtree\n onerror(os.listdir, path, sys.exc_info())\n File \"/opt/Python-2.6.1/lib/python2.6/shutil.py\", line 206, in rmtree\n names = os.listdir(path)\nOSError: [Errno 2] No such file or directory: 'mongo'\n>>> ast.literal_eval(\"shutil.rmtree('mongo')\")\nTraceback (most recent call last):\n File \"<stdin>\", line 1, in <module>\n File \"/opt/Python-2.6.1/lib/python2.6/ast.py\", line 68, in literal_eval\n return _convert(node_or_string)\n File \"/opt/Python-2.6.1/lib/python2.6/ast.py\", line 67, in _convert\n raise ValueError('malformed string')\nValueError: malformed string\n\n", "https://docs.python.org/3.8/library/json.html\nJSON can solve this problem though its decoder wants double quotes around keys and values. If you don't mind a replace hack...\nimport json\ns = \"{'muffin' : 'lolz', 'foo' : 'kitty'}\"\njson_acceptable_string = s.replace(\"'\", \"\\\"\")\nd = json.loads(json_acceptable_string)\n# d = {u'muffin': u'lolz', u'foo': u'kitty'}\n\nNOTE that if you have single quotes as a part of your keys or values this will fail due to improper character replacement. This solution is only recommended if you have a strong aversion to the eval solution.\nMore about json single quote: jQuery.parseJSON throws “Invalid JSON” error due to escaped single quote in JSON\n", "using json.loads:\n>>> import json\n>>> h = '{\"foo\":\"bar\", \"foo2\":\"bar2\"}'\n>>> d = json.loads(h)\n>>> d\n{u'foo': u'bar', u'foo2': u'bar2'}\n>>> type(d)\n<type 'dict'>\n\n", "To OP's example:\ns = \"{'muffin' : 'lolz', 'foo' : 'kitty'}\"\n\nWe can use Yaml to deal with this kind of non-standard json in string:\n>>> import yaml\n>>> s = \"{'muffin' : 'lolz', 'foo' : 'kitty'}\"\n>>> s\n\"{'muffin' : 'lolz', 'foo' : 'kitty'}\"\n>>> yaml.load(s)\n{'muffin': 'lolz', 'foo': 'kitty'}\n\n", "To summarize:\nimport ast, yaml, json, timeit\n\ndescs=['short string','long string']\nstrings=['{\"809001\":2,\"848545\":2,\"565828\":1}','{\"2979\":1,\"30581\":1,\"7296\":1,\"127256\":1,\"18803\":2,\"41619\":1,\"41312\":1,\"16837\":1,\"7253\":1,\"70075\":1,\"3453\":1,\"4126\":1,\"23599\":1,\"11465\":3,\"19172\":1,\"4019\":1,\"4775\":1,\"64225\":1,\"3235\":2,\"15593\":1,\"7528\":1,\"176840\":1,\"40022\":1,\"152854\":1,\"9878\":1,\"16156\":1,\"6512\":1,\"4138\":1,\"11090\":1,\"12259\":1,\"4934\":1,\"65581\":1,\"9747\":2,\"18290\":1,\"107981\":1,\"459762\":1,\"23177\":1,\"23246\":1,\"3591\":1,\"3671\":1,\"5767\":1,\"3930\":1,\"89507\":2,\"19293\":1,\"92797\":1,\"32444\":2,\"70089\":1,\"46549\":1,\"30988\":1,\"4613\":1,\"14042\":1,\"26298\":1,\"222972\":1,\"2982\":1,\"3932\":1,\"11134\":1,\"3084\":1,\"6516\":1,\"486617\":1,\"14475\":2,\"2127\":1,\"51359\":1,\"2662\":1,\"4121\":1,\"53848\":2,\"552967\":1,\"204081\":1,\"5675\":2,\"32433\":1,\"92448\":1}']\nfuncs=[json.loads,eval,ast.literal_eval,yaml.load]\n\nfor desc,string in zip(descs,strings):\n print('***',desc,'***')\n print('')\n for func in funcs:\n print(func.__module__+' '+func.__name__+':')\n %timeit func(string) \n print('')\n\nResults:\n*** short string ***\n\njson loads:\n4.47 µs ± 33.4 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)\nbuiltins eval:\n24.1 µs ± 163 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\nast literal_eval:\n30.4 µs ± 299 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\nyaml load:\n504 µs ± 1.29 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n\n*** long string ***\n\njson loads:\n29.6 µs ± 230 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\nbuiltins eval:\n219 µs ± 3.92 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\nast literal_eval:\n331 µs ± 1.89 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\nyaml load:\n9.02 ms ± 92.2 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n\nConclusion:\nprefer json.loads\n", "If the string can always be trusted, you could use eval (or use literal_eval as suggested; it's safe no matter what the string is.) Otherwise you need a parser. A JSON parser (such as simplejson) would work if he only ever stores content that fits with the JSON scheme.\n", "Use json. the ast library consumes a lot of memory and and slower. I have a process that needs to read a text file of 156Mb. Ast with 5 minutes delay for the conversion dictionary json and 1 minutes using 60% less memory!\n", "string = \"{'server1':'value','server2':'value'}\"\n\n#Now removing { and }\ns = string.replace(\"{\" ,\"\")\nfinalstring = s.replace(\"}\" , \"\")\n\n#Splitting the string based on , we get key value pairs\nlist = finalstring.split(\",\")\n\ndictionary ={}\nfor i in list:\n #Get Key Value pairs separately to store in dictionary\n keyvalue = i.split(\":\")\n\n #Replacing the single quotes in the leading.\n m= keyvalue[0].strip('\\'')\n m = m.replace(\"\\\"\", \"\")\n dictionary[m] = keyvalue[1].strip('\"\\'')\n\nprint dictionary\n\n", "Optimized code of Siva Kameswara Rao Munipalle\ns = s.replace(\"{\", \"\").replace(\"}\", \"\").split(\",\")\n \ndictionary = {}\n\nfor i in s:\n dictionary[i.split(\":\")[0].strip('\\'').replace(\"\\\"\", \"\")] = i.split(\":\")[1].strip('\"\\'')\n \nprint(dictionary)\n\n", "no any libs are used (python2):\ndict_format_string = \"{'1':'one', '2' : 'two'}\"\nd = {}\nelems = filter(str.isalnum,dict_format_string.split(\"'\"))\nvalues = elems[1::2]\nkeys = elems[0::2]\nd.update(zip(keys,values))\n\nNOTE: As it has hardcoded split(\"'\") will work only for strings where data is \"single quoted\".\nNOTE2: In python3 you need to wrap filter() to list() to get list.\n", "My string didn't have quotes inside:\ns = 'Date: 2022-11-29T10:57:01.024Z, Size: 910.11 KB'\nMy solution was to use str.split:\n{k:v for k, v in map(lambda d: d.split(': '), s.split(', '))}\n" ]

[ 1594, 368, 225, 52, 31, 25, 20, 11, 7, 6, 0 ]

[]

[]

[ "dictionary", "python", "string" ]

[ "pip3 show --files fb-idb shows where pip has installed all the files of the package. Run\npip3 show --files fb-idb | grep -F /bin/\n\nto extract the directory where pip installed scripts and entry points (On Windows it's \\Scripts\\). The directories are related to the header Location: so either do grep -F Location: separately or do it combined:\npip3 show --files fb-idb | grep 'Location:\\|/bin/'\n\n" ]

[ 1 ]

[]

[]

[ "pip", "python" ]

[ "I've found the solution here Windows: using decorator with multiprocessing\n" ]

[ 0 ]

[]

[]

[ "python" ]

[ "\nI want to read some variables from csv file, the problem that i cant use libraries in the c code so i am not able to read the csv file from the c code.\n\nAnd you don't need to. Simply this is how the file gets generated:\n[CSV File] -> Python -> [.h file]\n\nSo, you actually need to parse and convert the file in python, not in C.\nHow to read CSV data in python?\nYou've got plenty of ways. My preference is reading it the simplest way. Consider your csv is just some floating point number lines:\nwith open('file.csv', 'r') as file:\n data = file.readlines()\nrows = []\nfor line in data:\n rows.append([float(item.replace(' ', '')) for item in line.replace('\\n', '')])\n\nHow to generate header using extracted csv data?\nSimply, it depends on you. It's not hard. Are you going to use 2d arrays?\nLet me know if you've got any difficulties then I'll write a sample code for you. I don't find it hard to, though.\n" ]

[ 1 ]

[]

[]

[ "c", "python" ]

[ "For all links that you crawl from the main url, you need to do exactly the same as before (request, bs4, extract hrefs).\n\nThen check if href of link ends with \".pdf\"\nIf href is a relative path of the pdf file, use urllib to extract the domain from the website url and concatenate the domain and the pdf file name again\n\nE. g.:\nfrom urllib.parse import urlparse\n\ndomain = urlparse(\"https://www.europarl.europa.eu/doceo/document/EMPL-AM-544465_EN.html\").netloc\n\n\nDo another get request to retrieve the context of the pdf file\n\n", "You've done it right. You just need a simple modification:\nSolution for this specific problem\nYou have done the hardest part. Just replace \".html\" with \".pdf\" and then download the file is a manner like this:\nfor link in soup.find_all(\"a\",href=re.compile(\"AM\")):\n page_url = str(link.get('href'))\n pdf_url = page_url.replace('.html', '.pdf')\n pdf_response = requests.get(url)\n with open('/blah_blah_blah/metadata.pdf', 'wb') as f:\n f.write(response.content)\n\nWhat was the problem?\nDear friends, many websites have a dedicated webpage for any of their links. Usually it is so, for the sake of \"SEO\" purposes!\nThe good point is: \"There is often a relation between the page link and the target file's link.\nFor example here we have:\nhttps://www.europarl.europa.eu/doceo/document/EMPL-AM-544465_EN.html\n\nAnd when we check that page we find that the url is:\nhttps://www.europarl.europa.eu/doceo/document/EMPL-AM-544465_EN.pdf\n\nSo you just need to change the \".html\" tail to \".pdf\" and then go for a download.\nWhat to do if there is no relation?\nThen you should do exactly what a human does. Open/Fetch the page and then extract the content link explicitly. After that you can download the target file.\nDo we have even harder situations?\nYes. Sometimes the website has AJAX and ... operations to get the link. In those cases I suggest trying to follow the behavior of the browser and to check if there exists a pattern between the link (usually an API) and the content.\nBut there exists even harder situations that you will use alternatives like \"Selenium\".\n" ]

[ 1, 0 ]

[]

[]

[ "beautifulsoup", "parsing", "pdf", "python", "web_scraping" ]

[ "Chain both conditions by & for bitwise AND with helper Series:\ns = df['product name'].str[:6]\ndf.loc[df['ID'].isna() & ~s.str.contains(\"_\"), 'ID'] = s\nprint (df)\n ID product name\n0 1BJM10 1BJM10_RS2022_PK\n1 NaN L_RS2022_PK\n2 2PKL10 2PKL10_RS2022_PK\n3 3BDG10 3BDG10_RS2022_PK\n4 1BJM10 1BJM10_RS2022_PK\n\n", "Try:\ndf['ID'] = df['product name'].apply(lambda x: x[:x.find('_')] if x.find('_')>=6 else '')\n\n" ]

[ 2, 0 ]

[]

[]

[ "pandas", "python" ]

[]

[]

[ "It's because your json file is UTF-8. Add the UTF-8 enconding to your Choropleth class parameters:\nchoropleth = folium.Choropleth(\ngeo_data=state_geo,\nname='choropleth',\ndata=state_data,\ncolumns=['Name', 'Un'],\nkey_on='feature.id',\nfill_color='YlGn',\nfill_opacity=0.7,\nline_opacity=0.2,\nlegend_name='Unemployment',\nenconding='UTF-8'\n).add_to(m)\n\nSee post: Python - UnicodeDecodeError:\nYou could always run chardet to check:\nimport chardet \nrawdata=open(state_geo,'rb').read()\nchardet.detect(rawdata)\n\n" ]

[ -1 ]

[ "choropleth", "csv", "folium", "pandas", "python" ]

[ "You can analyse recursively you input data and remove all forbidden values:\ndef filter_data(data, forbidden_values):\n if isinstance(data, str):\n return data\n if isinstance(data, dict):\n return {\n key: filter_data(value, forbidden_values) for key, value in data.items()\n }\n if isinstance(data, list):\n return [\n filter_data(value, forbidden_values)\n for value in data\n if value not in forbidden_values\n ]\n\n\nforbidden_values = [\"aks9098\", \"aks9099\", \"aks9103\"]\nfiltered_data = filter_data(json_data, forbidden_values)\n\n" ]

[ 0 ]

[]

[]

[ "json", "python", "python_3.x" ]

[ "From the error message it looks like you are using a torch script model:\n\nSupport for converting Torch Script Models is experimental. If\npossible you should use a traced model for conversion.\n\nif possible try to use a traced model e.g.:\ndummy_input = torch.randn(batch, channels, width, height)\ntraceable_model = torch.jit.trace(model, dummy_input)\n\nfollowed by your original code:\nct.convert(traceable_model,...\n\n" ]

[ 0 ]

[]

[]

[ "coreml", "coremltools", "detectron", "python", "torchscript" ]

[ "Referring from this thread, you should remove many=True in PostModel_serializer.\nAlso it should be comment_list not comments_list.\n@api_view(['GET'])\ndef getPost(request, pk):\n post = PostModel.objects.filter(id=pk).first()\n comment_list = CommentModel.objects.filter(post_id=post.id)\n for i in comment_list:\n post.comments.add(i)\n serializer = PostModel_serializer(post)\n return Response(serializer.data)\n\n", "I think you did wrong while creating ManyToManyField().\nInstead of this:\ncomments = models.ManyToManyField('CommentModel') #Here you made mistake. you should not add single quote to CommentModel. I think that's why you got that error\n\nTry this:\ncomments = models.ManyToManyField(CommentModel)\n\n" ]

[ 1, 0 ]

[]

[]

[ "django", "django_queryset", "django_rest_framework", "django_serializer", "python" ]

[ "You're close! This should do what you want:\nimport numpy as np\nfrom PIL import Image\n\n# Load image and make into Numpy array\nim = np.array(Image.open('artistic-swirl.jpg'))\n\n# Make every second element on x-axis black\nim[:, 0::2] = [0,0,0]\n\n# Make every third element on y-axis black\nim[::3, :] = [0,0,0]\n\nNote: You should read a colon (:) as meaning \"all columns\" or \"all rows\" depending on whether it is specified for the first or second axis.\nTurns this:\n\ninto this:\n\nIf you are using Jupyter, based off your question, you'll need to add something like this at the end to actually display the result.\n%matplotlib notebook\nplt.imshow(im)\n\nOr, if you want to save a PNG with PIL, you can use:\nim.save('result.png')\n\n" ]

[ 0 ]

[]

[]

[ "image_processing", "jupyter", "jupyter_notebook", "python" ]

[ "\nNull values or empty strings: an ML algorithm will probably not accept such an input\nTruncated strings in numbers: ?\nString formatted numbers: numbers are always formatted as strings\nWeird date formats: an ML system will require huge samples before it learns rules that you can implement in two minutes\nBad or missing references between tables: how could an ML algorithm deal with this ???\n\nIMO, you forget the most important check: values out of the normal range. These ranges can be found by simple statistical observation or by... common sense.\n" ]

[ 2 ]

[]

[]

[ "algorithm", "anomaly_detection", "machine_learning", "python" ]

[ "The ListBucket() API call only returns 1000 objects at a time, so listing buckets with 100,000+ objects is very slow and best avoided. You have 3-4 million objects, so definitely avoid listing them!\nInstead, use Amazon S3 Inventory, which can provide a daily or weekly CSV file listing all objects in a bucket. Activate it for both buckets, and then use the provided CSV files as input to make the comparison.\nYou can also use the CSV file generated by Amazon S3 Inventory as a manifest file for Amazon S3 Batch Operations. So, your code could generate a file that lists only the objects that you would like deleted, then get S3 Batch Operations to process those deletions.\n" ]

[ 0 ]

[]

[]

[ "amazon_web_services", "boto3", "paginator", "python" ]

[ "Unfortunately that is not possible. However you could re-export the original model from PyTorch to onnx, and add the output of the desired layer to the return statement of the forward method of your model. (you might have to feed it through a couple of methods up to the first forward method in your model)\n" ]

[ 0 ]

[]

[]

[ "conv_neural_network", "image_processing", "onnx", "python", "tensorflow" ]

[ "Use Resample.agg with list of functions, then reshape by DataFrame.stack and remove second level of MultiIndex by Series.droplevel:\ns = df.resample('D')['Random'].agg(['min','mean','max']).stack().droplevel(1)\nprint (s)\n2022-01-01 0.162976\n2022-01-01 0.574074\n2022-01-01 0.980742\n2022-01-02 0.012299\n2022-01-02 0.467338\n2022-01-02 0.962570\n2022-01-03 0.000722\n2022-01-03 0.426793\n2022-01-03 0.947014\ndtype: float64\n\n" ]

[ 1 ]

[]

[]

[ "pandas", "python" ]

[ "Are you sure you are not trying to multiply sinc functions, sin(x*u)/(x*u)? Currently you are multiplying terms like u * sin(x*u) / x because there are not parenthesis in the denominator.\nYou should be able to fit your function for small a,b,c. But having a = 100, you should have a much higher resolution, I would say steps.\nI am asuming you are trying to fit using some local minimization method.\nIf you have a function with more than many maxima and minima while you are trying to fit you are likely to get stuck. You could try some of non-convex optimization methods available\nas well\n" ]

[ 0 ]

[]

[]

[ "custom_function", "integral", "numpy", "python" ]

[ "Try to use \"name\" field in Series Type to get index value, and then get the same index for the other DataFrame\nimport pandas as pd\nimport numpy as np\n\n\ndef postional_sum(inot, df1, df2, df3):\n \"\"\"\n Get input index and gather the same position for the other DataFrame collection\n \"\"\"\n\n position = inot.name\n\n x = df1.iloc[position].sum()\n y = df2.iloc[position].sum()\n z = df3.iloc[position].sum()\n return x + y + z\n\n\n# dataframe rows as lists\ntsol = pd.DataFrame(np.random.randn(10, 5), columns=range(5))\nrfsol = pd.DataFrame(np.random.randn(10, 5), columns=range(5))\nrbsol = pd.DataFrame(np.random.randn(10, 5), columns=range(5))\n\nout2 = pd.DataFrame(index=range(len(tsol)), columns=[\"output\"])\n\nout2[\"output\"] = out2.apply(lambda x: postional_sum(x, tsol, rfsol, rbsol), axis=1)\n\nout2\n\nHope this helps!\n", "When you run df.apply() with axis=1, it does not pass on the columns as individual arguments to the function, but as a Series object, as explained here. The correct way to do this would be\nout2['output'] = out2.apply(lambda row:\n simple_func(t=row[\"tsol\"],\n rf=row[\"rfsol\"],\n rb=row[\"rbsol\"]),\n axis=1)\n\n", "You can eliminate the simple function using this:\nout2[\"output\"] = tsol.sum(axis=1) + rfsol.sum(axis=1) + rbsol.sum(axis=1)\n\nHere is the complete code:\nt=[1,2,3,4,5]\nrf=[6,7,8,9,10]\nrb=[11,12,13,14,15]\n\n# dataframe rows as lists\ntsol=pd.DataFrame((t,t,t,t,t,t,t,t,t,t))\nrfsol=pd.DataFrame((rf,rf,rf,rf,rf,rf,rf,rf,rf,rf))\nrbsol=pd.DataFrame((rb,rb,rb,rb,rb,rb,rb,rb,rb,rb))\n\nout2 = pd.DataFrame(index=range(len(tsol)), columns=[\"output\"])\nout2[\"output\"] = tsol.sum(axis=1) + rfsol.sum(axis=1) + rbsol.sum(axis=1)\n\nprint(out2)\n\nOUTPUT:\n output\n0 120\n1 120\n2 120\n3 120\n4 120\n5 120\n6 120\n7 120\n8 120\n9 120\n\n" ]

[ 2, 0, 0 ]

[]

[]

[ "apply", "function", "lambda", "pandas", "python" ]

[ "The connection could not be established because the site can't be reached.\nJust execute the get request inside your try/except and it will work.\ndef abnormal_url(url):\n domainname = urlparse(url).netloc\n domain = whois.whois(domainname)\n try:\n response = requests.get(url,verify=False)\n if response.text == domain:\n return 0 # legitimate\n else:\n return 1 # phishing\n except:\n return 1 # phishing\n\n" ]

[ 1 ]

[]

[]

[ "dataframe", "jupyter_notebook", "python", "python_requests" ]

[ "You can use a filter to add your custom attribute. For example :\ndef add_my_custom_attribute(record):\n record.myAttribute = 'myValue'\n record.username = record.request.user.username \n return True\n\nLOGGING = {\n 'version': 1,\n 'disable_existing_loggers': False,\n 'filters': {\n ...\n 'add_my_custom_attribute': {\n '()': 'django.utils.log.CallbackFilter',\n 'callback': add_my_custom_attribute,\n }\n },\n 'handlers': {\n ...\n 'django.server': {\n 'level': 'INFO',\n 'class': 'logging.StreamHandler',\n 'filters': ['add_my_custom_attribute'],\n 'formatter': 'django.server',\n }, \n },\n ...\n}\n\n\nBy installing a filter, you can process each log record and decide whether it should be passed from logger to handler. \nThe filter get the log record which contains all the details of log (i.e : time, severity, request, status code). \nThe attributes of the record are used by the formatter to format it to string message. If you add your custom attributes to that record - they will also be available to the formatter.\n", "The extra keyword is not a workaround. That's the most eloquent way of writing customized formatters, unless you are writing a custom logging altogether.\nformat: '%(asctime).19s %(levelname)s - %(username)s: %(message)s'\nlogging.basicConfig(format=format)\nlogger.info(message, extra={'username' : request.user.username})\n\nSome note from the documentation (**kwars for Django logger):\n\nThe keys in the dictionary passed in extra should not clash with the keys used by the logging system.\nIf the strings expected by the Formatter are missing, the message will not be logged.\nThis feature is intended for use in specialized circumstances, and not always.\n\n", "I will provide one of many possible complete answers to this question:\n\nHow can Django use logging to log using custom attributes in the formatter? I'm thinking of logging the logged in username for example.\n\nOther answers touched on the way to add extra contextual info through python's logging utilities. The method of using filters to attach additional information to the log record is ideal, and best described in the docs:\nhttps://docs.python.org/3/howto/logging-cookbook.html#using-filters-to-impart-contextual-information\nThis still does not tell us how to get the user from the request in a universal way. The following library does this:\nhttps://github.com/ninemoreminutes/django-crum\nThus, combine the two, and you will have a complete answer to the question that has been asked.\nimport logging\nfrom crum import get_current_user\n\nclass ContextFilter(logging.Filter):\n \"\"\"\n This is a filter injects the Django user\n \"\"\"\n\n def filter(self, record):\n\n record.user = get_current_user()\n return True\n\nif __name__ == '__main__':\n logging.basicConfig(level=logging.DEBUG,\n format='User: %(user)-8s %(message)s')\n a1 = logging.getLogger('a.b.c')\n\n f = ContextFilter()\n a1.addFilter(f)\n a1.debug('A debug message')\n\nThis will need to happen within a Django request-response cycle with the CRUM library properly installed.\n", "I came accross this old question while looking for a solution to display username in Django request logs.\nResponse of napuzba helped a lot, but it didn't work right away, because the user doesn't always exist in the request object.\nIt resulted in this piece of code in Django settings.py:\ndef add_username(record):\n \"\"\"\n Adds request user username to the log\n :param record:\n :return:\n \"\"\"\n try:\n record.username = record.request.user.username\n except AttributeError:\n record.username = \"\"\n return True\n \n\nLOGGING = {\n ...\n \"formatters\": {\n \"verbose\": {\n \"format\": \"[%(asctime)s] %(levelname)s [%(name)s:%(module)s] [%(username)s] %(message)s\",\n \"datefmt\": \"%d.%m.%Y %H:%M:%S\",\n },\n },\n \"filters\": {\n \"add_username\": {\n \"()\": \"django.utils.log.CallbackFilter\",\n \"callback\": add_username,\n }\n },\n \"handlers\": { \n \"console\": {\n \"level\": \"DEBUG\",\n \"class\": \"logging.StreamHandler\",\n \"formatter\": \"verbose\",\n \"filters\": [\"add_username\"],\n },\n }\n}\n\n" ]

[ 33, 3, 2, 0 ]

[]

[]

[ "django", "logging", "python" ]

[ "It looks like you are using numpy arrays. With numpy, it is faster to use vectorized operations on whole arrays at the same time, compared to loops. It also usually gives cleaner code.\nAs I understand it, you want to extract the color corresponding to the smallest (euclidean) distance to a specific point.\nOutputs = []\n\nfor X2, Y2 in X:\n XD = X2 - Velocity_Momentum[:, 0]\n YD = Y2 - Velocity_Momentum[:, 1]\n Distance = (XD * XD) + (YD * YD)\n Color_Gradient = Color[Distance.argmin()]\n Outputs.append(Color_Gradient)\n print(\"X2 = \", X2, \" Y2 = \", Y2, \" Color = \", Color_Gradient)\n\n" ]

[ 3 ]

[]

[]

[ "performance", "process", "python", "python_3.x" ]

[ "This kind of solution worked for me.\nclass ArticleDetailSerializer(serializers.ModelSerializer):\n author = ArticleAuthorSerializer(read_only=True)\n comments = serializers.SerializerMethodField()\n\n class Meta:\n model = Article\n fields = '__all__'\n\n def get_comments(self, obj):\n qs = obj.comments.all().get_cached_trees()\n return CommentSerializer(qs, many=True).data\n\nThis reduced all queries from 11 SQL to 4.\n" ]

[ 0 ]

[]

[]

[ "django", "django_mptt", "django_rest_framework", "python" ]

[ "You've imported the wrong thing; you've done from datetime import datetime so that datetime now refers to the class, not the containing module.\nEither do:\nimport datetime\n...article.created_on + datetime.timedelta(...)\n\nor\nfrom datetime import datetime, timedelta\n...article.created_on + timedelta(...)\n\n" ]

[ 8 ]

[ "You should use the correct import:\nfrom datetime import timedelta\n\nnew_date = article.created_on + timedelta(0, elapsed_time_in_seconds)\n\n", "I got the same issue, i solved it by replacing\nfrom datetime import datetime as my_datetime\nwith\nimport datetime as my_datetime\n" ]

[ -1, -1 ]

[ "datetime", "django", "python", "python_3.x" ]

[ "I have reproduced above and got the below results.\nThis my csv file in data container.\n\nThis is my mounting:\n\nI have mounted this and when I tried to create table from CSV file, I got same error.\n\nThe above error arises when we don't give correct path in the csv file. In file path after /mnt give mount point(here onemount for me) not container name as we already mounted till container.\n\nResult:\n\n" ]

[ 0 ]

[]

[]

[ "azure", "azure_databricks", "databricks", "python", "sql" ]

[ "As you said, you can simply write a second serializer method. So you keep the other one for your /uploads API call.\n# User serializer\ndef serialize_user(self):\n if self.uploads:\n uploads = [upload.serialize_upload_bis() for upload in self.uploads]\n return {\n \"email\": self.email,\n \"password\": self.password,\n \"username\": self.username,\n \"uploads\": uploads\n }\n\n# Upload serializer\ndef serialize_upload_bis(self):\n return {\n \"name\": self.name,\n \"path_to_file\": self.path_to_file,\n }\n\ndef serialize_upload(self):\n if self.user:\n dict_user = self.user.serialize_user()\n return {\n \"name\": self.name,\n \"path_to_file\": self.path_to_file,\n \"user\": dict_user\n }\n\n", "Although the question is quite old, I want to add a more generic answer, because I also was faced with this issue. The following serializer works for all classes with relationships:\nfrom sqlalchemy.orm.collections import InstrumentedList\nclass Serializer(object):\n\n def serialize(self):\n serializedObject= {}\n for c in inspect(self).attrs.keys():\n attribute = getattr(self, c)\n if(type(attribute) is InstrumentedList ):\n serializedObject[c]= Component.serialize_list(attribute)\n else:\n serializedObject[c]= attribute \n return serializedObject\n\n @staticmethod\n def serialize_list(l):\n return [m.serialize() for m in l]\n\n" ]

[ 0, 0 ]

[]

[]

[ "flask_sqlalchemy", "json", "python", "sqlalchemy" ]

[ "Quoting from the docs (shortened):\n\nRaises a GeneratorExit at the point where the generator function was\npaused.\n\n\nclose() does nothing if the generator has already exited\n\nWhen it was not started, it couldn't be paused or exited. The case it was not started (with an initial next or send) is not mentioned.\nBut we could easily check that not a single statemenet from the generator function was executed before .close() and no statement could be executed after the .close().\nThat is the only logical behaviour, IMO. There is nothing to be cleaned up in the generator function, because it did not get a chance to run.\n" ]

[ 1 ]

[]

[]

[ "generator", "python" ]

[ "My suggestion is not the perfect one, I didn't find the way to use your time blocks as axis labels. But it could be a starting point)\nSo here is my suggestion:\n\ncreate a separate column with start of your time blocks using split function and transforming this column to a datetime type\n\ndata['start_block']=data['time_block'].str.split(\" - \", expand=True)[0]\ndata['start_block']= pd.to_datetime(data['start_block'], format='%I:%M%p')\n\n\nuse time for axis encoding\n\nalt.Chart(data).mark_line(point=True).encode(\n x=alt.X('start_block:T',title=\"Time\"),\n y='blocks'\n)\n\nso you will have something like that.\n\n" ]

[ 0 ]

[]

[]

[ "altair", "datetime", "pandas", "python", "x_axis" ]

[ "You can use built-in max and min functions:\ndef q():\n greater, smaller = max(x, y), min(x, y)\n que = int(input(\"what is {}-{}?\".format(greater, smaller)))\n ...\n\n", "You can use:\ndef q(x, y):\n if y > x:\n return q(y, x)\n return int(input(\"what is {}-{}?\".format(x,y)))\n\n", "You can sort the numbers as a tuple first:\nx, y = sorted((random.randint(1,10), random.randint(1,10)), reverse=True)\n\n", "I would do something like this:\nimport random\n\nbigger = random.randint(1, 10)\nlower = random.randit(1, bigger)\n\ndef q(bigger, lower):\n\n que = int(input(f'What is {bigger} - {lower}?'))\n return que\n\n", "You can also use abs() function\n#random floating number\nfloating = -1.33\nprint('Absolute value of -1.33 is:', abs(floating)) # 1.33\n\n" ]

[ 2, 1, 1, 0, 0 ]

[]

[]

[ "python" ]

[ "For psycopg2.OperationalError: SCRAM authentication requires libpq version 10 or above follow the below work arounds:\nSolution :1\nDownload libpq.dll from https://www.exefiles.com/en/dll/libpq-dll/ then replace old libpq.dll at php directory with the latest downloaded\nSolution :2\nChange authentication to md5, then reset your password and restart the postgresql service and here are step by step:\n\nFind file postgresql.conf in C:\\Program Files\\PostgreSQL\\13\\data then set password_encryption = md5\nFind file pg_hba.conf in C:\\Program Files\\PostgreSQL\\13\\data then change all METHOD to md5\nOpen command line (cmd,cmder,git bash...) and run psql -U postgres then enter your password when installed postgres sql\n-Then change your password by running ALTER USER postgres WITH PASSWORD 'new-password'; in command line\nRestart service postgresql in your Service\n\nSolution :3\nCheck if psycopg is using the additional copy of libpq that may be present on your computer. Recognize that file, then upgrade or remove it. Perhaps psycopg has to be updated for that.\n" ]

[ 0 ]

[]

[]

[ "kubernetes", "psycopg2", "python" ]

[ "The error you are facing is due to type annotation.\nAs per https://github.com/pydantic/pydantic/issues/156 this is not yet fixed, you can try using pydantic.Field and then pass the regex argument there like so\ninstall_component_version: str = Field(regex=r\"^[0-9]+.[0-9]+.[0-9]$\")\nThis way you get the regex validation and type checking.\nPS: This is not a 100% alternative to constr but if all you want is regex validation, the above alternative works and makes mypy happy.\n" ]

[ 1 ]

[]

[]

[ "pydantic", "python" ]

[ "Two solutions were provided to this problem on Github issue \"Define a Python class across multiple cells #1243\" which can be found here: https://github.com/jupyter/notebook/issues/1243\nOne solution is using a magic function from a package developed for this specific case called jdc - or Jupyter dynamic classes. The documentation on how to install it and how to use can be found on package url at https://alexhagen.github.io/jdc/\nThe second solution was provided by Doug Blank and which just work in regular Python, without resorting to any extra magic as follows:\nCell 1:\nclass MyClass():\n def method1(self):\n print(\"method1\")\n\nCell 2:\nclass MyClass(MyClass):\n def method2(self):\n print(\"method2\")\n\nCell 3:\ninstance = MyClass()\ninstance.method1()\ninstance.method2()\n\nI tested the second solution myself in both Jupyter Notebook and VS Code, and it worked fine in both environments, except that I got a pylint error [pylint] E0102:class already defined line 5 in VS Code, which is kind of expected but still runs fine. Moreover, VS Code was not meant to be the target environment anyway.\n", "I don't feel like that whole stuff to be a issue or a good idea... But maybe the following will work for you:\n\n# First cell\nclass Foo(object):\n pass\n\n\n# Other cell\ndef __init__(self, var):\n self.var = var\n\nFoo.__init__ = __init__\n\n\n# Yet another cell\ndef print_var(self):\n print(self.var)\nFoo.print_var = print_var\n\n\nI don't expect it to be extremely robust, but... it should work for regular classes.\nEDIT: I believe that there are a couple of situations where this may break. I am not sure if that will resist code inspection, given that the method lives \"far\" from the class. But you are using a notebook, so code inspection should not be an issue (?), although keep that in mind if debugging.\nAnother possible issue can be related to use of metaclasses. If you try to use metaclasses (or derive from some class which uses a metaclass) that may broke it, because metaclasses typically expect to be able to know all the methods of the class, and by dynamically adding methods to a class, we are bending the rules on the flow of class creation.\nWithout metaclasses or some \"quite-strange\" use cases, the approach should be safe-ish.\nFor \"simple\" classes, it is a perfectly valid approach. But... it is not exactly an expected feature, so (ab)using it may give some additional problems which I may not \n", "Here's a decorator which lets you add members to a class:\nimport functools\ndef update_class(\n main_class=None, exclude=(\"__module__\", \"__name__\", \"__dict__\", \"__weakref__\")\n):\n \"\"\"Class decorator. Adds all methods and members from the wrapped class to main_class\n\n Args:\n - main_class: class to which to append members. Defaults to the class with the same name as the wrapped class\n - exclude: black-list of members which should not be copied\n \"\"\"\n\n def decorates(main_class, exclude, appended_class):\n if main_class is None:\n main_class = globals()[appended_class.__name__]\n for k, v in appended_class.__dict__.items():\n if k not in exclude:\n setattr(main_class, k, v)\n return main_class\n\n return functools.partial(decorates, main_class, exclude)\n\nUse it like this:\n#%% Cell 1\nclass MyClass:\n def method1(self):\n print(\"method1\")\nme = MyClass()\n\n#%% Cell 2\n@update_class()\nclass MyClass:\n def method2(self):\n print(\"method2\")\nme.method1()\nme.method2()\n\nThis solution has the following benefits:\n\npure python\nDoesn't change the inheritance order\nEffects existing instances\n\n", "There is no way to split a single class,\nYou could however, add methods dynamically to an instance of it\nCELL #1\nimport types\nclass A:\n def __init__(self, var):\n self.var = var\n\na = A()\n\nAnd in a different cell:\nCELL #2\ndef print_var(self):\n print (self.var)\na.print_var = types.MethodType( print_var, a )\n\nNow, this should work:\nCELL #3\na.print_var()\n\n", "Medhat Omr's answer provides some good options; another one I found that I thought someone might find useful is to dynamically assign methods to a class using a decorator function. For example, we can create a higher-order function like the one below, which takes some arbitrary function, gets its name as a string, and assigns it as a class method.\ndef classMethod(func):\n setattr(MyClass, func.__name__, func)\n return func\n\nWe can then use the syntactic sugar for a decorator above each method that should be bound to the class;\n@classMethod\ndef get_numpy(self):\n return np.array(self.data)\n\nThis way, each method can be stored in a different Jupyter notebook cell and the class will be updated with the new function each time the cell is run.\nI should also note that since this initializes the methods as functions in the global scope, it might be a good idea to prefix them with an underscore or letter to avoid name conflicts (then replace func.__name__ with func.__name__[1:] or however characters at the beginning of each name you want to omit. The method will still have the \"mangled\" name since it is the same object, so be wary of this if you need to programmatically access the method name somewhere else in your program.\n", "thanks@Medhat Omr, it works for me for the @classmethod as well.\nBase class in the first cell\nclass Employee:\n # define two class variables\n num_empl = 0\n raise_amt = 1.05\n\n def __init__(self, first, last, pay):\n\n self.first = first \n self.last = last\n self.pay = pay\n ...\n ...\n\n@classmethod in an another cell:\nclass Employee(Employee):\n\n @classmethod\n def set_raise_amt(cls, amount):\n cls.raise_amt = amount\n\n\nempl = Employee(\"Jahn\", \"Smith\", 65000)\nEmployee.set_raise_amt(1.04)\nprint(empl.full_name() + \" is getting \" + str(empl.apply_raise()))\n\n" ]

[ 23, 8, 3, 1, 0, 0 ]

[]

[]

[ "jupyter_notebook", "python" ]

[ "What is it that you're trying to achieve?\nI don't think that key_fn is related as it only affects the ordering of the vocabulary (and top k when provided)\nCould you compute the vocabulary after appending the added information?\ntft.vocabulary(tf.strings.join([words, <mask>, <pad>]), ...)\nThis would result in the vocabulary including the added suffix\n" ]

[ 0 ]

[]

[]

[ "mlops", "python", "tensorflow", "tensorflow_transform", "tfx" ]

[ "Take a look at line ParseError: mismatched tag: line 449, column 3.\nline 449 is the line number in your source XML file.\nFind this line and look what is wrong with the content.\nProbably this line contains some tag (e.g. closing) which has no\nopening conterpart.\nAn alternative: Visit any XML validation site and check what is wrong\nwith your file.\n", "i had the same issue and simply renamed the name of my pictures that had this naming convention from frame15148 - Copy.jpg to frame15148.jpg and the mismatch error disappeared Picture of rename .\n" ]

[ 5, 0 ]

[ "I find that sometimes you get this error at a line number a lot later than the actual error. It only notices something wrong when you close a tag that you didn't correctly open. Look earlier in the file.\nFor example I had a similar problem where the error message claimed\n\nxml.etree.ElementTree.ParseError: mismatched tag: line 12, column 2\n\nbut the mistake wasn't at the mentioned line, I had misspelled one tag opening at the beginning.\nCheck backwards through your input for the opening tag of the one that's given you this error.\n" ]

[ -1 ]

[ "elementtree", "parsing", "python", "xml", "xml_parsing" ]

[ "I mean the most natural way to do this imo would be\nfor key,val in dict1.items():\n assert val == dict2[key]\n\n" ]

[ 1 ]

[]

[]

[ "pytest", "python" ]

[ "You can use groupby.idxmax to get the index of the max per custom group:\n# get the indices of the max value per group\nidx = (df[df['State']].groupby(['room_id', (~df['State']).cumsum()])\n ['temperature'].idxmax()\n )\n\n# assign the new value\ndf.loc[idx, 'max_temp'] = 'max'\n\nIf you want the temperature value instead of a literax 'max':\ndf.loc[idx, 'max'] = df.loc[idx, 'temperature']\n\nOutput:\n room_id temperature State max max_temp\n0 117 1.489000 True max 1.489000\n1 117 8.900000 False NaN NaN\n2 117 2.500000 False NaN NaN\n3 117 4.370000 False NaN NaN\n4 117 4.363333 True max 4.363333\n5 117 4.356667 True NaN NaN\n6 118 4.350000 True NaN NaN\n7 118 6.648000 True max 6.648000\n8 118 6.642667 True NaN NaN\n9 118 7.300000 False NaN NaN\n10 118 9.400000 False NaN NaN\n11 118 5.300000 True NaN NaN\n12 118 7.100000 True max 7.100000\n\n\n" ]

[ 0 ]

[]

[]

[ "group_by", "intervals", "max", "pandas", "python" ]

[]

[]

[ "Solved it after posting. Changed the variable in the for loop from i to smth else, it reset the whole thing. If anyone has an explanation, can write it for other people.\n" ]

[ -1 ]

[ "dataframe", "python" ]

[ "Remove (or rename) the file matplotlib.py from your current working directory. It shadows the real library with the same name.\n", "I have a simple solution，delete your __init__.pyc and __init__.py files in your project dictionary. because i encounter the problem too，I have been solve it perfectly use this method.\n", "It worked for me\npip install pyparsing==2.4.7\n\n" ]

[ 12, 0, 0 ]

[]

[]

[ "import", "pandas", "python" ]

[ "Do you have to use the Records-class? If not, pandas offers a very clean implementation of what you are looking for:\nimport datetime\nimport pandas as pd\nimport numpy as np\n\ndf = pd.DataFrame([[datetime.date(2017, 8, 14), datetime.date(2018, 3, 5)],\n [datetime.date(2019, 1, 14), datetime.date(2019, 3, 6)],\n [datetime.date(2019, 3, 7), datetime.date(2019, 3, 20)],\n [datetime.date(2023, 12, 30), datetime.date(9999, 12, 31)],\n [datetime.date(2020, 9, 8), datetime.date(2020, 9, 14)],\n [datetime.date(2020, 9, 15), datetime.date(2020, 10, 14)],\n [datetime.date(2019, 3, 21), datetime.date(2019, 3, 30)]],\n columns=['start', 'end'])\n\ndf = df.sort_values('start').reset_index(drop=True)\nmask = df['start'] - pd.to_timedelta('1 day') == df['end'].shift(1)\ndf.loc[mask.shift(-1).fillna(False), 'end'] = np.nan\ndf['end'] = df['end'].bfill()\ndf = df[~mask]\nprint(df)\n\nAnd even if you have to use your class, you could just create it after you have done the data handling in pandas by running:\nresultLst = df.apply(lambda x: Records(x['start'], x['end']), axis=1).tolist()\n\nEDIT:\nUnfortunately, it is not really easy to understand what your underlying rules are, but the following works out almost the same way as what you say:\ndf = df.groupby('end').min().reset_index() # If two end dates are identical, we keep the first?\ndf = df.sort_values('start').reset_index(drop=True)\ndf['start_reduced'] = df['start'] - pd.to_timedelta('1 day')\ndf['idx_orig'] = df.index\ncols_to_drop = [x+'_y' for x in df.columns]\nfirst_iter = True\nseed_start_idx = []\nwhile first_iter or mask.any():\n df = df.merge(df, how='left', left_on='end', right_on='start_reduced', suffixes=('', '_y'))\n mask = ~df['end_y'].isna()\n df.loc[mask, 'end'] = df.loc[mask, 'end_y'].values\n if first_iter:\n seed_start_idx = df.loc[~df['start'].isin(df.loc[mask, 'start_y']), 'idx_orig'].tolist()\n df = df.drop(columns=cols_to_drop)\n first_iter = False\ndf = df[df['idx_orig'].isin(seed_start_idx)].drop_duplicates(subset='idx_orig', keep='last').drop(columns=['start_reduced', 'idx_orig'])\n\nThe only difference is that it is not possible to distinguish which of the ones starting 2011-01-01 should be kept. You state that the one ending 2011-03-04 should not be kept, but the one ending 9999-12-31 should seemingly be kept. I cannot understand the logic behind that differentiation. The rest works though.\n" ]

[ 1 ]

[]

[]

[ "python" ]

[ "In addition to\nimport nltk\nnltk.download(\"wordnet\")\n\nI also had to run this:\nimport nltk\nnltk.download('omw-1.4')\n\nCan you share the contents of your nltk_data directory (usually it is either C:\\Users\\yourusername\\nltk_data\\ or C:\\Users\\yourusername\\AppData\\Roaming\\nltk_data)? Also which version of nltk are you running? (you can get the version by running nltk.version )\n" ]

[ 0 ]

[]

[]

[ "lemmatization", "nltk", "python" ]

[ "You can do something like follows:\ndef get_repr(value): \n if callable(value):\n return '%s' % value()\n return value\n\ndef get_field(instance, field):\n field_path = field.split('.')\n attr = instance\n for elem in field_path:\n try:\n attr = getattr(attr, elem)\n except AttributeError:\n return None\n return attr\n\nfor field in fields:\n context['project'][field] = get_repr(get_field(project, field))\n\n", "Here's an equivalent recursive getattr -\ndef get_foreign_key_attr(obj, field: str):\n \"\"\"Get attr recursively by following foreign key relations\n\n For example ...\n get_foreign_key_attr(\n <Station: Spire operational 1m>, \"idproject__idcountry__name\"\n )\n ... splits \"idproject__idcountry__name\" into [\"idproject\", \"idcountry\", \"name\"]\n & follows finds the value of each foreign key relation\n \"\"\"\n fields = field.split(\"__\")\n if len(fields) == 1:\n return getattr(obj, fields[0], \"\")\n else:\n first_field = fields[0]\n remaining_fields = \"__\".join(fields[1:])\n return get_foreign_key_attr(getattr(obj, first_field), remaining_fields)\n\n" ]

[ 19, 0 ]

[]

[]

[ "callable", "getattr", "model", "python", "relational_database" ]