Datasets:

---

suriyagunasekar

/

stackoverflow-python-with-meta-data

like 10

[ "This is not really something to worry about in most cases. The warning comes from this function, here the most important part:\n...\n env_configured = False\n n_cores = detect_number_of_cores()\n if 'NUMEXPR_MAX_THREADS' in os.environ:\n # The user has configured NumExpr in the expected way, so suppress logs.\n env_configured = True\n n_cores = MAX_THREADS\n...\n if 'NUMEXPR_NUM_THREADS' in os.environ:\n requested_threads = int(os.environ['NUMEXPR_NUM_THREADS'])\n elif 'OMP_NUM_THREADS' in os.environ:\n requested_threads = int(os.environ['OMP_NUM_THREADS'])\n else:\n requested_threads = n_cores\n if not env_configured:\n log.info('NumExpr defaulting to %d threads.'%n_cores)\n\nSo if neither NUMEXPR_MAX_THREADS nor NUMEXPR_NUM_THREADS nor OMP_NUM_THREADS are set, NumExpr uses so many threads as there are cores (even if the documentation says \"at most 8\", yet this is not what I see in the code).\nYou might want to use another number of threads, e.g. while really huge matrices are calculated and one could profit from it or to use less threads, because there is no improvement. Set the environment variables either in the shell or prior to importing numexpr, e.g.\nimport os\nos.environ['NUMEXPR_MAX_THREADS'] = '4'\nos.environ['NUMEXPR_NUM_THREADS'] = '2'\nimport numexpr as ne \n\n" ]

[ 0 ]

[]

[]

[ "numexpr", "numpy", "python", "warnings" ]

[ "you need to print(tmp) instead of print(n)\nthis will print 3 2 1.\nTo get 3 3 3 2 2 1 you need to change your code more:\nn = int(input())\n\ntmp=n\n\nwhile tmp > 0:\n for _ in range(tmp)\n print(tmp)\n tmp -= 1\n\n\n", "You need to add if statement when tmp equals 0 then subtract n by 1 and assign tmp back to n:\nn = int(input('input : '))\ntmp = n\nwhile tmp > 0:\n print(n)\n tmp -= 1\n if tmp == 0:\n n -= 1\n tmp = n\n\nOutput:\n# input : 3\n# 3\n# 3\n# 3\n# 2\n# 2\n# 1\n\n" ]

[ 1, 0 ]

[]

[]

[ "python" ]

[ "There seem to be frequent complaints about lack of documentation, but there is some available on the Google OR-Tools site at https://developers.google.com/optimization .\nI learned a lot from the user manual from the old Google OR-Tools ConstraintSolver, to be found at https://www.scribd.com/document/482135694/user-manual-A4-v-0-5-2-pdf or Google \"or-tools user_manual_A4.v.0.5.2 Nikolaj van Omme Laurent Perron Vincent Furnon\". Even if you're using the new solver, a lot of concepts from that document still apply.\nThe Google team maintains a list of recommended literature at https://developers.google.com/optimization/support/resources\nTo do your two-level optimization:\nDefine a new variable bestFirstObjective and constrain it to be equal to the expression of your optimization value:\n# here I simply use 1000 as the horizon for whatever maximum value would be plausible\nfirstObjective = model.NewIntVar(0, 1000, \"Objective1\")\nmodel.Add(firstObjective == sum(shift_requests[n][d][s] * shifts[(n, d, s)] for n in all_employees\n for d in all_oncall_shifts for s in all_shifts))\nmodel.Maximize(bestFirstObjective)\n\nThen after solving, retrieve the value for firstObjective, and save it in a new Python integer bestFirstObjective (not an IntVar).\nNow in a new model2, add all the original variables and constraints, but instead of calling model.Maximize(bestFirstObjective) do this:\nmodel2.Add(firstObjective == bestFirstObjective)\n\nIn this new model, construct the variables and constraints you need for the subsidiary objective, and then add:\nmodel2.Maximize(secondObjective)\n\nand solve.\nAs Laurent Perron mentioned in Multiple objective functions with binary variables Google OR-tools, you can add the optimized solution to the original model as a hint during the second solution, that should speed things up. See https://developers.google.com/optimization/reference/python/sat/python/cp_model#addhint.\nI'm not sure if the second model is really needed, it might suffice to reuse the first one, it depends on whether calling Maximize a second time replaces the original objective or confuses the model, I simply don't know which is the case.\nIn your last comment it sounds like you need more help to construct a measure for \"evenly distributed\", something like trying to maximize the lowest interval between two consecutive shifts of any employee. The challenge here is to compute the length of this interval from the Booleans that describe whether a shift is worked or not.\nLet's assume that the number of days between successive shifts should be \"evenly distributed\". This would be the case if the shortest break between any two shifts for all the employees would be as large as possible.\nOne approach would be to first construct a Boolean which is true if an employee worked on day d (pseudo-code here, I haven't actually tested anything):\n# For each employee e and day d:\nworkedOnDay[e, d] = model2.NewBoolvar(\"Worked on day xxx\")\nmodel2.AddMaxEquality(workedOnDay[e, d], {shifts[(e, d, s)] for s in all_shifts})\n\nnew IntVar offDaysCounter for each employee/day. If the employee worked on the previous day (d-1) constrain its value to be 0. If the employee did not work on the previous day, constrain it to be the value of the previous day for that employee plus 1. This could be done something like this:\nOn the first day there is no \"previous day\" so constrain offDaysCounter to be an arbitrary large number without condition.\n# For each employee e \noffDaysCounter[e, 0] = model.NewIntVar(num_oncall_shifts, num_oncall_shifts, \"OffDaysCounter xxx\")\n\nOn the subsequent days:\n# For each employee e and day d (except the first day)\noffDaysCounter[e, d] = model.NewIntVar(0, 2 * num_oncall_shifts, \"OffDaysCounter xxx\")\nmodel2.Add(offDaysCounter[e, d] == 0).OnlyEnforceIf(workedOnDay[e, (d - 1)])\nmodel2.Add(offDaysCounter[e, d] == offDaysCounter[e, (d - 1)] + 1).OnlyEnforceIf(workedOnDay[e, (d - 1)].Not())\n\nThe variable will count up on the off days and be reset to 0 on the worked days; its value just before the next worked day is the length of the break.\nNow we can compute the actual number of off days between shifts. Create a new IntVar offDays for each employee/day. If the employee did not work on a day, set its value to be equal to offDays for the following day. If the employee worked on a day, set its value to be equal to offDaysCounter for the same day. This variable will now have a value on each day equal to the break between the last and next worked day relative to this day.\n# For each employee e and day d (except the last day):\noffDays[e, d] = model.NewIntVar(0, 2 * num_oncall_shifts, \"OffDays xxx\")\nmodel.Add(offDays[e, d] == offDays[e, (d + 1)]).OnlyEnforceIf(workedOnDays[e, d].Not())\nmodel.Add(offDays[e, d] == offDaysCounter[e, d]).OnlyEnforceIf(workedOnDays[e, d])\n\nOn the last day, there is no following day, so set the value of offDays to an arbitrary large number (like the total number of days), if the day was not worked, otherwise to the offDaysCounter[e, d]\nThe result could look something like this for a given employee:\n\n\n\n\nday\n0\n1\n2\n3\n4\n5\n6\n7\n8\n\n\n\n\nworkedOnDay\n1\n0\n0\n1\n0\n1\n0\n0\n0\n\n\noffDaysCounter\n18\n0\n1\n2\n0\n1\n0\n1\n2\n\n\noffDays\n18\n2\n2\n2\n1\n1\n18\n18\n18\n\n\n\n\nNow we can construct the objective: maximize the minimum interval between any employees shifts:\nsecondObjective = model.NewIntVar(0, 2 * num_oncall_shifts, \"second objective\")\nmodel.AddMinEquality(secondObjective, {offDays[e, d] for d in all_oncall_shifts, for e in all_employees})\nmodel.Maximize(secondObjective)\n\nAs I said, I haven't actually tested anything...\n" ]

[ 1 ]

[]

[]

[ "constraint_programming", "cp_sat_solver", "or_tools", "python", "python_3.x" ]

[ "update_state measures the metrics (mean, auc, accuracy), and stores them in the object, so it can later be retrieved with result:\nimport tensorflow as tf\n\nmean_object = tf.metrics.Mean()\n\nvalues = [1, 2, 3, 4, 5]\n\nfor ix, val in enumerate(values):\n mean_object.update_state(val)\n print(mean_object.result().numpy(), 'is the mean of', values[:ix+1])\n\n1.0 is the mean of [1]\n1.5 is the mean of [1, 2]\n2.0 is the mean of [1, 2, 3]\n2.5 is the mean of [1, 2, 3, 4]\n3.0 is the mean of [1, 2, 3, 4, 5]\n\nreset_states resets the metric to zero:\nmean_object.reset_states()\nmean_object.result().numpy()\n\n0.0\n\nI'm not sure I made it more clear than the documentation, it's pretty well explained in my opinion.\nCalling the object, e.g., mean_object([1, 2, 3, 4]) will update the metric, and return the result.\nimport tensorflow as tf\n\nmean_object = tf.metrics.Mean()\n\nvalues = [1, 2, 3, 4, 5]\n\nprint(mean_object.result())\nreturned_mean = mean_object(values)\nprint(mean_object.result())\nprint(returned_mean)\n\ntf.Tensor(0.0, shape=(), dtype=float32)\ntf.Tensor(3.0, shape=(), dtype=float32)\ntf.Tensor(3.0, shape=(), dtype=float32)\n\n", "I think by default the .result() method computes the mean of all the values that the .update_states was given.\nHere is an example:\nimport tensorflow as tf\n\nm = tf.keras.metrics.Accuracy()\nm.update_state([[1], [2], [3], [4]], [[0], [2], [3], [4]])\nresult = m.result().numpy()\nprint(f\"result: {result}\")\nm.update_state([[1], [2], [3], [4]], [[0], [0], [3], [4]])\nresult = m.result().numpy()\nprint(f\"result: {result}\")\n\nHere the accuracy of the first set of values is 0.75 and the accuracy of the second set of values is 0.5. So the last call to the method \"result\" will print 0.625. This is equal to (0.75+0.5)/2.\n" ]

[ 4, 0 ]

[]

[]

[ "keras", "metrics", "python", "tensorflow", "tensorflow2.0" ]

[ "If the text is not on the element and is on childrens of the element, change .text to .get_attribute(\"innerText\")\nOther solution could be accurate more on the element you want to extract, but sometimes you need this cause text is mixed on multiple elements\ntrainerColour1 = driver.find_element(By.XPATH,\"//span[@class='display-color-name color-name-mobile font-body regular pdp-update-event-triggerd']\").get_attribute(\"innerText\") \n\n" ]

[ 0 ]

[]

[]

[ "console", "pycharm", "python", "selenium", "selenium_webdriver" ]

[ "Yes, you can achieve this with Celery. Celery is a task queue that allows you to schedule tasks to be executed at a later time. It is designed to be used in distributed systems and can be used to manage the execution of callbacks in your Dash application.\nA small working example of using Celery with Dash would look something like this:\n# Import Celery and create a Celery instance\nfrom celery import Celery\ncelery_app = Celery('tasks', broker='redis://localhost:6379/0')\n\n# Define the callback function that will be added to the queue\n@celery_app.task(name='update_graph_live')\ndef update_graph_live(relayout, n):\n if ctx.triggered_id == 'relayout':\n * code that affects the y axis * \n return fig \n else:\n satellite = Orbital('TERRA')\n data = {\n 'time': [],\n 'Latitude': [],\n 'Longitude': [],\n 'Altitude': []\n }\n\n # Collect some data \n for i in range(180): \n time = datetime.datetime.now() - datetime.timedelta(seconds=i*20) \n lon, lat, alt = satellite.get_lonlatalt( time ) data['Longitude'].append(lon) data['Latitude'].append(lat) data['Altitude'].append(alt) data['time'].append(time)\n\n # Create the graph with subplots fig = plotly.tools.make_subplots(rows=2, cols=1, vertical_spacing=0.2) fig['layout']['margin'] = { 'l': 30, 'r': 10, 'b': 30, 't': 10 } fig['layout']['legend'] = {'x': 0, 'y': 1, 'xanchor': 'left'}\n\n fig.append_trace({ 'x': data['time'], 'y': data['Altitude'], 'name': 'Altitude', 'mode': 'lines+markers', 'type': 'scatter' }, 1, 1) fig.append_trace({ 'x': data['Longitude'], 'y': data['Latitude'], 'text': data['time'], 'name': 'Longitude vs Latitude', 'mode': 'lines+markers', 'type': 'scatter' }, 2, 1)\n\n return fig\n\n# Update the callback to add the task to the queue\[email protected](Output('live-update-graph', 'figure'),\n Input('live-update-graph', 'relayout'),\n Input('interval-component', 'n_intervals'))\ndef update_graph_live(relayout, n):\n # Add the task to the queue update_graph_live.delay(relayout, n) # Return an empty figure until the task is completed return {}\n\nif __name__ == '__main__':\n app.run_server(debug=True)\n\nIn this example, the callback is updated to add the task to the queue using Celery's delay() method. The callback then returns an empty figure until the task is completed.\n", "Yes, you can use Celery to create a job queue for your Plotly Dash app. Celery is a task queue and distributed job queue that can be used to run long-running tasks in the background, such as data processing or API calls.\nHere is a simple example of how you can use Celery to set up a job queue for your Dash app:\nFirst, install Celery and the required dependencies:\npip install celery\npip install eventlet\n\nNext, create a celery.py file that defines a Celery app and sets up the required configuration:\nfrom celery import Celery\n\napp = Celery(__name__)\napp.config_from_object('celeryconfig')\n\nThen, create a celeryconfig.py file that defines the broker and backend for the Celery app:\nBROKER_URL = 'amqp://localhost:5672'\nCELERY_RESULT_BACKEND = 'redis://localhost:6379'\n\nNext, modify your Dash app to use Celery to run the long-running tasks in the background. In your update_graph_live function, you can use the delay method to add a task to the Celery queue, and the get method to retrieve the result of the task when it is finished:\[email protected](Output('live-update-graph', 'figure'),\n Input('live-update-graph', 'relayout'),\n Input('interval-component', 'n_intervals'))\ndef update_graph_live(relayout, n):\n if ctx.triggered_id == 'relayout':\n # Add the \"code that affects the y axis\" task to the Celery queue\n result = code_that_affects_the_y_axis.delay()\n\n # Wait for the task to finish and retrieve the result\n result = result.get()\n return result\n else:\n satellite = Orbital('TERRA')\n data = {\n 'time': [],\n 'Latitude': [],\n 'Longitude': [],\n 'Altitude': []\n }\n\n # Collect some data\n for i in range(180):\n time = datetime.datetime.now() - datetime.timedelta(seconds=i*20)\n lon, lat, alt = satellite.get_lonlatalt(\n time\n )\n data['Longitude'].append(lon)\n data['Latitude'].append(lat)\n data['Altitude'].append(alt)\n data['time'].\n\n" ]

[ 0, 0 ]

[]

[]

[ "celery", "plotly", "plotly_dash", "plotly_python", "python" ]

[ "I found a solution that works, it is to target the element in XPATH from the javascript interface of selenium. You must then execute two commands on an instruction (targeting and scroll)\ndriver.executeScript(\"var el = document.evaluate('/html/body/jsl/div[3]/div[10]/div[8]/div/div[1]/div/div/div[4]/div[1]', document, null, XPathResult.FIRST_ORDERED_NODE_TYPE, null).singleNodeValue; el.scroll(0, 5000);\");\n\nthis is the only solution that worked for me\n", "The search results in the google map are located with //div[contains(@aria-label,'dentists in Austin Texas')]//div[contains(@jsaction,'mouseover')] XPath.\nSo, to select 6-th element there you can do the following\nfrom selenium.webdriver.common.action_chains import ActionChains\n\nresults = driver.find_elements_by_xpath('//div[contains(@aria-label,\"dentists in Austin Texas\")]//div[contains(@jsaction,\"mouseover\")]')\n\nActionChains(driver).move_to_element(results[6]).click(button).perform()\n\n", "I was just implementing scrolling on google map sidebar, it's working on my side. check this code please\n# selecting scroll body\ndriver.find_element_by_xpath('/html/body/div[3]/div[9]/div[9]/div/div/div[1]/div[2]/div/div[1]/div/div/div[2]/div[1]').click()\n\n\n#start scrolling your sidebar\nhtml = driver.find_element_by_xpath('/html/body/div[3]/div[9]/div[9]/div/div/div[1]/div[2]/div/div[1]/div/div/div[2]/div[1]')\nhtml.send_keys(Keys.END)\n\nalso add the \"KEYS\" library\nfrom selenium.webdriver.common.keys import Keys\n\nI hope it would help you.\nby the way I have implemented scrapping of google map with its available data and used above code to scroll. check if you have any problem, let me know then\n" ]

[ 1, 0, 0 ]

[]

[]

[ "html", "python", "scroll", "selenium", "selenium_webdriver" ]

[ "You should be able to use this formula from the points directly. So, you'd have something like:\nimport math\n\nclass Point:\n def distance_to_line(self, p1, p2):\n x_diff = p2.x - p1.x\n y_diff = p2.y - p1.y\n num = abs(y_diff*self.x - x_diff*self.y + p2.x*p1.y - p2.y*p1.x)\n den = math.sqrt(y_diff**2 + x_diff**2)\n return num / den\n\n", "The distance formula between two points is Distance =sqrt((x2−x1)^2+(y2−y1)^2).\nAnd the formula to calculate slope is slope = (y2 - y1) / (x2 - x1). \nso below is a simple method to calculate the distance\ndef distance_from_other_point(self, other_point):\n return math.sqrt( ( other_point.getX() - self.getX() )**2 + ( other_point.getY() - self.getY() )**2 )\n\ndef slope(self, otehr_point):\n return ( other_point.getY() - self.getY() )*1.0 / ( other_point.getX() - self.getX() )\n\nIn the second method, slope, I multiplied with 1.0 so that result will be in float.\nNote - I used the syntax of python 2.7.6 though hopefully, it will work in python 3.x as well.\n", "You can install FastLine via pip and then use it in this way:\nfrom FastLine import Line\n# define a line by two points\nl1 = Line(p1=(0,0), p2=(10,10))\n# or define a line by slope and intercept\nl2 = Line(m=0.5, b=-1)\n\n# compute distance\nd1 = l1.distance_to((20,50))\n# returns 21.213203435596427\nd2 = l2.distance_to((-15,17))\n# returns 22.807893370497855\n\n", "\nIt is\nQ-P0 = α (P1-P0) + β k×(P1-P0)\n\nwhere k×(P1-P0) is the vector multiplication between the z versor and the position vector.\nWriting the above equation as a scalar system, we have\nΔX = α dx - β dy\nΔY = α dy + β dx\n\nsolving for α, β\n(dx²+dy²) α = + ΔX dx + ΔY dy\n(dx²+dy²) β = - ΔX dy + ΔY dx\n\nBecause (dx²+dy²) = |P1-P0|²\n ΔY dx - ΔX dy\nβ |P1-P0| = distance = ---------------\n |P1-P0|\n\nOf course, our result is equivalent to (P1-P0)×(Q-P0)/|P1-P0|.\nFinal remark, the distance of Q from (P1-P0) is oriented, and maybe you need the absolute value of the distance.\n" ]

[ 6, 0, 0, 0 ]

[]

[]

[ "distance", "line", "point", "python", "python_3.x" ]

[ "For better numerical stability, I would suggest to use slogdet, which is your main aim in any case. There may also be a very minimal gain if you use np.inner(J, J) instead of J.dot(J.T). For really speeding things up, I would recommend using jax.numpy.\nimport numpy as np\nimport jax\nimport jax.numpy as jnp\n\nm, n = 100, 150\nJ = np.random.randn(m, n)\n\ndef a(J):\n return np.log(np.linalg.det(J.dot(J.T)))\n\ndef b(J):\n return np.linalg.slogdet(np.inner(J, J))[1]\n\ndef c(J):\n return jnp.linalg.slogdet(jnp.inner(J, J))[1]\n\n# jit + compile\nd = jax.jit(c)\nd(J)\n\n# check correctness\nprint(np.allclose(a(J), b(J))) # True\nprint(np.allclose(a(J), c(J))) # True\nprint(np.allclose(a(J), d(J))) # True\n\nChecking run times, on Google Colab:\n%timeit -n 1000 -r 10 a(J)\n# 240 µs ± 16.2 µs per loop (mean ± std. dev. of 10 runs, 1000 loops each)\n\n%timeit -n 1000 -r 10 b(J)\n# 227 µs ± 10.2 µs per loop (mean ± std. dev. of 10 runs, 1000 loops each)\n\nJ_dev = jax.device_put(J)\n\n%timeit -n 1000 -r 10 c(J_dev).block_until_ready()\n# 112 µs ± 4.46 µs per loop (mean ± std. dev. of 10 runs, 1000 loops each)\n\n%timeit -n 1000 -r 10 d(J_dev).block_until_ready()\n# 96.2 µs ± 4.23 µs per loop (mean ± std. dev. of 10 runs, 1000 loops each)\n\nSo rougly about ~2x speedup is possible this way.\n" ]

[ 2 ]

[]

[]

[ "determinants", "numpy", "python", "scipy" ]

[ "Try changing your code to something like this:\nimport asyncio\nimport telethon\n\napi_id = 13******\napi_hash = '8ba0a***********************'\nmessage_file = 'messages.txt'\n\n# Better style would be to have these not initiated\n# at the top-level and instead behind the if name == main\n# guard, but lets keep it simple.\nclient = telethon.TelegramClient('sender', api_id, api_hash)\nwith open(message_file, 'r') as f:\n messages = f.readlines()\n\nasync def main():\n for message in messages:\n await client.send_message('@Example_recipient', message.strip())\n await asyncio.sleep(60)\n\nif __name__ == \"__main__\":\n # The context manager will release the resource\n # (i.e. call client.disconnect) for us when the\n # block exits.\n with client:\n client.loop.run_until_complete(main())\n \n\nwhich is much closer to the example in the documentation. Also it's not clear why you have the 60 second wait, or why you are using asyncio if you want to wait between sending messages. It looks like you could just use telethon.sync and block and not have any of the asyncio boilerplate?\nUpdate based on comments\nwith open(file) as handle:\n messages = list(handle.readlines())\n\n...\n\n while len(messages):\n first = messages.pop()\n # check here before popping that the list\n # isn't already empty\n second = messages.pop()\n await client.send_message(first)\n await client.send_message(second)\n\n" ]

[ 0 ]

[]

[]

[ "async_await", "python", "telegram", "telethon" ]

[ "Let's see if I can help. It would appear that what you want to accomplish is to replace a subset/filtered portion of a column with values derived from other one or more other columns.\nFor example, if you are attempting to accomplish this:\nERA_COL = \"era\"\n\ntest_split = [\"01\", \"02\", \"10\"]\ntest_split_index = data[ERA_COL].isin(test_split)\n\ndata.loc[test_split_index, \"pred1\"] = -2 * data[\"pred3\"]\nprint(data)\n\n>>> print(data)\n era pred1 pred2 pred3 something_else\n0 01 -6 2 3 5\n1 01 -10 4 5 4\n2 02 -12 5 6 3\n3 02 -16 6 8 67\n4 03 5 7 9 5\n5 10 -2 8 1 4\n\nWe would accomplish the above in Polars using a when/then/otherwise expression:\n(\n pl.from_pandas(data)\n .with_column(\n pl.when(pl.col(ERA_COL).is_in(test_split))\n .then(-2 * pl.col('pred3'))\n .otherwise(pl.col('pred1'))\n .alias('pred1')\n )\n)\n\nshape: (6, 5)\n┌─────┬───────┬───────┬───────┬────────────────┐\n│ era ┆ pred1 ┆ pred2 ┆ pred3 ┆ something_else │\n│ --- ┆ --- ┆ --- ┆ --- ┆ --- │\n│ str ┆ i64 ┆ i64 ┆ i64 ┆ i64 │\n╞═════╪═══════╪═══════╪═══════╪════════════════╡\n│ 01 ┆ -6 ┆ 2 ┆ 3 ┆ 5 │\n├╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┤\n│ 01 ┆ -10 ┆ 4 ┆ 5 ┆ 4 │\n├╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┤\n│ 02 ┆ -12 ┆ 5 ┆ 6 ┆ 3 │\n├╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┤\n│ 02 ┆ -16 ┆ 6 ┆ 8 ┆ 67 │\n├╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┤\n│ 03 ┆ 5 ┆ 7 ┆ 9 ┆ 5 │\n├╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┤\n│ 10 ┆ -2 ┆ 8 ┆ 1 ┆ 4 │\n└─────┴───────┴───────┴───────┴────────────────┘\n\nIs this what you were looking to accomplish?\n", "A few things as general points.\n\npolars syntax doesn't attempt to match that of pandas.\n\nIn polars, you can only assign a whole df, you can't assign part of a df.\n\npolars doesn't use an internal index so there's no index to record\n\n\nFor your problem, assuming there isn't already a natural index, you'd want to make an explicit index.\npldata=pl.from_pandas(data).with_row_count(name='myindx)\n\nthen recording the index would be\ntest_split_index = pldata.filter(pl.col(ERA_COL).is_in(test_split)).select('myindx').to_series()\n\nFor your last bit on the final assignment, without knowing anything about somefunction my best guess is that you'd want to do that with a join.\nMaybe something like:\npldata=pldata.join(\n pl.from_pandas(some_function()['another_column']) \\\n .with_column(test_split_index.alias('myindex')),\n on='myindex')\n\nYour test_split_index is actually a bool not the index whereas the above is the actual index so take that with a grain of salt.\nAll that being said, polars has free copies of data so rather than keeping track of index positions manually (as error prone as that can be), you can just make 2 new dfs since, under the hood, it just references the data, it doesn't make a physical copy of it.\nSomething like:\ntestdata=pldata.filter(pl.col(ERA_COL).is_in(test_split))\n\ntraindata=pldata.filter(~pl.col(ERA_COL).is_in(test_split))\n\n" ]

[ 0, 0 ]

[]

[]

[ "python", "python_polars" ]

[ "For each multi-key you need to build the tree, add a {} for each one except the last, then use the last one to assign the value\nvalue = {'a': 1, 'd': [1, '#d_i1', 3], 'c.a': '#a_val', 'c.b.x': '#x_value', 'c.b.y': '#y'}\n\nresult = {}\nfor k, v in value.items():\n tmp = result\n *keys, last = k.split(\".\")\n for key in keys:\n tmp = tmp.setdefault(key, {})\n tmp[last] = v\n\nprint(result)\n# {'a': 1, 'd': [1, '#d_i1', 3], 'c': {'a': '#a_val', 'b': {'x': '#x_value', 'y': '#y'}}}\n\n" ]

[ 1 ]

[]

[]

[ "python" ]

[ "Welcome to Stackoverflow!\nYou were getting there, the following code will yield what you want (I simply added some fake data + 1 line of code to your plotting code):\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport matplotlib.ticker as ticker\n\ny = [10**x for x in np.arange(0, 5, 0.1)]\nx = [x for x in np.linspace(2018, 2023, len(y))]\n\n#Plot the Data \nplt.style.use('fivethirtyeight') \nplt.figure(figsize =(20, 10))\nplt.title(\"Bitcoin Price\", fontsize=18)\nplt.yscale(\"log\")\nplt.plot(x, y)\nplt.xlabel(\"Date\", fontsize=15)\nplt.ylabel(\"Price\", fontsize=15)\nplt.gca().get_yaxis().set_major_formatter(ticker.ScalarFormatter())\nplt.show()\n\nThis generates the following figure:\n\nThe fundamental lines are these:\nimport matplotlib.ticker as ticker\nplt.gca().get_yaxis().set_major_formatter(ticker.ScalarFormatter())\n\nExplanation: plt.gca() gets the currently active axis object. This object is the one we want to adapt. And the actual thing we want to adapt is the way our ticks get formatted for our y axis. Hence the latter part: .get_yaxis().set_major_formatter(). Now, we only need to choose which formatter. I chose ScalarFormatter, which is the default for scalars. More info on your choices can be found here.\nHope this helps!\n" ]

[ 0 ]

[]

[]

[ "bitcoin", "price", "python", "scale" ]

[ "You need a merge_asof:\nout = pd.merge_asof(Wins, Levels, left_on='Wins Needed', right_on='Wins',\n direction='forward')[['Wins Needed', 'Level']]\n\nOr\nWins['Level'] = pd.merge_asof(Wins, Levels, left_on='Wins Needed', right_on='Wins',\n direction='forward')['Level']\n\nNB. the keys must be sorted for a merge_asof.\nOutput:\n Wins Needed Level\n0 1 1\n1 2 1\n2 3 1\n3 4 2\n4 5 2\n.. ... ...\n95 96 35\n96 97 35\n97 98 36\n98 99 36\n99 100 37\n\n[100 rows x 2 columns]\n\nIf the values are not initially sorted:\nWins['Level'] = pd.merge_asof(Wins[['Wins Needed']].reset_index().sort_values(by='Wins Needed'),\n Levels.sort_values(by='Wins'),\n left_on='Wins Needed', right_on='Wins',\n direction='forward').set_index('index')['Level']\n\n" ]

[ 1 ]

[]

[]

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

[ "Solution\nThe problem was in effect with the grad function or rather with the ddx and ddy functions. These two functions were actually the same, except I was swapping x and y.\nIn other words, I wasn't computing the partial derivatives correctly which caused my gradient to be wrong and my algorithm to not work properly.\nHere is the part of the code that has been modified, the rest is unchanged:\ndef f_replace_y(f, y):\n return lambda x: f(x, y)\n\ndef f_replace_x(f, x):\n return lambda y: f(x, y)\n\n# Dérivée partielle de f par rapport à x\ndef ddx(f):\n return lambda x, y : derivative(f_replace_y(f, y), x, dx=precision, n=1)\n\n# Dériavée partielle de f par rapport à y\ndef ddy(f):\n return lambda x, y : derivative(f_replace_x(f, x), y, dx=precision, n=1)\n\nHowever, I still wonder how my previous code could pass through the lowest point given that my partial derivatives were wrong.\n" ]

[ 0 ]

[]

[]

[ "algorithm", "gradient", "gradient_descent", "math", "python" ]

[ "It depends what you are comparing.\n\nif you compare the two methods \"with preprocessing allowed\", then you don't include the preprocessing in the experiment; and in principle you should test several (identical) queries;\n\nif you compare \"with no preprocessing allowed\", then include everything in the measurement.\n\n\n" ]

[ 1 ]

[]

[]

[ "algorithm", "comparison", "machine_learning", "python", "theory" ]

[ "I am not sure if this is the correct approach to handle this, but I had a similar issue when trying to test a deeply nested function that was using Depends and thus was not willing to use the override approach.\nMy \"mock\" was pretty simple, instead of trying to mock fastapi.Depends I just passed in my desired return value to replace the default parameter value.\nSo in your case something like:\ndef test_method_call():\n result = demo(user_id=\"me\")\n assert result[\"user_id\"] == \"me\"\n\n" ]

[ 0 ]

[]

[]

[ "fastapi", "mocking", "pytest", "python" ]

[ "You can select values in c1 and c2 without first letters and aggregate sum:\ndf = df.groupby([df.c1.str[1:], df.c2.str[1:]]).sum().reset_index()\nprint (df)\n c1 c2 c3\n0 x y 4\n1 y x 6\n\n" ]

[ 1 ]

[]

[]

[ "pandas", "python" ]

[ "My guess is that you want:\nrule all:\n input:\n expand('{reference_genome}.{type}', reference_genome=['fa1', 'fa2', 'fa3'], type=['amb', 'ann', 'pac'])\n\nrule index: \n input: \n #reference_genomeFile\n ref_genome=lambda wildcards:config['reference_genome'][wildcards.reference_genome]\n output: \n expand('{{reference_genome}}.{type}', type=['amb', 'ann', 'pac'])\n log: \n 'log/rule_index_{reference_genome}.log'\n shell: \n \"bwa index -a bwtsw {input.ref_genome} > {log} 2>&1\"\n\nsnakemake -p -n -j 1 --configfile config.yaml\n\nThat is: Run rule index for each genome file, i.e., three times here. Each run of index generates the three index files. Note the use of double curly braces {{reference_genome}} to tell expand that this wildcard does not need to be expanded.\n\nExample config.yaml:\nreference_genome:\n fa1: \"genome_1.fa\"\n fa2: \"genome_2.fa\"\n fa3: \"genome_3.fa\"\n\n", "Building upon dariober's answer and judging from your comments, I think this is what you are looking for?\nconfigfile: \"config.yaml\"\n\n\nrule all:\n input:\n expand(\n \"{reference_genome}.{type}\",\n reference_genome=list(config[\"reference_genome\"].values()),\n type=[\"amb\", \"ann\", \"pac\"],\n ),\n\n\nrule index:\n input:\n #reference_genomeFile\n ref_genome=\"{reference_genome}\",\n output:\n expand(\"{{reference_genome}}.{type}\", type=[\"amb\", \"ann\", \"pac\"]),\n log:\n \"log/rule_index_{reference_genome}.log\",\n shell:\n \"bwa index -a bwtsw {input.ref_genome} > {log} 2>&1\"\n\nwith config.yaml\nreference_genome:\n fa1: \"path/to/genome1\"\n fa2: \"path/to/genome2\"\n fa3: \"path/to/genome3\"\n\nI've modified the rule all to use the filepaths from config.yaml rather than the list ['fa1','fa2','fa3']. I've also removed the lambda wildcard from the input of rule index as it seems unnecessary.\n" ]

[ 1, 1 ]

[]

[]

[ "bioinformatics", "pipeline", "python", "snakemake", "wildcard" ]

[ "You can split the string, then parse each JSON string into a dictionary:\nimport json\n\ncontents = r'{\"payload\":{\"device\":{\"serial\":213}}}{\"payload\":{\"device\":{\"serial\":123}}}'\n\njson_strings = contents.replace('}{', '}|{').split('|')\njson_dicts = [json.loads(string) for string in json_strings]\n\nOutput:\n[{'payload': {'device': {'serial': 213}}}, {'payload': {'device': {'serial': 123}}}]\n\n" ]

[ 1 ]

[]

[]

[ "amazon_web_services", "arrays", "json", "ndjson", "python" ]

[ "First of all, your shown data has no tabs, so it's not clear why you've shown code that splits lines on tabs and finds the max. Not an average.\nTo find an average, you'll need to collect all seen values into a list (values.append(int(val))), then you can from statistics import mean and call mean(values) at the end of the loop\nI'd highly suggest that you use mrjob or pyspark instead\n" ]

[ 0 ]

[]

[]

[ "hadoop", "hadoop_streaming", "mapreduce", "python" ]

[ "def print_matrix(matrix):\n for i in range(len(matrix)):\n if i % 6 == 0 and i != 0:\n print('')\n print(matrix[i], end = ' ')\n print('')\n\nprint_matrix(matrix)\n\n" ]

[ 0 ]

[]

[]

[ "pygame", "python" ]

[ "Have you tried the python lib version of the GSPBOX?\nIf yes, why it does not work for you?\nhttps://pygsp.readthedocs.io/en/stable/reference/graphs.html\n", "To display a 3D Sphere graph structure using VTK, you can use the vtkSphereSource class to generate the sphere geometry, and the vtkGraphLayoutView class to visualize the graph.\nHere is an example of how you can create a 3D Sphere graph structure in VTK:\nFirst, import the required modules:\nsphere = vtk.vtkSphereSource()\nsphere.SetRadius(1.0)\nsphere.SetThetaResolution(32)\nsphere.SetPhiResolution(32)\n\nNext, create a vtkSphereSource object and set the radius and resolution of the sphere:\nsphere = vtk.vtkSphereSource()\nsphere.SetRadius(1.0)\nsphere.SetThetaResolution(32)\nsphere.SetPhiResolution(32)\n\nThen, create a vtkPolyData object and set the points and polygons of the sphere using the output of the vtkSphereSource object:\n polydata = vtk.vtkPolyData()\npolydata.SetPoints(sphere.GetOutput().GetPoints())\npolydata.SetPolys(sphere.GetOutput().GetPolys())\n\nNext, create a vtkPoints object and set the coordinates of the nodes in the graph:\n\npoints = vtk.vtkPoints()\npoints.InsertNextPoint(1.0, 0.0, 0.0)\npoints.InsertNextPoint(0.0, 1.0, 0.0)\npoints.InsertNextPoint(0.0, 0.0, 1.0)\n\nThen, create a vtkCellArray object and add the edges of the graph to the vtkCellArray object:\ncells = vtk.vtkCellArray()\ncells.InsertNextCell(2)\ncells.InsertCellPoint(0)\ncells.InsertCellPoint(1)\ncells.InsertNextCell(2)\ncells.InsertCellPoint(0)\ncells.InsertCellPoint(2)\n\nNext, create a vtkPolyData object and set the points and edges of the graph using the vtkPoints and vtkCellArray objects:\ngraph = vtk.vtkPolyData()\n\ngraph.SetPoints(points)\ngraph.SetLines(cells)\nThen, create a vtkGraphLayoutView object and add the vtkPolyData objects for the sphere and the graph to the vtkGraphLayoutView object:\n view = vtk.vtkGraphLayoutView()\nview.AddRepresentationFromInput(polydata)\nview.\n\n" ]

[ 0, 0 ]

[]

[]

[ "3d", "graph", "python" ]

[ "Two possibilities from the top of my head:\n\n[Linux] Use cron job https://help.ubuntu.com/community/CronHowto\n[Any OS] Use scheduler https://schedule.readthedocs.io/en/stable/\n[Windows] Scheduling a .py file on Task Scheduler in Windows 10\n\n" ]

[ 0 ]

[]

[]

[ "python" ]

[ "You run your function in the wrong place.\nTo run your function for each iterator, run the function there.\n#reading the file\n\nfor line in file:\n line = line.strip()\n if \">\" in line:\n header = line\n elif (len(line) == 0):\n continue\n else:\n seq = line\n #run function for each line, each time.\n rc = reverse_complement(seq)\n print(rc)\n\nIn your previous code, all iteration is successful. But you didn't put the line to the function to run each time. In your previous code, after all, iterations, only the last line is assigned. Therefore you put the last line to the function at the end. This is why your code prints only one line.\nThe solution.\nfrom itertools import repeat\n\n#reverse complement\n\ndef reverse_complement(seq):\n compline = ''\n for n in seq:\n if n == 'A':\n compline += 'T'\n elif n == 'T':\n compline += 'A'\n elif n == 'C':\n compline += 'G'\n elif n == 'G':\n compline += 'C'\n return((compline)[::-1])\n\n\n#opening file\n\nfilename = \"MySequences.fasta\"\nfile = open(filename, 'r')\n\n\n#reading the file\n\nfor line in file:\n line = line.strip()\n if \">\" in line:\n header = line\n elif (len(line) == 0):\n continue\n else:\n seq = line\n #run each line\n rc = reverse_complement(seq)\n print(rc) \n\nAlso, this is your other mistake.\nYou put seq as input instead line.\nBut even if you fix this, this code won't work for the same reason I told you before.\nfor line in file:\n rc = reverse_complement(seq)\n print(rc) \n\n" ]

[ 0 ]

[]

[]

[ "bioinformatics", "python" ]

[ "So I was able to fix my problem by first converting my matrix into a stack.\ncooc_matrix = matrix(matrixLabel, texts)\nmatrix = pd.DataFrame(cooc_matrix.todense(), index=matrixLabel, columns=matrixLabel)\nprint(matrix)\n\n#This fixed my problem\nstw = matrix.stack()\nstw = stw[stw >= 1].rename_axis(('source', 'target')).reset_index(name='weight')\nprint(stw)\n\nG = nx.from_pandas_edgelist(stw, edge_attr=True)\n\n", "I got the same problem. I am using a square pandas data frame (as indicated by df.shape) but I am getting the error Adjacency matrix not square. Using stack() and from_pandas_edgelist() solved the problem for me.\n" ]

[ 2, 0 ]

[]

[]

[ "matrix", "networkx", "python" ]

[ "There are several tutorials on the internet. Here's what I found...\n\n\nFirst you need to check for the correct formatting and for this you can use regular expressions like this:\n import re\n \n addressToVerify ='[email protected]'\n match = re.match('^[_a-z0-9-]+(\\.[_a-z0-9-]+)*@[a-z0-9-]+(\\.[a-z0-9-]+)*(\\.[a-z]{2,4})$', addressToVerify)\n \n if match == None:\n print('Bad Syntax')\n raise ValueError('Bad Syntax')\n\nDNS\nNext we need to get the MX record for the target domain, in order to start the email verification process. Note that you are allowed in the RFCs to have a mail server on your A record, but that's outside of the scope of this article and demo script.\n import dns.resolver\n \n records = dns.resolver.query('scottbrady91.com', 'MX')\n mxRecord = records[0].exchange\n mxRecord = str(mxRecord)\n\nPython DNS Python doesn't have any inbuilt DNS components, so we've\npulled in the popular dnspython library. Any library that can resolve\nan MX record from a domain name will work though.\nMailbox\nNow that we have all the preflight information we need, we can now find out if the email address exists.\n import socket\n import smtplib\n \n # Get local server hostname\n host = socket.gethostname()\n \n # SMTP lib setup (use debug level for full output)\n server = smtplib.SMTP()\n server.set_debuglevel(0)\n \n # SMTP Conversation\n server.connect(mxRecord)\n server.helo(host)\n server.mail('[email protected]')\n code, message = server.rcpt(str(addressToVerify))\n server.quit()\n \n # Assume 250 as Success\n if code == 250:\n print('Success')\n else:\n print('Bad')\n\nWhat we are doing here is the first three commands of an SMTP conversation for sending an email, stopping just before we send any data.\nThe actual SMTP commands issued are: HELO, MAIL FROM and RCPT TO. It is the response to RCPT TO that we are interested in. If the server sends back a 250, then that means we are good to send an email (the email address exists), otherwise the server will return a different status code (usually a 550), meaning the email address does not exist on that server.\nAnd that's email verification!\n\nsource: https://www.scottbrady91.com/Email-Verification/Python-Email-Verification-Script\n\nHere are some other alternatives from another website...\n\nLet’s make it more sophisticated and assume we want the following criteria to be met for an account@domain email address:\n\nBoth consist of case-insensitive alphanumeric characters. Dashes,\nperiods, hyphens or underscores are also allowed\nBoth can only start and end with alphanumeric characters\nBoth contain no white spaces account has at least one character and\ndomain has at least two domain includes at least one period\nAnd of course, there’s an ‘@’ symbol between them\n\n ^[a-z]([w-]*[a-z]|[w-.]*[a-z]{2,}|[a-z])*@[a-z]([w-]*[a-z]|[w-.]*[a-z]{2,}|[a-z]){4,}?.[a-z]{2,}$\n\nValidating emails with Python libraries\nAnother way of running sophisticated checks is with ready-to-use packages, and there are plenty to choose from. Here are several popular options:\nemail-validator 1.0.5\nThis library focuses strictly on the domain part of an email address, and checks if it’s in an [email protected] format.\nAs a bonus, the library also comes with a validation tool. It checks if the domain name can be resolved, thus giving you a good idea about its validity.\npylsEmail 1.3.2\nThis Python program to validate email addresses can also be used to validate both a domain and an email address with just one call. If a given address can’t be validated, the script will inform you of the most likely reasons.\npy3-validate-email\nThis comprehensive library checks an email address for a proper structure, but it doesn’t just stop there.\nIt also verifies if the domain’s MX records exist (as in, is able to send/receive emails), whether a particular address on this domain exists, and also if it’s not blacklisted. This way, you avoid emailing blacklisted accounts, which would affect deliverability.\n\nsource: https://mailtrap.io/blog/python-validate-email/\nThese are just extracts from these websites, make sure to visit them and confirm this can be useful for your specific case, they also include much more methods and explain this in more detail, hope it helps.\n", "You can send a verification email with a code to the entered email to verify its presence. This can be done using any python framework. Django, flask, etc.\n", "You can use gmass.co, connect to your google account , go to settings , apikeys , then create an api key and use it this way :\nsent = requests.get('https://verify.gmass.co/verify?email='+email+'&key='+api,\n headers={'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/87.0.4280.141 Safari/537.36'}\n )\nif (\"valid\" in sent.content):\n print(email+\"valid\")\nelse: \n print(email+\"invalid\")\n\n", "There are many services offering such tools that can validate your emails in bulk. If you are doing this for a marketing campaign to avoid bounce rate and getting your campaign marked as spammed then there are services that can help you checking bulk emails for your marketing campaigns.\nHowever, there is some online platform too but you can't check multiple emails.\n\nValidating email addresses is easy using valid email checkers.\nYou can use SMTPlib, a python library used for sending emails and using python regex too. But It will only check for the email if it is an email or something else.\n\nimport re\n\nregex = r'\\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Z|a-z]{2,}\\b'\ndef check(email):\n\n if(re.fullmatch(regex, email)):\n print(\"Valid Email\")\n \n else:\n print(\"Invalid Email\")\n\nif __name__ == '__main__':\n email = input(\"Enter Email:\")\n check(email)\n\nThis is not the actual solution to your question, but an example of matching email with regex.\nCheck out the services that offer emails validation in bulk.\nI'd recommend EmailChecks\n", "The best method would be to send a verification mail, with some code they need to submit. Not only this is simple, it also proves the email belongs to them\n", "I believe there are three ways to approach this.\n1. REGEX\nOther thread participants have already suggested this. The problem with this approach is that it simply checks a string for the correct syntax and does not ensure that the email address exists.\n2. Bulk Software\nThere are plenty of GUI tools that allow you to import lists that are then being validated. However, this usually does not fit the software development case.\n3. Email Validation API\nThe previously mentioned GUI tools often offer an API that allows you to conduct real-time validation of your user's email addresses.\nA few vendors that provide this:\n\nemailvalidation.io\nemailable.com\ndatavalidation.com\n\n" ]

[ 2, 1, 1, 1, 0, 0 ]

[]

[]

[ "python" ]

[ "Maybe the issue is that the emp_list is very large, and it takes a long time to insert it in the template, see the page won't be shown. \nYou can limit the data to for example 10 documents, using:\nemp_list = mongo.db.employee_entry.find().limit(10)\n\nand see if it solves the problem.\n", "OK, I'm sorry for that msg earlier\ntry this:\n{% for key,value in emp_list %}\n<tr>\n <th scope=\"row\">{{loop.index}}</th>\n <td>{{value}}</td>\n <td>{{value}}</td>\n <td>{{value}}</td>\n</tr>\n{% endfor %}\n\nwhere loop.index is used for incrementing you can search it's use\nI am 100% sure that just pasting it without editing it will not work.\nBut what i am actually trying to say is try using key,value function of for loops in python(jinja2)\ntry edit this so that it can fit in your code\n" ]

[ 0, 0 ]

[]

[]

[ "flask", "mongodb", "python" ]

[ "In your view, you cannot use {% url '' %}.\nTo resolve a URL dynamically in your utils.py, use build_absolute_uri instead. You can also combine this with reverse() like so (note: you will have to pass your request object):\nrequest.build_absolute_uri(reverse('columndetails', args=('2022-007', )))\n\n" ]

[ 1 ]

[]

[]

[ "django", "django_urls", "html", "python", "url" ]

[ "I don't know if I understand exactly what you need, but I think after the modifications you should use the to_csv() function to export the changes to the CSV file, and connect it for example with a Button click.\nIn case you can not find the file after saving, you should note that the to_csv() function usually saves the files in the root directory.\n# in case you have a save Button in your GUI\nself.save_Button.clicked.connect(self.SaveChanges)\n\ndef SaveChanges(self):\n df_subj.to_csv(\"file_ame.csv\", index=False)\n\n" ]

[ 0 ]

[]

[]

[ "csv", "pyqt5", "python" ]

[ "The problem with Python and parallel computing is that there is that thing called GIL (Global Interpreter Lock). The GIL prevents a process to run multiple threads at the same time. So for that to work you would need to spawn a new process which has quiet some overhead. Furthermore it is cumbersome to exchange data between the processes.\nThere is that multiprocessing package which helps with creating new processes: https://docs.python.org/3/library/multiprocessing.html\nStill in your case the question is, how do you split up the work. You could split up the elements in subblocks and create a process for each subblock which gets calculated there. The messaging between the processes can be implemented using pipes, as also described in 1. So you need to send a lot of objects to the new processes and then send them back via the pipes. I am not sure if the overhead of all that will make it faster in the end or even slower.\n" ]

[ 0 ]

[]

[]

[ "background_process", "parallel_processing", "python", "python_multithreading" ]

[ "It would be very helpful if you could attach some sample records from the input .csv file (anonymized if possible).\nWithout that, have you tried shuffling the original list instead of using the permutations?\nimport glob\nimport random\nimport csv\n\nall_list = []\nfor glob in glob.glob(\"random_friend/csv/*\"):\n file1 = open(glob, \"r+\")\n reader = csv.reader(file1, delimiter=',')\n for i in reader:\n all_list.append(i)\n all_list.pop(0)\n\nrandom.shuffle(all_list)\n\ngift = random.choice(['chocolat', 'Squeeze', 'fridge magnet', 'popcorn door cushion kit', 'cocktail shaker kit', 'Suspense book'])\n\nfor i in range(0, len(all_list), 2):\n name_one = all_list[i][1]\n name_two = all_list[i+1][1]\n mail_one = all_list[i][2]\n mail_two = all_list[i+1][2]\n\n print(f\"\"\"{name_one} took {name_two} and present with a {gift} and send it by e-mail to {mail_two}\"\"\")\n\n" ]

[ 1 ]

[]

[]

[ "python", "python_itertools", "random" ]

[ "With the help of Peter i managed to come up with this final test:\ndef mock_mutate_my_dict(my_dict):\n my_dict[\"new_key\"] = \"new_value\"\n return True\n\n\ndef test_function_under_test():\n with patch(\"stack_over_flow.mutate_my_dict\") as mutate_my_dict_mock:\n mutate_my_dict_mock.side_effect = mock_mutate_my_dict\n result = function_under_test()\n assert result == \"new_value\"\n\nHow it works is that with a side effect you can run a function instead of the intended function.\nIn this function you need to both change all of the mutating arguments and return the value returned.\n" ]

[ 0 ]

[]

[]

[ "pytest", "python", "python_3.x", "python_unittest.mock", "unit_testing" ]

[ "Removed the for loop and rewrote the code as below:\[email protected]('/')\ndef Results():\n try:\n Project_List_Col = db.ppm_master_db_collection.find()\n return render_template('Results.html',tasks=Project_List_Col)\n except Exception as e:\n return dumps({'error': str(e)})\n\nif __name__ == '__main__': \n app.run(debug = True)\n\nDocuments are displayed on the HTML Page as is.\n(***Will work on the formatting part. Meanwhile any pointers are greatly appreciated.)\n", "Try using key,value function of for loop and also remove that loop in the python app file from route like upper answer @Dinakar suggested\n" ]

[ 0, 0 ]

[]

[]

[ "flask", "mongodb", "python" ]

[ "This is the line interpreter is complaining about, you cannot pass two directories to os.listdir function\nfiles = os.listdir(source1, source2)\n\nYou have to have a nested loop (or list comprehension) to do what you want so:\nimport os\nsources = [source1, source2, ..., sourceN]\nfiles_to_move = []\nfor source in sources:\n current_source_files =[f\"{source}{filename}\" for filename in os.listdir(source)]\n files_to_move.extend(current_source_files)\nfor f in files_to_move:\n shutil.move(f, f\"{destination}{f.split(os.sep)[-1]}\")\n\nFor \"cleaner\" solution it's worth to look at:\nhttps://docs.python.org/3/library/os.path.html#module-os.path\n" ]

[ 1 ]

[]

[]

[ "directory", "file", "python", "shutil" ]

[ "This code should work:\nif True:(\nx:='hello_x',\nprint('hello'),\nprint(x)\n)\n\n## hello\n## hello_x\n\nIn your case, you are using a tuple to break python's indentation logic, so you need to separate each element with a comma. And since you are in a tuple, you need to use the Walrus Operator := to assign a value.\n" ]

[ 3 ]

[]

[]

[ "python", "python_3.x" ]

[ "I'm pretty sure you need to install once the extensions are built (otherwise where are the built extension and how python/pytest should know where to look?). This is how my workflow looked some time ago (not sure it still applies but should be close enough):\npython setup.py build_ext --inplace -j 4\npython -m pip install -e . --no-build-isolation --no-use-pep517\n\n...\n\npytest pandas/tests/xxxx/yyyy.py\n\nInstalling in development mode (-e) is the most convenient option in my opinion for development.\n" ]

[ 1 ]

[]

[]

[ "cython", "pandas", "pytest", "python", "visual_studio_code" ]

[ "We can use list to split every character of a string, also we can use \"end\" in \"print\" to indicate the deliminter in the print out.\ndef sq(num):\n words = list(str(num)) # split the text\n for word in words: # for each word in the line:\n print(int(word)**2, end=\"\") # print the word\n\nnum = 9119\nsq(num)\n\nAlternatively\nreturn ''.join(str(int(i)**2) for i in str(num))\n\n", "def sq(num):\n z = ''.join(str(int(i)**2) for i in str(num))\n return int(z)\n\n", "number=str(input(\"Enter the number :\"))\n\ndef pc(number):\n\n digits = list(number)\n\n for j in digits:\n\n print(int(j)**2,end=\"\")\n \npc(number)\n\n", "We can also support input of negative numbers and zeros. Uses arithmetic operators (% and //) for fun.\ndef sq(num):\n num = abs(num) #Handle negative numbers\n output = str((num % 10)**2) #Process rightmost digit \n\n while(num > 0): \n num //= 10 #Remove rightmost digit \n output = str((num % 10)**2) + output #Add squared digit to output\n print(output)\n\n", "Also you can try this variant:\ndef square_digits(num):\nreturn int(''.join(str(int(i)**2) for i in str(num)))\n" ]

[ 3, 2, 1, 0, 0 ]

[ "def square_digits(num):\n num = str(num)\n result = ''\n for i in num:\n result += str(int(i)**2)\n return int(result)\nvar = square_digits(123)\nprint(var)\n\n" ]

[ -1 ]

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

[ "There is some issues with your code as @azro pointed in the comments and there is no need for a loop here. You can simply use numpy.select to create a multi-conditionnal column.\nHere is an example to give you the general logic :\ndf[\"Attack\"] = df[\"Attack\"].astype(int)\n \nconditions = [df[\"Type 1\"].eq(\"Grass\"), df[\"Type 1\"].isin([\"Fire\", \"Rock\"])]\nchoices = [df[\"Attack\"].div(2), df[\"Attack\"].mul(2)]\n \ndf[\"grass_attack\"] = np.select(conditions, choices, default=df[\"Attack\"]).astype(int)\n\n# Output :\nprint(df)\n\n Type 1 Attack grass_attack\n0 Grass 62 31\n1 Grass 82 41\n2 Dragon 100 100\n3 Fire 52 104\n4 Rock 100 200\n\n", "You could use apply to do the necessary calculations. In the following code, the modify_Attack() function is used to calculate the Grass Attack values based on the Type1 and Attack values.\n\nType 1 values that are in the bad list will have their attack values halved.\nType 1 values that are in the good list will have their attack values doubled.\nAll other attack values will remain unchanged.\n\nHere is the code:\nimport pandas as pd\n\n# Create dataframe\ndf = pd.DataFrame({ 'Type 1': ['Grass', 'Grass', 'Dragon', 'Fire', 'Rock'],\n 'Attack': [62, 82, 100, 52, 100]})\n\n\n# Function to modify the Attack value based on the Type 1 value\ndef modify_Attack(type_val, attack_val):\n bad = ['Fire', 'Flying', 'Poison', 'Bug', 'Steel', 'Grass', 'Dragon']\n good = ['Ground','Water']\n \n result = attack_val # default value is unchanged\n \n if type_val in bad:\n result /= 2\n\n elif type_val in good:\n result *= 2\n \n return result\n \n\n# Create the Grass Attack column\ndf['Grass Attack'] = df.apply(lambda x: modify_Attack(x['Type 1'], x['Attack']), axis=1).astype(int)\n\n# print the dataframe\nprint(df)\n\nOUTPUT:\n Type 1 Attack Grass Attack\n0 Grass 62 31\n1 Grass 82 41\n2 Dragon 100 50\n3 Fire 52 26\n4 Rock 100 100\n\n" ]

[ 1, 1 ]

[]

[]

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

[ "To my knowledge YahooFinance uses CoinMarketCap to retrive crypto market information.\nCoinMarketCap offers the API you request here: (not free)\nhttps://pro-api.coinmarketcap.com/v1/exchange/market-pairs/latest\nI suggest you transfer to the Binance API. It includes the endpoint GET /api/v1/exchangeInfo as documented here:\nhttps://github.com/binance/binance-spot-api-docs/blob/master/rest-api.md\nYour direct endpoint would be https://api.binance.com/api/v1/exchangeInfo\n", "You could use https://python-binance.readthedocs.io/en/latest/general.html#id4 :\nfrom binance import Client \n\ninfo = client.get_exchange_info()\nsymbols = [x['symbol'] for x in info['symbols']]\n\n", "\"BTC-USD\"works pretty well for me\n", "you can use python-binance library\nfrom binance import Client\nfrom tqdm.autonotebook import tqdm\nimport pandas as pd\nimport numpy as np\n\ndef get_binance_data(ticker, interval='4h', start='1 Jan 2018', end=None):\n client = Client()\n intervals = {\n '15m': Client.KLINE_INTERVAL_15MINUTE,\n '1h': Client.KLINE_INTERVAL_1HOUR, \n '4h': Client.KLINE_INTERVAL_4HOUR,\n '1d': Client.KLINE_INTERVAL_1DAY\n }\n interval = intervals.get(interval, '4h')\n# print(f'Historical interval {interval}')\n klines = client.get_historical_klines(symbol=ticker, interval=interval, start_str=start, end_str=end)\n data = pd.DataFrame(klines)\n data.columns = ['open_time','open', 'high', 'low', 'close', 'volume','close_time', 'qav','num_trades','taker_base_vol','taker_quote_vol', 'ignore']\n data.index = [pd.to_datetime(x, unit='ms').strftime('%Y-%m-%d %H:%M:%S') for x in data.open_time]\n usecols=['open', 'high', 'low', 'close', 'volume', 'qav','num_trades','taker_base_vol','taker_quote_vol']\n data = data[usecols]\n data = data.astype('float')\n return data\n\n\nclient = Client()\nexchange_info = client.get_exchange_info()\nsymbols=[s['symbol'] for s in exchange_info['symbols'] if s['status'] == 'TRADING']\nticker_list = symbols[:50]\n# tiker_list = np.random.choice(symbols, size=50)\nprint('Number of crypto pairs: ', len(symbols))\nprint('First 50 pairs: ', *ticker_list)\n\n# collect pair closes in one dataframe\ncoins = []\nfor ticker in tqdm(ticker_list):\n try:\n close_price = get_binance_data(ticker, interval='1d', start='1 Jan 2018', end='1 Jul 2022')['close'].to_dict()\n info = {'name': ticker}\n info.update(close_price)\n coins.append(info)\n except Exception as err:\n print(err)\n continue\n\ncoins = pd.DataFrame(coins)\n# print(coins.head())\ncoins.head()\n\nresult:\n\nif you need data of all pairs it is better use multithreading or asynchronous requests\n", "you can get crypto symbol list from yahoo.finance only as \"coin-USD\"\nimport requests\nfrom requests_html import HTMLSession\nsession = HTMLSession()\nnum_currencies=250\nresp = session.get(f\"https://finance.yahoo.com/crypto?offset=0&count={num_currencies}\")\ntables = pd.read_html(resp.html.raw_html) \ndf = tables[0].copy()\nsymbols_yf = df.Symbol.tolist()\nprint(symbols_yf[:15])\nprint(df.head(5))\n\nresult:\n\n" ]

[ 0, 0, 0, 0, 0 ]

[]

[]

[ "cryptocurrency", "python", "yfinance" ]

[ "Here is an article describing a Python multithreaded approach to bulk loading into Snowflake Zero to Snowflake: Multi-Threaded Bulk Loading with Python. Also note to optimize the number of parallel operations for a load, Snowflake recommends data files roughly 100-250 MB (or larger) in size compressed.\n" ]

[ 0 ]

[]

[]

[ "python", "snowflake_cloud_data_platform" ]

[ "It looks like you are setting the whole image as a mask, this is why the rounded corners have no effect at all from your pink background. I myself was struggling a lot with this task aswell and ended up using pillow instead of OpenCV. I don't know if it is more performant, but I got it running.\nHere the code that works for your example:\nfrom PIL import Image\n\n# load images\nbackground = Image.open(r\"pink.png\")\n# load image and scale it to the same size as the background\nforeground = Image.open(r\"yellow.png\").resize(background.size)\n# split gives you the r, g, b and alpha channel of the image.\n# For the mask we only need alpha channel, indexed at 3\nmask = background.split()[3]\n\n# we combine the two images and provide the mask that is applied to the foreground.\nim = Image.composite(background, foreground, mask)\nim.show()\n\nIf your background is not monochrome as in your example, and you want to use the version, where you paste your original image, you have to create an empty image with the same size as the background, then paste your foreground to the position (your gridpos), e.g. like this:\ncanvas = Image.new('RGBA', background.size)\ncanvas.paste(foreground, gridpos)\nforeground = canvas\n\nHope this helps!\n" ]

[ 1 ]

[]

[]

[ "alpha_transparency", "opencv", "python" ]

[ "Sometimes fast and readable are the same, but more often they depend on your input and wishes. For example, what you are doing is still considered fast for most people, less than a second for a grid of 2500x2500 or 6.250.000 entries.\nNow for your question it is noticable, that if you know the row and column counts of 1's, you can quickly look up if there are more than one 1 in a row or column.\nEventhough fast and readable are very personal, I will share with you my solutions to the problem.\nPure python\nThis solution uses only pure python that is available in the library.\nimport itertools\n\nrows = [row.count(1) for row in grid]\ncols = [col.count(1) for col in zip(*grid)]\n\nwidth, height = len(grid), len(grid[0])\nindices = itertools.product(range(width), range(height))\n\nanswer = sum(1 for irow, icol in indices if grid[irow][icol] == 1 and (rows[irow] > 1 or cols[icol] > 1))\n\nFor this I am using several small tricks or perks of the python language, which are the following:\n\nrow.count(1), which counts the number of times the element 1 occurs in a list.\nzip(*grid), which is a reverse zipping, see this post from Mike Corcoran for more information.\nitertools.product, which generates all indices for me (documentation)\nsum(1 for each in val if <condition on each>), which is a generator expression.\n\nFor me the above solution, is considered very readable and concise. Now on the speed front, the solution is a bit faster than yours, but how much you might wonder? Well if we go to a grid of 5000x5000 or 25.000.000 elements, the time difference is about 4 seconds (from 12s to 8s).\nWell that is kinda bad... All that trouble for 4 seconds on something that would be considered enormous for most programmers, and probably only be used once. And if you have that many elements, you will probably not use a list anymore, but a numpy array (docs).\nPure Numpy\nSo how would the solution look like if it was a numpy array?\nimport numpy as np\n\n# Convert grid from `list` to a `np.ndarray`\narr = np.array(grid)\n\n# Get to row and column totals\ncols = arr.sum(axis=0)\nrows = arr.sum(axis=1)\n\n# Get all indices that are `1`, and combine the x and y coordinates by zip.\nindices = zip(*np.where(arr == 1))\nanswer = sum(1 for irow, icol in indices if rows[irow] > 1 or cols[icol] > 1)\n\nThe above is even more readable than before, you just need to know a few numpy convetions and methods. Namely, that numpy uses row, col for indexing, instead of the traditional (col, row) or x, y indexing. Therefore axis=0, loops over the rows, which give you the column totals. And that np.where returns the indices of where the condition inside the brackets is True.\nNow the solution is slightly slower than the pure python solution, because it has to convert the list to an array, but if your data was already a numpy array, your solution and the pure python solution would take almost twice the amount of time compared to the numpy solution.\nConclusion\nAbove are two different alternative solutions that would solve the same problem, but the readability and speed are depending on your knowledge of the language and your needs.\nIf the grid is not going to be much langer than a 1000 elements, all solutions are equally quick, namely instantaneously from a user perspective.\nAnd the readability is as good as your skills are.\n" ]

[ 1 ]

[]

[]

[ "hash", "list", "optimization", "python", "python_3.x" ]

[ "You're spot on, it looks like just a formatting issue. Can you remove any back quotes and make everything into a single line?\n" ]

[ 0 ]

[]

[]

[ "azure_functions", "prefect", "python" ]

[ "That's a very general question... Using plt.scatter() is certainly a good option. Then just add the two lines to the plot (e.g. using axhline and axvline).\nSlightly adapting this example:\nimport numpy as np\nimport matplotlib.pyplot as plt\n# don't show right and top axis[![enter image description here][1]][1]\nimport matplotlib as mpl\nmpl.rcParams['axes.spines.right'] = False\nmpl.rcParams['axes.spines.top'] = False\n\n# some random data\nN = 50\nx = np.random.randint(-10, high=11, size=N, dtype=int)\ny = np.random.randint(-10, high=11, size=N, dtype=int)\ncolors = np.random.rand(N)\narea = (30 * np.random.rand(N))**2 # 0 to 15 point radii\n\n# creating a vertical and a horizontal line\nplt.axvline(x=0, color='grey', alpha=0.75, linestyle='-')\nplt.axhline(y=0, color='grey', alpha=0.75, linestyle='-')\n# scatter plot\nplt.scatter(x, y, s=area, c=colors, alpha=0.5)\n\nplt.show()\n\n\n" ]

[ 0 ]

[]

[]

[ "plot", "python", "scatter_plot" ]

[ "Using \"subscriber\" and \"consumer\" synonymously in regards to the contract using the subscription. Also, VRF v2 Mock can be found at Chainlink GitHub\nYou can add the contract as a subscriber upon deployment in two ways:\n\nset the contract as a subscriber within the .py file you're using for deployment\ninitialize the contract as a subscriber within the .sol contract itself\n\nIn both methods you will need to do two things in the following order: create a subscription, and add the contract as a consumer to that subscription. Your code only adds a consumer but does not actually create a subscription.\n1. Set contract as subscriber within the .py file:\n\nsource: Medium article\n.py deployment file: --> assuming deploying mock\n\ndef deploy_VRFTest():\naccount = accounts[0]\n# deploy mock VRF coordinator ####################################################\n# VRFCoordinatorV2Mock.deploy(_baseFee, _gasPriceLink, {\"from\": account})\nvrf_contract = VRFCoordinatorV2Mock.deploy(25 * 10e15, 10e9, {\"from\": account})\n# create subscription & get subscription ID #######################################\nsub_id_tx = vrf_contract.createSubscription({\"from\": account})\n# add these waits so Web3 doesn't freak out and error, can also use time.sleep()\nsub_id_tx.wait(1)\nsub_id = sub_id_tx.events[\"SubscriptionCreated\"][\"subId\"] # get subscription ID\n# fund subscription ################################################################\n# must be greater than set _baseFee\nfund_amount_link = 30 * 10e15\nfund_vrf_tx = vrf_contract.fundSubscription(\n sub_id, fund_amount_link, {\"from\": account}\n)\n# request random words #############################################################\n# goerli key_hash, can use any for development\nkey_hash = \"0x79d3d8832d904592c0bf9818b621522c988bb8b0c05cdc3b15aea1b6e8db0c15\"\n# blocks to wait before confirming random words (can be any #)\nmin_request_confirm = 3\ngas_lim = 10e6 # 10e5 is recommended, added extra to be sure\nnum_words = 1 # number of random words\nrequest_tx = vrf_contract.requestRandomWords(\n key_hash, sub_id, min_request_confirm, gas_lim, num_words, {\"from\": account}\n)\nrequest_tx.wait(1)\nrequest_id = request_tx.events[\"RandomWordsRequested\"][\"requestId\"]\n# fulfill request ####################################################################\n# deploy \"ExampleContract.sol\"\ncontract = ExampleContract.deploy(vrf_contract.address, {\"from\": account})\n# fulfill random words ###############################################################\n# vrf_contract.fulfillRandomWords(_requestId, _consumer, {\"from\": account})\nfulfill_tx = vrf_contract.fulfillRandomWords(\n request_id, contract.address, {\"from\": account}\n)\nfulfill_tx.wait(1)\n# return the random number ###########################################################\nsuccess = fulfill_tx.events[\"RandomWordsFulfilled\"][\"success\"]\nif success:\n random_word = contract.s_randomWords(0)\n print(f\"random number is {random_word}\")\ntime.sleep(10)\n\n\n.sol contract file:\npragma solidity >=0.6.0 <0.9.0;\nimport \"@chainlink/contracts/src/v0.8/interfaces/VRFCoordinatorV2Interface.sol\";\nimport \"@chainlink/contracts/src/v0.8/VRFConsumerBaseV2.sol\";\n contract ExampleContract is VRFConsumerBaseV2 {\n constructor(address _VRFCoordinator) VRFConsumerBaseV2(_VRFCoordinator) {}\n uint256[] public s_randomWords;\n function fulfillRandomWords(uint256, uint256[] memory randomWords)\n internal\n override\n {\n s_randomWords = randomWords;\n }\n }\n\n\n\n2. Set contract as subscriber within the .sol file:\n\nsource: Chainlink docs\n\nthe actual source has a few other functions but I removed them to codense\n\n.sol file:\n\nnotice how the constructor for this file calls the createNewSubcription() function which not only creates a new subscription, but also adds the contract as a consumer to the subscription\npragma solidity ^0.8.7;\nimport \"@chainlink/contracts/src/v0.8/interfaces/VRFCoordinatorV2Interface.sol\";\nimport \"@chainlink/contracts/src/v0.8/VRFConsumerBaseV2.sol\";\n contract VRFTest is VRFConsumerBaseV2 {\n VRFCoordinatorV2Interface COORDINATOR;\n\n // Goerli coordinator\n address vrfCoordinator = 0x2Ca8E0C643bDe4C2E08ab1fA0da3401AdAD7734D;\n // The goerli gas lane to use, which specifies the maximum gas price to bump to.\n bytes32 keyHash =\n 0x79d3d8832d904592c0bf9818b621522c988bb8b0c05cdc3b15aea1b6e8db0c15;\n uint32 callbackGasLimit = 100000;\n uint16 requestConfirmations = 3;\n uint32 numWords = 2;\n\n // Storage parameters\n uint256[] public s_randomWords;\n uint256 public s_requestId;\n uint64 public s_subscriptionId;\n address s_owner;\n\n constructor() VRFConsumerBaseV2(vrfCoordinator) {\n COORDINATOR = VRFCoordinatorV2Interface(vrfCoordinator);\n s_owner = msg.sender;\n //Create a new subscription when you deploy the contract.\n createNewSubscription();\n }\n\n // Assumes the subscription is funded sufficiently.\n function requestRandomWords() external onlyOwner {\n // Will revert if subscription is not set and funded.\n s_requestId = COORDINATOR.requestRandomWords(\n keyHash,\n s_subscriptionId,\n requestConfirmations,\n callbackGasLimit,\n numWords\n );\n }\n\n function fulfillRandomWords(\n uint256, /* requestId */\n uint256[] memory randomWords\n ) internal override {\n s_randomWords = randomWords;\n }\n\n // Create a new subscription when the contract is initially deployed.\n function createNewSubscription() private onlyOwner {\n s_subscriptionId = COORDINATOR.createSubscription();\n // Add this contract as a consumer of its own subscription.\n COORDINATOR.addConsumer(s_subscriptionId, address(this));\n }\n\n function addConsumer(address consumerAddress) external onlyOwner {\n // Add a consumer contract to the subscription.\n COORDINATOR.addConsumer(s_subscriptionId, consumerAddress);\n }\n\n modifier onlyOwner() {\n require(msg.sender == s_owner);\n _;\n }\n }\n\n\n\n" ]

[ 0 ]

[]

[]

[ "brownie", "chainlink", "python", "solidity" ]

[ "You are trying to run python directly within powershell. But the powershell-interpreter speaks only powershell and cannot natively interprete python code.\nYou have to put the python code in a python file, e.g. my_code.py and call/execute it with python my_code.py from within powershell. Now the python interpreter is used to run the script. See How to run python code for details.\n", "I've split the code block in your question into 3 distinct parts, one for each error.\nWhile, yes, you should put Python code into an actual script, the REPL will work too, but pip install is not Python code. Neither is plain text like Windows 11 Powershell. If you're copying code off some blog/docs, then only do actual Python code...\nPython cares about indentation, so reviews not defined was because your for loop was not indented where the variable was defined.\n'str' object has no attribute 'numpy' is on this line print(review.numpy().decode(\"utf-8\")) because strings don't have numpy functions... It should already be decoded, as well. So just print out the review text directly.\nI see that the video/image you've linked is from 2019... Tensorflow has had several new releases since then, so the API may have changed. You'll want to consult official documentation for the version you've installed\n" ]

[ 1, 0 ]

[]

[]

[ "python" ]

[ "You might want to try something like following:\ndef get_pos_nums(num):\n pos_nums = []\n while num != 0:\n pos_nums.append(num % 10)\n num = num // 10\n return pos_nums\n\nAnd call this method as following.\n>>> get_pos_nums(9876)\n[6, 7, 8, 9]\n\nThe 0th index will contain the units, 1st index will contain tens, 2nd index will contain hundreds and so on...\nThis function will fail with negative numbers. I leave the handling of negative numbers for you to figure out as an exercise.\n", "Like this?\na = str(input('Please give me a number: '))\n\nfor i in a[::-1]:\n print(i)\n\n\nDemo:\nPlease give me a number: 1324\n4\n2\n3\n1\n\nSo the first number is ones, next is tens, etc.\n", "num = 1234\n\nthousands = num // 1000\nhundreds = (num % 1000) // 100\ntens = (num % 100) // 10\nunits = (num % 10)\n\nprint(thousands, hundreds, tens, units)\n# expected output: 1 2 3 4\n\n\"//\" in Python stands for integer division. It largely removes the fractional part from the floating-point number and returns an integer\nFor example:\n4/3 = 1.333333\n4//3 = 1\n\n", "You could try splitting the number using this function:\ndef get_place_values(n):\n return [int(value) * 10**place for place, value in enumerate(str(n)[::-1])]\n\nFor example:\nget_place_values(342)\n>>> [2, 40, 300]\n\nNext, you could write a helper function:\ndef get_place_val_to_word(n):\n n_str = str(n)\n num_to_word = {\n \"0\": \"ones\",\n \"1\": \"tens\",\n \"2\": \"hundreds\",\n \"3\": \"thousands\"\n }\n return f\"{n_str[0]} {num_to_word[str(n_str.count('0'))]}\"\n\nThen you can combine the two like so:\ndef print_place_values(n):\n for value in get_place_values(n):\n print(get_place_val_to_word(value))\n\nFor example:\nnum = int(input(\"Please give me a number: \"))\n# User enters 342\nprint_place_values(num)\n>>> 2 ones\n4 tens\n3 hundreds\n\n", "num=1234\ndigit_at_one_place=num%10\nprint(digit_at_one_place)\ndigits_at_tens_place=(num//10)%10\ndigits_at_hund_place=(num//100)%10\ndigits_at_thou_place=(num//1000)%10\nprint(digits_at_tens_place)\nprint(digits_at_hund_place)\nprint(digits_at_thou_place)\n\nthis does the job. it is simple to understand as well.\n", "Please note that I took inspiration from the above answer by 6pack kid to get this code. All I added was a way to get the exact place value instead of just getting the digits segregated. \nnum = int(input(\"Enter Number: \"))\nc = 1\npos_nums = []\nwhile num != 0:\n z = num % 10\n pos_nums.append(z *c)\n num = num // 10\n c = c*10\nprint(pos_nums)\n\nOnce you run this code, for the input of 12345 this is what will be the output:\nEnter Number: 12345\n[5, 40, 300, 2000, 10000]\n\nThis helped me in getting an answer to what I needed.\n", "money = int(input(\"Enter amount: \"))\nthousand = int(money // 1000)\nfive_hundred = int(money % 1000 / 500)\ntwo_hundred = int(money % 1000 % 500 / 200)\none_hundred = int(money % 1000 % 500 % 200 / 100)\nfifty = int(money % 1000 % 500 % 200 % 100 / 50)\ntwenty = int(money % 1000 % 500 % 200 % 100 % 50 / 20)\nten = int(money % 1000 % 500 % 200 % 100 % 50 % 20 / 10)\nfive = int(money % 1000 % 500 % 200 % 100 % 50 % 20 % 10 / 5)\none = int(money % 1000 % 500 % 200 % 100 % 50 % 20 % 10 % 5 / 1)\nif thousand >=1: \n print (\"P1000: \" , thousand)\nif five_hundred >= 1:\n print (\"P500: \" , five_hundred)\nif two_hundred >= 1:\n print (\"P200: \" , two_hundred)\nif one_hundred >= 1:\n print (\"P100: \" , one_hundred)\nif fifty >= 1:\n print (\"P50: \" , fifty)\nif twenty >= 1:\n print (\"P20: \" , twenty)\nif ten >= 1:\n print (\"P10: \" , ten)\nif five >= 1:\n print (\"P5: \" , five)\nif one >= 1:\n print (\"P1: \" , one)\n\n", "Quickest way:\nnum = str(input(\"Please give me a number: \"))\nprint([int(i) for i in num[::-1]])\n\n", "This will do it, doesn't use strings at all and handles any integer passed for col sensibly.\ndef tenscol(num: int, col: int):\n ndigits = 1\n while (num % (10**ndigits)) != num:\n ndigits += 1\n x = min(max(1, col), ndigits)\n y = 10**max(0, x - 1)\n return int(((num % 10**x) - (num % y)) / y)\n\nusage:\nprint(tenscol(9785,-1))\nprint(tenscol(9785,1))\nprint(tenscol(9785,2))\nprint(tenscol(9785,3))\nprint(tenscol(9785,4))\nprint(tenscol(9785,99))\n\nOutput:\n5\n5\n8\n7\n9\n9\n\n", "def get_pos(num,unit):\n return int(abs(num)/unit)%10\n\nSo for \"ones\" unit is 1 while for \"tens\", unit is 10 and so forth.\nIt can handle any digit and even negative numbers effectively.\nSo given the number 256, to get the digit in the tens position you do\nget_pos(256,10)\n>> 5\n\n", "I had to do this on many values of an array, and it's not always in base 10 (normal counting - your tens, hundreds, thousands, etc.). So the reference is slightly different: 1=1st place (1s), 2=2nd place (10s), 3=3rd place (100s), 4=4th place (1000s). So your vectorized solution:\nimport numpy as np\ndef get_place(array, place):\n return (array/10**(place-1)%10).astype(int)\n\nWorks fast and also works on arrays in different bases.\n", "# method 1\n\nnum = 1234\nwhile num>0:\n print(num%10)\n num//=10\n\n# method 2\n\nnum = 1234\n\nprint('Ones Place',num%10)\nprint('tens place',(num//10)%10)\nprint(\"hundred's place\",(num//100)%10)\nprint(\"Thousand's place \",(num//1000)%10)\n\n" ]

[ 14, 6, 4, 1, 1, 0, 0, 0, 0, 0, 0, 0 ]

[ "In Python, you can try this method to print any position of a number.\nFor example, if you want to print the 10 the position of a number,\nMultiply the number position by 10, it will be 100,\nTake modulo of the input by 100 and then divide it by 10.\nNote: If the position get increased then the number of zeros in modulo and in divide also increases:\ninput = 1234\n\nprint(int(input % 100) / 10 )\n\nOutput:\n3\n\n", "So I saw what another users answer was and I tried it out and it didn't quite work, Here's what I did to fix the problem. By the way I used this to find the tenth place of a number\n# Getting an input from the user\n\ninput = int(input())\n\n# Finding the tenth place of the number\n\nprint(int(input % 100) // 10)\n\n\n" ]

[ -1, -1 ]

[ "numbers", "operators", "python" ]

[ "Although this question is old, I just faced the same issue and solved it.\nThe issue is after adding patch to the axes, you need to give the coordinates/points in display reference frame. This can be performed with:\nax.transData.transform()\n\nI added one line to your code ignoring import statements. So the code goes here:\nimport matplotlib.pyplot as plt\nimport matplotlib.patches as patches\n\nfig, ax = plt.subplots()\nrect = patches.Rectangle([0.2, 0.3], 0.8, 0.5)\n\npnt = [0.4, 0.45] # point inside rect\nprint(\"Before adding patch to axes:\")\nprint(rect.get_path().vertices)\nprint(rect.contains_point(pnt))\nprint(rect.get_path().contains_point(pnt))\n\nprint(\"After adding patch to axes\")\nax.add_patch(rect)\nprint(rect.get_path().vertices)\n# added lines\npnt_in_display_coordiantes = ax.transData.transform(pnt)\nprint(rect.contains_point(pnt_in_display_coordiantes))\nprint(rect.get_path().contains_point(pnt))\n\nplt.show()\n\nOutput:\nBefore adding patch to axes:\n[[0. 0.]\n[1. 0.]\n[1. 1.]\n[0. 1.]\n[0. 0.]]\nTrue\nTrue\nAfter adding patch to axes\n[[0. 0.]\n[1. 0.]\n[1. 1.]\n[0. 1.]\n[0. 0.]]\nTrue\nTrue\n\nFor more: https://matplotlib.org/stable/tutorials/advanced/transforms_tutorial.html\n" ]

[ 0 ]

[]

[]

[ "matplotlib", "patch", "path", "python" ]

[ "Simple thing, you should load the template tag in news.html template which is registered.\nJust load the tag in news.html template:\n{% load tag_name %} #Add here tag name to load\n\nNote: Please ensure that template tag setting is added in settings.py file\n", "You need to define the python code containing your tag codes in the TEMPLATES variable in settings.py.\nTEMPLATES = [\n {\n 'BACKEND': 'django.template.backends.django.DjangoTemplates',\n 'DIRS': [\n str(BASE_DIR.joinpath('templates'))\n ],\n 'APP_DIRS': True,\n 'OPTIONS': {\n 'context_processors': [\n 'django.template.context_processors.debug',\n 'django.template.context_processors.request',\n 'django.contrib.auth.context_processors.auth',\n 'django.contrib.messages.context_processors.messages',\n ],\n 'libraries':{\n 'tagname': 'appname.news', # your template tag\n\n }\n },\n },\n]\n\n" ]

[ 0, 0 ]

[]

[]

[ "django", "django_templates", "python" ]

[ "You can print the min(rows, n - rows - 1) instead of rows -\nn = 5\nfor rows in range(n):\n for cols in range(n):\n if((rows == cols) or (rows+cols)==n-1 ):\n print(min(rows, n - rows - 1),end=\"\")\n else:\n print(\" \",end=\"\")\n print()\n\nOutput:\n0 0\n 1 1 \n 2 \n 1 1 \n0 0\n\nFor n = 7 -\n0 0\n 1 1 \n 2 2 \n 3 \n 2 2 \n 1 1 \n0 0\n\nFor n = 6 -\n0 0\n 1 1 \n 22 \n 22 \n 1 1 \n0 0\n\n", "This is a longer alternative, but it does work. However you will run into a weird issue when entering an even number.\nn = int(input(\"Enter Total Rows: \"))\n\nmatrix = []\n\n# Append each row and column to the matrix.\nfor rows in range(n):\n row = []\n for cols in range(n):\n row.append(\" \")\n\n matrix.append(row)\n\nfor i in range(n):\n if i > n/2: # After we reach the center, start to subtract numbers instead of adding\n matrix[i][i] = abs(i - (n - 1)) # Abs is so we can invert the negative number to a positive\n else:\n matrix[i][i] = i\n\n if i > n/2: # Ditto, but now instead of top left to bottom right, it goes top right to bottom left\n matrix[i][n - (i + 1)] = abs(i - (n - 1))\n else:\n matrix[i][n - (i + 1)] = i\n\n# Print each row of the matrix\nfor i in range(len(matrix)):\n print(*matrix[i])\n\n" ]

[ 0, 0 ]

[]

[]

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

[ "It's not very clear what you are trying to do. Your line '[I] am trying to get multiple outputs from multiple inputs accordingly' sounds like you want to take an array as input and then perform your function on every element.\nI am not exactly clear on what your function is, because it sounds like you want to calculate the nth power of 2, but it also sounds like you want to do some sort of modular arithmetic.\nI'll assume it's the first case. Then, what you want is a function that's taking inputs, and as output giving 2 to the input power. There's a lot of ways you could do this, the simplest would be some sort of list comprehension, which is like a neat 1-line for loop that Python is a big fan of. It would look like this:\ndef nth_power(input_list: list[Int]) -> list[Int]:\n return [2**n for n in input_list]\n\nThis gives you a list of numbers where each number is 2 raised to the power of the number in that place in the first list.\nI think numpy arrays would also suit your needs well. Start now with a numpy array of the numbers you want, so presumably 0... 1000. This looks like:\ninput_list = np.arange(1000) # generates the np.array([0... 999])\n\nThen, you can just pass this to np.pow(), like this:\noutput_list = np.power(2, input_list)\n\nHere, this will give output_list as\narray([ 1, 2, 4, 8, 16, 32, 64, 128, 256, 512], dtype=int32)\n\nHope that helps, feel free to ask if you need some clarification:)\n" ]

[ 0 ]

[]

[]

[ "arrays", "numpy_ndarray", "python" ]

[ "Without seeing the rest of the code I think that the error description is correct. Your import is unused. That means that you imported it but in the actual script the shuffle function was never accessed.\nTry using it in your code:\nfrom random import shuffle\n\nmy_shuffled_number = shuffle([1, 2, 3, 4, 5])\n\nprint(my_shuffled_number)\n\nThis small script makes use of the import function.\n" ]

[ 0 ]

[]

[]

[ "import", "python" ]

[ "xtrabackup is not entirely helpful showing what it did recognise, it would help if it also showed the entire command line as received (especially if it turns out it read a defaults file which means the problem could be entirely elsewhere).\nIn the meantime, you need to see exactly what trove/guestagent/strategies/backup/mysql_impl.py is doing at line 94.\nI would hazard guess that it's building up a command-line string and somewhere (other than the --datadir=/var/lib/mysql/data which xtrabackup says it recognized) is another occurrence of /var/lib/mysql/data. If you can print out the full command line before executing it should offer a clue.\n" ]

[ 0 ]

[]

[]

[ "backup", "mysql", "percona", "python" ]

[ "Try this,\ndata = [[10.2427, 43.5703, 0.63],\n [10.2427, 43.5703, 0.94],\n [10.2427, 43.5703, 0.76],\n [10.2427, 43.5703, 0.76],\n [10.2427, 43.5703, 1.03],\n [10.2427, 43.5703, 0.86],\n [10.2427, 43.5703, 0.74],\n [10.2427, 43.5703, 1.71],\n [10.2427, 43.5703, 3.07],\n [12.199, 41.708, 0.96],\n [12.199, 41.708, 0.89],\n [12.199, 41.708, 1.29],\n [12.199, 41.708, 0.24],\n [12.199, 41.708, 1.59],\n [12.199, 41.708, 1.78],\n [12.199, 41.708, 0.39],\n [12.199, 41.708, 1.54],\n [12.199, 41.708, 1.62]]\n \ndf = pd.DataFrame(data)\n\nprint(df.groupby([0, 1])[2].apply(list).reset_index().apply(lambda x: [x[0], x[1]]+x[2], axis=1).values.tolist())\n\nExplanation:\n\nCreate dataframe out of your current output\ngroupby first two cols and get other elements as list\nRestructure to get the expected output\n\nO/P:\n[[10.2427, 43.5703, 0.63, 0.94, 0.76, 0.76, 1.03, 0.86, 0.74, 1.71, 3.07], [12.199, 41.708, 0.96, 0.89, 1.29, 0.24, 1.59, 1.78, 0.39, 1.54, 1.62]]\n\nNote: The above code is just to give you an idea, it can be further improved. If I get some free time, I will post that as well.\n" ]

[ 0 ]

[]

[]

[ "dataframe", "list", "loops", "pandas", "python" ]

[ "You need a specific condition in which to put the \"break\". Here, at every iteration you are breaking.\nBy looking at the code, i think you have to indent break another time and put it under \"print(\"Bad VOTE\", name)\". So in the case of a bad vote, the loop brakes.\n" ]

[ 0 ]

[]

[]

[ "python", "while_loop" ]

[ "This might be trivial, but have you tried closing your command line window and opening a new one? This is supposed to reload all the environment variables.\nTry typing\necho %PATH%\n\ninto the command prompt and see if you can find your Python directory there.\nAlso, the second part of your addition to the PATH environment variable is indeed unnecessary.\n", "I had the same problem: python not being recognized, with python in the path which was was not truncated. \nFollowing the comment of eryksun in yossim's answer:\n\nAlso, if you installed for all users you should have %SystemRoot%\\py.exe, which >is typically C:\\Windows\\py.exe. So without setting Python's directory in PATH >you can simply run py to start Python; if 2.x is installed use py -3 since >Python 2 is the default. – eryksun \n\nI tried to use py instead of python and it worked.\nMeaning: \npython setup.py build -> does NOT work.\npy setup.py build -> does work.\nHope it helps\n", "I was also having the same problem.\nTurns out the path I added included '..\\python.exe' at the end, which as turns out was not required. I only needed to add the directory in which 'python.exe' is in (which in my case is the Anaconda's distribution directory in Users folder), similar to what we do when installing JDK in our system's PATH variable.\nHope it helps!\n", "It wasn't working for me even after adding the path. What finally did the trick, was changing the order of listed paths in the PATH variable. I moved %USERPROFILE%\\AppData\\Local\\Microsoft\\WindowsApps down vs. having it the first path listed there.\n", "Environment PATH Length Limitation is 1024 characters\nIf restarting your cmd window does not work you might have reached the character limit for PATH, which is a surprisingly short 1024 characters.\nNote that the user interface will happily allows you to define a PATH that is way longer than 1024, and will just truncate anything longer than this. Use\necho %PATH%\n\nin your cmd window to see if the PATH being truncated.\nSolution\nUnfortunately, there is no good way to fix this besides removing something else from your PATH.\n\nNOTE: Your PATH = SYSTEM_PATH + USER_PATH, so you need to make sure the combined is < 1024.\n", "Also, make sure to leave no spaces after the semi-colon.\nFor example, this didn't work for me:\nC:\\Windows\\system32; C:\\Python27; C:\\Python27\\Scripts;\nBut, this did:\nC:\\Windows\\system32;C:\\Python27;C:\\Python27\\Scripts;\n", "I'm late to the game here, but I'd like to share my solution for future users. The previous answers were on the right track, but if you do not open the CMD as an administrator, then you will be thrown that same error. I know this seems trivial and obvious, but after spending the past 8 hours programming before attempting to install Django for the first time, you might be surprised at the stupid mistakes you might make.\n", "I have faced same problem even though my path contains 400 characters.\nTry to update the path from the command line(Run as administrator)\nCommand to update path: setx path \"%path%;c:\\examplePath\"\nAfter this command I could see that paths that I configured earlier in environment variables got updated and working.\nTo check the configured paths: echo %PATH%\n", "I was facing similar porblem. What helped me is where command.\n\nC:\\WINDOWS\\system32>where python\nC:\\Users\\xxxxxxx\\AppData\\Local\\Microsoft\\WindowsApps\\python.exe\nC:\\Program Files (x86)\\Microsoft Visual\nStudio\\Shared\\Python39_86\\python.exe\n\nOn updating PATH variable to point to only one desired directory (basically I removed %USERPROFILE%\\AppData\\Local\\Microsoft\\WindowsApps from PATH) fixed my problem.\n", "I had the same issue with Python 2.7 on Windows 10 until I changed the file path in Enviroment Variables to the folder path, ie C:\\Python27\\python.exe didn't work but C:\\Python27\\ did work.\n", "I did everything:\n\nAdded Python to PATH\nUninstall all the Pythons - Both from downloaded python.org and Microsoft Store and reinstall from python.org\nChange the order of PATH\nDeleted %USERPROFILE%\\AppData\\Local\\Microsoft\\WindowsApps from PATH\n\nBut nothing worked. What worked for me was:\nSettings > Application > App execution aliases. Then disable all the Pyhtons from here and it worked!\n\n" ]

[ 24, 22, 8, 6, 4, 2, 1, 1, 1, 0, 0 ]

[ "For me, installing the 'Windows x86-64 executable installer' from the official python portal did the trick.\nPython interpreter was not initially recognized, while i had installed 32 bit python.\nUninstalled python 32 bit and installed 64 bit.\nSo, if you are on a x-64 processor, install 64bit python.\n", "I tried it multiple times with the default installer option, the first one, (Python 3.7.3) with both 'add to environment variable' and 'all users' checked, though the latter was greyed out and couldn't be unchecked.\nIt failed to work for other users except for the user I installed it under until I uninstalled it and chose \"Custom Install\". It then clearly showed the install path being in the C:\\Program Files\\Python37 directory when it was failing to install it there the other way even though the 'All Users' option was checked.\n", "Same thing was happening with me when i was trying to open the python immediately with CMD.\nThen I kept my in sleep mode and started CMD using these Key Windows_key+R, typed cmd and OK. Then the package of python worked perfectly.\n", "\nUninstall python and pyqt\nThen go to pyqt setup and open installation but don't install. You will see a message box saying something like pyqt version built with python version 32bit/64bit. \nThen see python version bit and download that version from python.org from all release menu. \nThen first install python and then install pyqt. It will work like butter.\n\n", "I spent sometime checking and rechecking the path and restarting to no avail.\nThe only thing that worked for me was to rename the executable C:\\Python34\\python.exe to C:\\Python34\\python34.exe. This way, calling typing python34 at the command line now works.\nOn windows it seems that when calling 'python', the system finds C:\\Python27 in the path before it finds C:\\Python34\nI'm not sure if this is the right way to do this, seems like a hack, but it seems to work OK.\n" ]

[ -1, -1, -1, -1, -2 ]

[ "cmd", "command_line", "python", "python_3.x", "windows" ]

[ "If I recall correctly:\nx1 = (center_X-width/2)*image_width\nx2 = (center_X+width/2)*image_width\ny1 = (center_y-height/2)*image_height\ny2 = (center_y+height/2)*image_height\n\n", "Given that the upper-left corner of the image is [0,0]: For the upper-left corner you have to do [x,y] = [center_X, center_Y] - 1/2 * [width, height] . For the bottom-right corner [x,y] = [center_X, center_Y] + 1/2 * [width, height] .\n", "def get_coord(label_file, img_width, img_height):\n\nlfile = open(label_file)\ncoords = []\nall_coords = []\n\nfor line in lfile:\n l = line.split(\" \")\n \n coords = list(map(float, list(map(float, l[1:5]))))\n x1 = float(img_width) * (2.0 * float(coords[0]) - float(coords[2])) / 2.0\n y1 = float(img_height) * (2.0 * float(coords[1]) - float(coords[3])) / 2.0\n x2 = float(img_width) * (2.0 * float(coords[0]) + float(coords[2])) / 2.0\n y2 = float(img_height) * (2.0 * float(coords[1]) + float(coords[3])) / 2.0\n tmp = [x1, y1, x2, y2]\n all_coords.append(list(map(int, tmp)))\nlfile.close()\nreturn all_coords\n\n" ]

[ 2, 1, 0 ]

[]

[]

[ "computer_vision", "python", "yolo" ]

[ "I assume you have already installed Visual C++ Latest Build Tools\nfrom Visual Studio 2022 Build Tools: https://aka.ms/vs/17/release/vs_buildtools.exe\nNow, Check the Pycocotools/PythonAPI folder and modify the setup.py file\nchange these couple of lines as the following code and run this command again\nPreviously :\nextra_compile_args=['-Wno-cpp', '-Wno-unused-function', '-std=c99']\nRemove '-Wno-cpp', '-Wno-unused-function' from the ext_modules Extension in setup.py\nOnly keep '-std=c99', and it will look like this.\nextra_compile_args=['-std=c99']\nNow run this command once again.\npython setup.py build_ext --inplace\nthis should successfully build the pycocotools in windows 11.\n" ]

[ 0 ]

[]

[]

[ "python" ]

[ "You need to add each player name entered to a list. Here is a starting point of what you need in your code:\nfrom random import choice\n\nnumber_of_players = int(input(\"How many players are there: \"))\nplayers = []\n\nfor _ in range(number_of_players):\n players.append(input(\"name the players: \"))\n\nprint(choice(players))\n\n", "That loop reassigns the ask_player variable on each iteration, erasing the previous value.\nPresumably you meant to save each value in a list:\nplayers = []\nfor i in range(player_number):\n players.append(input(\"Player name: \"))\n\nprint(random.choice(players))\n\n", "The problem is that in each for-loop iteration, you are reassigning the ask_player var to a string. When you pass a string to random.choice(...), it picks a random letter of that string (since strings can be indexed like arrays). Just define an array before the loop and append on each iteration:\nimport random\n\nplayer_numberCount = input(\"How many players are there: \")\nplayer_number = int(player_numberCount)\n\nplayers = []\nfor i in range(player_number):\n players.append(input(f\"name player {i + 1}: \"))\n\nprint(random.choice(players))\n\n", "import random\n\nplayer_number = int(input(\"How many players are there: \"))\nplayer_list = []\n\n\nfor _ in range(player_number):\n ask_player = input(\"name the players: \")\n player_list.append(ask_player)\n\nprint(player_list[random.randint(0, player_number)])\n\n" ]

[ 2, 0, 0, 0 ]

[]

[]

[ "input", "python" ]

[ "According to this issue, from ipywidgets 8, widgets uses _repr_mimebundle_ instead of _ipython_display_.\nSo _widget_attrs['_ipython_display_'] will raise KeyError. Use _repr_mimebundle_ as key will sovle problem.\n" ]

[ 0 ]

[]

[]

[ "ipywidgets", "jupyter_notebook", "python" ]

[ "Stop reinventing the wheel. Also, class names are supposed to be named with PascalCase.\nUse AbstractUser model:\nfrom django.contrib.auth.models import AbstractUser\n\nclass Student(AbstractUser):\n ...\n\nand in your main urls.py:\nfrom django.contrib.auth import views as auth_views\n\nurlpatterns = [\n ...\n path('login/', auth_views.LoginView.as_view(), name='login'),\n ...\n]\n\nIt is much faster and SAFER way to create new user.\n", "So I figured out out how to solve my problem. By using the AbstractUser model,i was able to create a custom user and then create another model which i extended a ForeignKey on the User model therefore allowing me to tie every user to their profile.\nHere is my models.py\nfrom django.db import models\nfrom django.contrib.auth.models import AbstractUser\n\n\n# Create your models here.\nclass User(AbstractUser):\n pass\n def __str__(self):\n return self.username\n\nclass UserProfile(models.Model):\n \"\"\"\n This is the one for model.py\n \"\"\"\n username = models.ForeignKey(User, on_delete=models.CASCADE, null=True, default=\"\")\n profile_picture = models.ImageField(blank=True, null=True, default=\"\")\n matricno = models.CharField(max_length=9, default=\"\", primary_key=True)\n email = models.EmailField(default=\"\")\n first_name = models.CharField(max_length=200, default=\"\")\n last_name = models.CharField(max_length=255, default=\"\")\n\n class Meta:\n verbose_name_plural = \"Users Profile\"\n\n def __str__(self):\n return self.first_name+ \" \"+self.last_name\n\nAnd here is my views.py\ndef signup(request):\n if request.method == \"POST\":\n form = Signup(request.POST)\n if form.is_valid():\n username = request.POST[\"username\"]\n email = request.POST[\"email\"]\n password = request.POST[\"password\"]\n password2 = request.POST[\"password2\"]\n\n user = User.objects.create_user(\n username=username,\n password=password,\n email=email,\n )\n user.save()\n login(request, user)\n messages.success(request, \"Account Created successfully for \" + username)\n return redirect(details)\n else:\n form = Signup()\n return render(request, \"accounts/register.html\", {\"form\": form})\n\ndef details(request, username):\n user = User.objects.get(username=username)\n form = Details()\n if request.method == \"POST\":\n form = Details(request.POST, request.FILES)\n if form.is_valid():\n detail = form.save(commit=False)\n detail.username = request.user\n detail.save()\n return redirect(success, pk=detail.pk)\n else:\n form = Details(initial={\"matricno\":request.user.username})\n return render(request, \"details.html\", {\"form\":form})\n\nAnd finally my forms.py that i use in creating a signup form and perfoming validation\nclass Signup(forms.Form):\n username = forms.CharField(\n max_length=9,\n widget=forms.TextInput(attrs={\"placeholder\": \"Enter Your Matric Number\"}),\n )\n\n email = forms.EmailField(\n max_length=255,\n widget=forms.EmailInput(attrs={\"placeholder\": \"Enter Your E-mail Address\"}),\n )\n\n password = forms.CharField(\n max_length=255,\n widget=forms.PasswordInput(\n attrs={\"placeholder\": \"Enter Your Password\", \"id\": \"password\"}\n ),\n )\n\n password2 = forms.CharField(\n max_length=255,\n widget=forms.PasswordInput(\n attrs={\"placeholder\": \"Confirm Your Password\", \"id\": \"password2\"}\n ),\n )\n \n def clean(self):\n super(Signup, self).clean()\n password = self.cleaned_data.get(\"password\")\n password2 = self.cleaned_data.get(\"password2\")\n username = self.cleaned_data.get(\"username\")\n email = self.cleaned_data.get(\"email\")\n\n if password != password2:\n self.errors[\"\"] = self.error_class([\"The two password fields must match\"])\n\n for instance in User.objects.all():\n if instance.username == str(username):\n self.errors[\"\"] = self.error_class([\"User already exist\"])\n elif instance.email == email:\n self.errors[\"\"] = self.error_class([\"E-mail already in use\"])\n else:\n pass\n\n return self.cleaned_data\n\n" ]

[ 0, 0 ]

[]

[]

[ "authentication", "django", "django_forms", "django_models", "python" ]

[ "you can chek it and check the paths with these commands :\n\nwhich nvidia-smi\nwhich nvcc\ncat /usr/local/cuda/version.txt\n\n" ]

[ 0 ]

[]

[]

[ "python", "pytorch" ]

[ "Simplejson 2.1 and higher has native support for Decimal type:\n>>> json.dumps(Decimal('3.9'), use_decimal=True)\n'3.9'\n\nNote that use_decimal is True by default:\ndef dumps(obj, skipkeys=False, ensure_ascii=True, check_circular=True,\n allow_nan=True, cls=None, indent=None, separators=None,\n encoding='utf-8', default=None, use_decimal=True,\n namedtuple_as_object=True, tuple_as_array=True,\n bigint_as_string=False, sort_keys=False, item_sort_key=None,\n for_json=False, ignore_nan=False, **kw):\n\nSo:\n>>> json.dumps(Decimal('3.9'))\n'3.9'\n\nHopefully, this feature will be included in standard library.\n", "I would like to let everyone know that I tried Michał Marczyk's answer on my web server that was running Python 2.6.5 and it worked fine. However, I upgraded to Python 2.7 and it stopped working. I tried to think of some sort of way to encode Decimal objects and this is what I came up with:\nimport decimal\n\nclass DecimalEncoder(json.JSONEncoder):\n def default(self, o):\n if isinstance(o, decimal.Decimal):\n return str(o)\n return super(DecimalEncoder, self).default(o)\n\nNote that this will convert the decimal to its string representation (e.g.; \"1.2300\") to a. not lose significant digits and b. prevent rounding errors.\nThis should hopefully help anyone who is having problems with Python 2.7. I tested it and it seems to work fine. If anyone notices any bugs in my solution or comes up with a better way, please let me know.\nUsage example:\njson.dumps({'x': decimal.Decimal('5.5')}, cls=DecimalEncoder)\n\n", "How about subclassing json.JSONEncoder?\nclass DecimalEncoder(json.JSONEncoder):\n def default(self, o):\n if isinstance(o, decimal.Decimal):\n # wanted a simple yield str(o) in the next line,\n # but that would mean a yield on the line with super(...),\n # which wouldn't work (see my comment below), so...\n return (str(o) for o in [o])\n return super(DecimalEncoder, self).default(o)\n\nThen use it like so:\njson.dumps({'x': decimal.Decimal('5.5')}, cls=DecimalEncoder)\n\n", "The native Django option is missing so I'll add it for the next guy/gall that looks for it.\nStarting on Django 1.7.x there is a built-in DjangoJSONEncoder that you can get it from django.core.serializers.json.\nimport json\nfrom django.core.serializers.json import DjangoJSONEncoder\nfrom django.forms.models import model_to_dict\n\nmodel_instance = YourModel.object.first()\nmodel_dict = model_to_dict(model_instance)\n\njson.dumps(model_dict, cls=DjangoJSONEncoder)\n\nPresto!\n", "In my Flask app, Which uses python 2.7.11, flask alchemy(with 'db.decimal' types), and Flask Marshmallow ( for 'instant' serializer and deserializer), i had this error, every time i did a GET or POST. The serializer and deserializer, failed to convert Decimal types into any JSON identifiable format. \nI did a \"pip install simplejson\", then \nJust by adding\nimport simplejson as json\n\nthe serializer and deserializer starts to purr again. I did nothing else...\nDEciamls are displayed as '234.00' float format.\n", "I tried switching from simplejson to builtin json for GAE 2.7, and had issues with the decimal. If default returned str(o) there were quotes (because _iterencode calls _iterencode on the results of default), and float(o) would remove trailing 0.\nIf default returns an object of a class that inherits from float (or anything that calls repr without additional formatting) and has a custom __repr__ method, it seems to work like I want it to.\nimport json\nfrom decimal import Decimal\n\nclass fakefloat(float):\n def __init__(self, value):\n self._value = value\n def __repr__(self):\n return str(self._value)\n\ndef defaultencode(o):\n if isinstance(o, Decimal):\n # Subclass float with custom repr?\n return fakefloat(o)\n raise TypeError(repr(o) + \" is not JSON serializable\")\n\njson.dumps([10.20, \"10.20\", Decimal('10.20')], default=defaultencode)\n'[10.2, \"10.20\", 10.20]'\n\n", "For Django users:\nRecently came across TypeError: Decimal('2337.00') is not JSON serializable\nwhile JSON encoding i.e. json.dumps(data)\nSolution:\n# converts Decimal, Datetime, UUIDs to str for Encoding\nfrom django.core.serializers.json import DjangoJSONEncoder \n\njson.dumps(response.data, cls=DjangoJSONEncoder)\n\nBut, now the Decimal value will be a string, now we can explicitly set the decimal/float value parser when decoding data, using parse_float option in json.loads:\nimport decimal \n\ndata = json.loads(data, parse_float=decimal.Decimal) # default is float(num_str)\n\n", "3.9 can not be exactly represented in IEEE floats, it will always come as 3.8999999999999999, e.g. try print repr(3.9), you can read more about it here: \nhttp://en.wikipedia.org/wiki/Floating_point\nhttp://docs.sun.com/source/806-3568/ncg_goldberg.html \nSo if you don't want float, only option you have to send it as string, and to allow automatic conversion of decimal objects to JSON, do something like this:\nimport decimal\nfrom django.utils import simplejson\n\ndef json_encode_decimal(obj):\n if isinstance(obj, decimal.Decimal):\n return str(obj)\n raise TypeError(repr(obj) + \" is not JSON serializable\")\n\nd = decimal.Decimal('3.5')\nprint simplejson.dumps([d], default=json_encode_decimal)\n\n", "My $.02!\nI extend a bunch of the JSON encoder since I am serializing tons of data for my web server. Here's some nice code. Note that it's easily extendable to pretty much any data format you feel like and will reproduce 3.9 as \"thing\": 3.9\nJSONEncoder_olddefault = json.JSONEncoder.default\ndef JSONEncoder_newdefault(self, o):\n if isinstance(o, UUID): return str(o)\n if isinstance(o, datetime): return str(o)\n if isinstance(o, time.struct_time): return datetime.fromtimestamp(time.mktime(o))\n if isinstance(o, decimal.Decimal): return str(o)\n return JSONEncoder_olddefault(self, o)\njson.JSONEncoder.default = JSONEncoder_newdefault\n\nMakes my life so much easier...\n", "For those who don't want to use a third-party library... An issue with Elias Zamaria's answer is that it converts to float, which can run into problems. For example:\n>>> json.dumps({'x': Decimal('0.0000001')}, cls=DecimalEncoder)\n'{\"x\": 1e-07}'\n>>> json.dumps({'x': Decimal('100000000000.01734')}, cls=DecimalEncoder)\n'{\"x\": 100000000000.01733}'\n\nThe JSONEncoder.encode() method lets you return the literal json content, unlike JSONEncoder.default(), which has you return a json compatible type (like float) that then gets encoded in the normal way. The problem with encode() is that it (normally) only works at the top level. But it's still usable, with a little extra work (python 3.x):\nimport json\nfrom collections.abc import Mapping, Iterable\nfrom decimal import Decimal\n\nclass DecimalEncoder(json.JSONEncoder):\n def encode(self, obj):\n if isinstance(obj, Mapping):\n return '{' + ', '.join(f'{self.encode(k)}: {self.encode(v)}' for (k, v) in obj.items()) + '}'\n if isinstance(obj, Iterable) and (not isinstance(obj, str)):\n return '[' + ', '.join(map(self.encode, obj)) + ']'\n if isinstance(obj, Decimal):\n return f'{obj.normalize():f}' # using normalize() gets rid of trailing 0s, using ':f' prevents scientific notation\n return super().encode(obj)\n\nWhich gives you:\n>>> json.dumps({'x': Decimal('0.0000001')}, cls=DecimalEncoder)\n'{\"x\": 0.0000001}'\n>>> json.dumps({'x': Decimal('100000000000.01734')}, cls=DecimalEncoder)\n'{\"x\": 100000000000.01734}'\n\n", "From the JSON Standard Document, as linked in json.org:\n\nJSON is agnostic about the semantics of numbers. In any programming language, there can be a variety of\n number types of various capacities and complements, fixed or floating, binary or decimal. That can make\n interchange between different programming languages difficult. JSON instead offers only the representation of\n numbers that humans use: a sequence of digits. All programming languages know how to make sense of digit\n sequences even if they disagree on internal representations. That is enough to allow interchange.\n\nSo it's actually accurate to represent Decimals as numbers (rather than strings) in JSON. Bellow lies a possible solution to the problem.\nDefine a custom JSON encoder:\nimport json\n\n\nclass CustomJsonEncoder(json.JSONEncoder):\n\n def default(self, obj):\n if isinstance(obj, Decimal):\n return float(obj)\n return super(CustomJsonEncoder, self).default(obj)\n\nThen use it when serializing your data:\njson.dumps(data, cls=CustomJsonEncoder)\n\nAs noted from comments on the other answers, older versions of python might mess up the representation when converting to float, but that's not the case anymore.\nTo get the decimal back in Python:\nDecimal(str(value))\n\nThis solution is hinted in Python 3.0 documentation on decimals:\n\nTo create a Decimal from a float, first convert it to a string.\n\n", "This is what I have, extracted from our class\nclass CommonJSONEncoder(json.JSONEncoder):\n\n \"\"\"\n Common JSON Encoder\n json.dumps(myString, cls=CommonJSONEncoder)\n \"\"\"\n\n def default(self, obj):\n\n if isinstance(obj, decimal.Decimal):\n return {'type{decimal}': str(obj)}\n\nclass CommonJSONDecoder(json.JSONDecoder):\n\n \"\"\"\n Common JSON Encoder\n json.loads(myString, cls=CommonJSONEncoder)\n \"\"\"\n\n @classmethod\n def object_hook(cls, obj):\n for key in obj:\n if isinstance(key, six.string_types):\n if 'type{decimal}' == key:\n try:\n return decimal.Decimal(obj[key])\n except:\n pass\n\n def __init__(self, **kwargs):\n kwargs['object_hook'] = self.object_hook\n super(CommonJSONDecoder, self).__init__(**kwargs)\n\nWhich passes unittest:\ndef test_encode_and_decode_decimal(self):\n obj = Decimal('1.11')\n result = json.dumps(obj, cls=CommonJSONEncoder)\n self.assertTrue('type{decimal}' in result)\n new_obj = json.loads(result, cls=CommonJSONDecoder)\n self.assertEqual(new_obj, obj)\n\n obj = {'test': Decimal('1.11')}\n result = json.dumps(obj, cls=CommonJSONEncoder)\n self.assertTrue('type{decimal}' in result)\n new_obj = json.loads(result, cls=CommonJSONDecoder)\n self.assertEqual(new_obj, obj)\n\n obj = {'test': {'abc': Decimal('1.11')}}\n result = json.dumps(obj, cls=CommonJSONEncoder)\n self.assertTrue('type{decimal}' in result)\n new_obj = json.loads(result, cls=CommonJSONDecoder)\n self.assertEqual(new_obj, obj)\n\n", "You can create a custom JSON encoder as per your requirement.\nimport json\nfrom datetime import datetime, date\nfrom time import time, struct_time, mktime\nimport decimal\n\nclass CustomJSONEncoder(json.JSONEncoder):\n def default(self, o):\n if isinstance(o, datetime):\n return str(o)\n if isinstance(o, date):\n return str(o)\n if isinstance(o, decimal.Decimal):\n return float(o)\n if isinstance(o, struct_time):\n return datetime.fromtimestamp(mktime(o))\n # Any other serializer if needed\n return super(CustomJSONEncoder, self).default(o)\n\nThe Decoder can be called like this,\nimport json\nfrom decimal import Decimal\njson.dumps({'x': Decimal('3.9')}, cls=CustomJSONEncoder)\n\nand the output will be:\n>>'{\"x\": 3.9}'\n\n", "Based on stdOrgnlDave answer I have defined this wrapper that it can be called with optional kinds so the encoder will work only for certain kinds inside your projects. I believe the work should be done inside your code and not to use this \"default\" encoder since \"it is better explicit than implicit\", but I understand using this will save some of your time. :-)\nimport time\nimport json\nimport decimal\nfrom uuid import UUID\nfrom datetime import datetime\n\ndef JSONEncoder_newdefault(kind=['uuid', 'datetime', 'time', 'decimal']):\n '''\n JSON Encoder newdfeault is a wrapper capable of encoding several kinds\n Use it anywhere on your code to make the full system to work with this defaults:\n JSONEncoder_newdefault() # for everything\n JSONEncoder_newdefault(['decimal']) # only for Decimal\n '''\n JSONEncoder_olddefault = json.JSONEncoder.default\n\n def JSONEncoder_wrapped(self, o):\n '''\n json.JSONEncoder.default = JSONEncoder_newdefault\n '''\n if ('uuid' in kind) and isinstance(o, uuid.UUID):\n return str(o)\n if ('datetime' in kind) and isinstance(o, datetime):\n return str(o)\n if ('time' in kind) and isinstance(o, time.struct_time):\n return datetime.fromtimestamp(time.mktime(o))\n if ('decimal' in kind) and isinstance(o, decimal.Decimal):\n return str(o)\n return JSONEncoder_olddefault(self, o)\n json.JSONEncoder.default = JSONEncoder_wrapped\n\n# Example\nif __name__ == '__main__':\n JSONEncoder_newdefault()\n\n", "If someone is still looking for the answer, it is most probably you have a 'NaN' in your data that you are trying to encode. Because NaN is considered as float by Python.\n", "If you want to pass a dictionary containing decimals to the requests library (using the json keyword argument), you simply need to install simplejson:\n$ pip3 install simplejson \n$ python3\n>>> import requests\n>>> from decimal import Decimal\n>>> # This won't error out:\n>>> requests.post('https://www.google.com', json={'foo': Decimal('1.23')})\n\nThe reason of the problem is that requests uses simplejson only if it is present, and falls back to the built-in json if it is not installed.\n", "For anybody that wants a quick solution here is how I removed Decimal from my queries in Django\ntotal_development_cost_var = process_assumption_objects.values('total_development_cost').aggregate(sum_dev = Sum('total_development_cost', output_field=FloatField()))\ntotal_development_cost_var = list(total_development_cost_var.values())\n\n\nStep 1: use , output_field=FloatField() in you r query\nStep 2: use list eg list(total_development_cost_var.values())\n\nHope it helps\n", "This question is old, but there seems to be a better and much simpler solution in Python3 for most use-cases:\nnumber = Decimal(0.55)\nconverted_number = float(number) # Returns: 0.55 (as type float)\n\nYou can just convert Decimal to float.\n", "My 2 cents for easy solution, if you're sure Decimal is the only bad guy on your json dumps method:\nprint(json.loads(json.dumps({\n 'a': Decimal(1230),\n 'b': Decimal(11111111123.22),\n}, default=lambda x: eval(str(x)))))\n\n>>> {'a': 1230, 'b': 11111111123.22}\n\nThe \"smart\" thing here is using default to convert Decimal to int or float, automatically, taking advantage of eval function: default=lambda x: eval(str(x))\nBut always be careful using eval on your code as it can lead to security issues ;)\n", "Decimal is not suitable to be converted through:\n\nfloat due to precision problems\nstr due to openapi restrictions\n\nWe still need direct decimal to a number json serialisation.\nHere is our extension of @tesdal 's fakefloat solution (closed in v3.5.2rc1).\nIt uses fakestr + monkeypatching to avoid quotation and \"floatation\" of decimals.\nimport json.encoder\nfrom decimal import Decimal\n\n\ndef encode_fakestr(func):\n def wrap(s):\n if isinstance(s, fakestr):\n return repr(s)\n return func(s)\n return wrap\n\n\njson.encoder.encode_basestring = encode_fakestr(json.encoder.encode_basestring)\njson.encoder.encode_basestring_ascii = encode_fakestr(json.encoder.encode_basestring_ascii)\n\n\nclass fakestr(str):\n def __init__(self, value):\n self._value = value\n def __repr__(self):\n return str(self._value)\n\n\nclass DecimalJsonEncoder(json.encoder.JSONEncoder):\n def default(self, o):\n if isinstance(o, Decimal):\n return fakestr(o)\n return super().default(o)\n\n\njson.dumps([Decimal('1.1')], cls=DecimalJsonEncoder)\n\n[1.1]\n\nI don't understand why python developers force us using floats in places where it is not suitable.\n", "I will share what worked for me with flask 2.1.0\nWhen I was creating the dictionary which had to be used from jsonify I used rounding:\njson_dict['price'] = round(self.price, ndigits=2) if self.price else 0\n\nSo this way I could return D.DD number or 0 without using some global configuration. And this is nice because some Decimals has to be bigger, like latitude and longitude coordinates.\nreturn jsonify(json_dict)\n\n" ]

[ 265, 239, 178, 64, 57, 38, 28, 14, 14, 13, 11, 7, 4, 2, 1, 0, 0, 0, 0, 0, 0 ]

[]

[]

[ "decimal", "floating_point", "json", "python" ]

[ "x_out has a single dimension after the squeeze operation (I assume, hard to say from your code). Try printing x_out.shape to check. In this case, the solution would be torch.cat(x_out,dim = 0).\nYou may save yourself some code comprehension headaches by not performing the squeeze operation until the end. For instance, if each output i in x_out has some dimension [1,d_i], it's straightforward to see that you want to concatenate along dimension 1. Then you can squeeze x_out at the very end.\n" ]

[ 0 ]

[]

[]

[ "machine_learning", "python", "pytorch" ]

[ "Try omitting the -t from ping when running the script on your phone.\nping works differently on your phone. On windows you have to add -t for it to ping nonstop, on other operating systems that is the default behaviour.\n" ]

[ 0 ]

[]

[]

[ "pydroid", "python" ]

[ "it's more like, writing \\n a new line character after each json. so each line is JSON. somehow the link jsonlines throw server not found error on me.\nyou can have these options:\n\nwrite \\n after each line:\n\nimport json\nwith open(\"sample2_op1.json\", \"w\") as outfile:\n for e_json in dictionary:\n json.dump(e_json, outfile)\n outfile.write('\\n')\n#read file, as it has \\n, read line by line and load as json\nwith open(\"sample2_op1.json\",\"r\") as file:\n for line in file:\n print(json.loads(line),type(json.loads(line)))\n\n\nwhich have way to read file too, its jsonlines\ninstall the module !pip install jsonlines\n\nimport jsonlines\n#write to file\nwith jsonlines.open('sample2_op2.jsonl', 'w') as outfile:\n outfile.write_all(dictionary)\n#read the file\nwith jsonlines.open('sample2_op2.jsonl') as reader:\n for obj in reader:\n print(obj)\n\n" ]

[ 1 ]

[]

[]

[ "gpt_3", "nlp", "python" ]

[ "the problem here come from indexes try reseting index before adding like this:\nfb_us_data = us_data[us_data.app_id == fb_key].copy()\nfb_ww_data = ww_data[ww_data.app_id == fb_key].copy()\nfb_us_data.reset_index(drop=True, inplace=True)\nfb_ww_data.reset_index(drop=True, inplace=True)\n\n" ]

[ 0 ]

[]

[]

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

[ "You did set up a local tile server, not a geocoder. To perform local bulk geocoding you need a local geocoding server. For an OSM-based geocoder take a look at Nominatim.\n" ]

[ 0 ]

[]

[]

[ "docker", "geocode", "local", "openstreetmap", "python" ]

[ "Try this!\nbl = [\"-\", \"1)\", \"2)\"]\nstring = \"-2)Hello1) World!\"\n\nfor item in bl:\n string = string.replace(item, \"\")\n\nprint(string)\n\n" ]

[ -1 ]

[]

[]

[ "python", "string" ]

[ "Using a service name for testing purposes is not supported at the moment.\nI suspect this is why python manage.py test does not work.\nSee the ticket here https://code.djangoproject.com/ticket/33685\nIf that is the cause of your issue, python manage.py runserver should work.\nIf you need to use tests, try the old config.\n" ]

[ 0 ]

[]

[]

[ "django", "docker", "postgresql", "python", "python_3.x" ]

[ "One solution would be to not use the pivot_table function and instead use the pivot function, which allows you to specify the aggregation function. Instead of using the default mean function, you can use the function that simply concatenates the values together, such as the sum function.\nHere is an example:\ntrainingload = trainingload.pivot(index=('Date', 'About'), columns='NHV Training', values=['Duration', 'sRPE Cardio', 'sRPE Biomechanical'], aggfunc=sum)\n\nThis will result in the values for Duration, sRPE Cardio, and sRPE Biomechanical being concatenated together for each person and each day, rather than being averaged.\n" ]

[ 0 ]

[]

[]

[ "dataframe", "pivot", "python" ]

[ "As it seems cases accept a \"guard\" clause starting with Python 3.10, which you can use for this purpose:\nmatch x:\n case w if w in a:\n # this was the \"case in a\" in the question\n case w if w in b:\n # this was the \"case in b\" in the question\n ...\n\nthe w here actually captures the value of x, part of the syntax here too, but it's more useful in some other fancy cases listed on the linked whatsnew page.\n" ]

[ 2 ]

[]

[]

[ "python", "python_3.x" ]

[ "I think this will help you a lot (tweaks are on you) but it doesn't write in a CSV, I leave that for you to try, just take into account that there are thousands of possible combinations and in some cases, the results are really huge (see comments in main()).\nI've added missing types in function declarations for clarity and successful linting (compatible with older Python versions).\nAlso, I think that the function combinations() is not needed so I removed it.\nIn my proposed code, the function solve() is the one doing the magic :)\nSaid all that, here's the full code:\nimport numpy as np\nfrom itertools import product\nfrom math import factorial\nfrom typing import Union, Callable, Tuple, List, Set\n\n\ndef plus(a: int, b: int) -> int:\n return a + b\n\n\ndef minus(a: int, b: int) -> int:\n return a - b\n\n\ndef mult(a: int, b: int) -> int:\n return a * b\n\n\ndef div(a: int, b: int) -> Union[int, float]:\n try:\n retval = int(a / b)\n except (ValueError, ZeroDivisionError):\n retval = np.nan\n return retval\n\n\ndef the_factorial(a: int) -> Union[int, float]:\n try:\n return factorial(int(a))\n except ValueError:\n return np.nan\n except OverflowError:\n return np.inf\n\n\ndef power_to(a: int, b: int) -> Union[int, float]:\n try:\n return int(a ** b)\n except (ValueError, ZeroDivisionError):\n return np.nan\n\n\ndef root(a: int, b: int) -> Union[int, float]:\n try:\n return int(b ** (1 / a))\n except (TypeError, ZeroDivisionError, ValueError):\n return np.nan\n\n\ndef solve(values: Tuple[int, int, int], ops: List[Callable]) -> list[Tuple[str, int]]:\n # Iterate over available functions.\n combs = list()\n for f in FACTORS:\n # Get values to operate with.\n x, y, z = values\n sx, sy, sz = x, y, z\n a, b, c = f\n # Calculate the factorial for the values (if applicable).\n if a == 1:\n sx = f\"{x}!\"\n x = the_factorial(x)\n if b == 1:\n sy = f\"{y}!\"\n y = the_factorial(y)\n if c == 1:\n sz = f\"{z}!\"\n z = the_factorial(z)\n for ext_op in ops: # External operation.\n for int_op in ops: # Internal operation.\n # Create equations by grouping the first 2 elements, e.g.: ((x + y) * z).\n eq_str = f\"{ext_op.__name__}({int_op.__name__}({sx}, {sy}), {sz})\"\n eq_val = ext_op(int_op(x, y), z)\n combs.append((eq_str, eq_val))\n # Create equations by grouping the last 2 elements, e.g.: (x + (y * z)).\n eq_str = f\"{ext_op.__name__}({sx}, {int_op.__name__}({sy}, {sz}))\"\n eq_val = ext_op(x, int_op(y, z))\n combs.append((eq_str, eq_val))\n return combs\n\n\ndef inverter(data: List[int]) -> List[Tuple[int, int, int]]:\n inverted_data = [-x for x in data]\n res = list(product(*zip(data, inverted_data)))\n return res\n\n\n# Data to process.\nINITIAL_DATA: List[str] = [\n \"518-2\",\n '533-3',\n # '534-0',\n # '000-3',\n # '000-4'\n]\n# Available functions.\nFUNCTIONS: List[Callable] = [ # the_factorial() removed, see solve().\n plus,\n minus,\n mult,\n div,\n power_to,\n root\n]\n# Get posible combinations to apply the factor operation.\nFACTORS: Set[Tuple] = set(product([1, 0, 0], repeat=3))\n\n\ndef main():\n cases = 0 # Count all possible cases (for each input value).\n data = list() # List with all final data to be dumped in CSV.\n print(\"number, solution, number_of_solutions\")\n # Iterate over all initial data.\n for eq in INITIAL_DATA:\n # Get values before and after the hyphen.\n nums, res = eq.split('-')\n res = int(res)\n # Get combinations with inverted values.\n combs = inverter([int(n) for n in list(nums)])\n # Iterate over combinations and generate a list with their many possible solutions.\n sol_cnt = 0 # Number of solutions (for each input value).\n solutions = list() # List with all final data to be dumped in CSV.\n for i in [solve(i, FUNCTIONS) for i in combs]:\n for j in i:\n str_repr, value = j\n # Some values exceed the 4300 digits, hence the 'try-catch'.\n # The function 'sys.set_int_max_str_digits()' may be used instead to increase the str() capabilites.\n try:\n str(value)\n except ValueError:\n value = np.inf\n if value == res:\n sol_cnt += 1\n solutions.append((eq, str_repr, value))\n cases += 1\n # Iterate over all data gathered, and add number of solutions.\n for i in range(len(solutions)):\n eq, str_repr, value = solutions[i]\n solutions[i] += (sol_cnt,)\n print(f\"{eq}, {str_repr} = {value}, {sol_cnt}\")\n data.extend(solutions)\n # Print all the solutions for this input.\n print(f\"\\nThese are the {sol_cnt} solutions for input {eq}:\")\n solutions = [s for s in solutions if (type(s[2]) is int and s[2] == res)]\n for i in range(len(solutions)):\n print(f\" {i:4}. {solutions[i][1]}\")\n print()\n print(f\"\\nTotal cases: {cases}\")\n\nAnd for the output, note that solutions are printed/formatted using the name of your functions, not mathematical operators. This is just an excerpt of the output generated for the first value in initial_data using factorials in the 1st and 3rd digits:\nnumber, solution, number_of_solutions\n518-2, plus(plus(5!, 1), 8!) = 40441, 12\n518-2, plus(5!, plus(1, 8!)) = 40441, 12 \n518-2, plus(minus(5!, 1), 8!) = 40439, 12 \n518-2, plus(5!, minus(1, 8!)) = -40199, 12 \n518-2, plus(mult(5!, 1), 8!) = 40440, 12 \n518-2, plus(5!, mult(1, 8!)) = 40440, 12 \n518-2, plus(div(5!, 1), 8!) = 40440, 12 \n518-2, plus(5!, div(1, 8!)) = 120, 12\n518-2, plus(power_to(5!, 1), 8!) = 40440, 12 \n518-2, plus(5!, power_to(1, 8!)) = 121, 12 \n518-2, plus(root(5!, 1), 8!) = 40321, 12 \n518-2, plus(5!, root(1, 8!)) = 40440, 12\n\n...\n\nThese are the 12 solutions for input 518-2:\n 0. plus(minus(-5, 1!), 8)\n 1. minus(-5, minus(1!, 8))\n 2. plus(minus(-5, 1), 8)\n 3. minus(-5, minus(1, 8))\n 4. minus(-5, plus(1!, -8))\n 5. minus(minus(-5, 1!), -8)\n 6. minus(-5, plus(1, -8))\n 7. minus(minus(-5, 1), -8)\n 8. plus(plus(-5, -1), 8)\n 9. plus(-5, plus(-1, 8))\n 10. plus(-5, minus(-1, -8))\n 11. minus(plus(-5, -1), -8)\n\nTotal cases: 4608\n\nNote that 4608 cases were processed just for the first value in initial_data, so I recommend you to try with this one first and then add the rest, as for some cases it could take a lot of processing time.\nAlso, I noticed that you are truncating the values in div() and root() so bear it in mind. You will see lots of nan and inf in the full output because there are huge values and conditions like div/0, so it's expected.\n" ]

[ 0 ]

[]

[]

[ "functools", "numpy", "python" ]

[ "You can use set_title() and set_text() to set the names of the legend title & labels. Similarly, use plt.setp() to change the font to the size you need it to be... an example is shown below.\npenguins = sns.load_dataset(\"penguins\")\ng=sns.pairplot(penguins, hue=\"species\")\n\ng._legend.set_title(\"New Title\") ## Change text of Title\nnew_labels = ['Label 1', 'Label 2', 'Label 3']\nfor t, l in zip(g._legend.texts, new_labels):\n t.set_text(l) ## Change text of labels\n \nplt.setp(g._legend.get_title(), fontsize=30) ## Set the Title font to 30\nplt.setp(g._legend.get_texts(), fontsize=20) ## Set the label font to 20\n\nplt.show()\n\n\n" ]

[ 0 ]

[]

[]

[ "matplotlib", "python", "seaborn" ]

[ "I think that you need to pass a string in the y.evalf(subs={'x': x_i}) part of your code.\n" ]

[ 1 ]

[]

[]

[ "nameerror", "python" ]

[ "The purpose of join, is to put the element between each\nreduce(lambda x, y: x + \"\\n\" + y, [\"dog\", \"cat\"])\n\n", "###################### METHOD 1 ######################\n\nstrings = [\"This\", \"is\", \"a\", \"list\", \"of\", \"strings\"]\n\n# join the strings using lambda\njoined = lambda strings: \"\\n\".join(strings)\nprint(joined(strings))\n\n###################### METHOD 2 ######################\n\nmylist = [\"a\", \"b\", \"c\", \"d\", \"e\"]\n\n# use list comprehension to join the list of strings\nmystring = \"\\n\".join([str(x) for x in mylist])\nprint(mystring)\n\nstrings = [\"This\", \"is\", \"a\", \"list\", \"of\", \"strings\"]\n\n###################### METHOD 3 ######################\nimport functools\nlist_of_strings = [\"a\", \"b\", \"c\"]\nprint(functools.reduce(lambda x, y: x + \"\\n\" + y, list_of_strings))\n\n" ]

[ 2, 0 ]

[]

[]

[ "python" ]

[ "This can be achieved simply by manipulating the rendered html with a simple regular expression:\nimport re\n\ndf = pd.DataFrame(1, index=[1, 2], columns=list('AB'))\n\nhtml = df.to_html(classes=\"table\")\nhtml = re.sub(\n r'<table([^>]*)>',\n r'<table\\1 attribute=\"value\" attribute2=\"value2\">',\n html\n)\n\nprint(html.split('\\n')[0])\n\n<table border=\"1\" class=\"dataframe table\" attribute=\"value\" attribute2=\"value2\">\n\n", "For a pandas-only solution, you can use a Styler object:\ndf = pd.DataFrame(1, index=[1, 2], columns=list('AB'))\nstyled = df.style.set_table_attributes('foo=\"foo\" bar=\"bar\"')\nprint(styled.to_html())\n\nOutput:\n<style type=\"text/css\">\n</style>\n<table id=\"T_7d7d8\" foo=\"foo\" bar=\"bar\">\n <thead>\n <tr>\n <th class=\"blank level0\" >&nbsp;</th>\n <th id=\"T_7d7d8_level0_col0\" class=\"col_heading level0 col0\" >A</th>\n <th id=\"T_7d7d8_level0_col1\" class=\"col_heading level0 col1\" >B</th>\n </tr>\n </thead>\n <tbody>\n <tr>\n <th id=\"T_7d7d8_level0_row0\" class=\"row_heading level0 row0\" >1</th>\n <td id=\"T_7d7d8_row0_col0\" class=\"data row0 col0\" >1</td>\n <td id=\"T_7d7d8_row0_col1\" class=\"data row0 col1\" >1</td>\n </tr>\n <tr>\n <th id=\"T_7d7d8_level0_row1\" class=\"row_heading level0 row1\" >2</th>\n <td id=\"T_7d7d8_row1_col0\" class=\"data row1 col0\" >1</td>\n <td id=\"T_7d7d8_row1_col1\" class=\"data row1 col1\" >1</td>\n </tr>\n </tbody>\n</table>\n\n" ]

[ 6, 0 ]

[]

[]

[ "html", "pandas", "python" ]

[ "Using a kernel_initializer of tf.zeros_initializer() for your dense layers in the actor and critic networks can lead to the issue you are experiencing, where the loss becomes NaN and the model repeats the same action. This is because using a kernel_initializer of tf.zeros_initializer() initializes all of the weights in the dense layers to zeros, which can prevent the network from learning.\nIn general, it is better to use a different kernel_initializer for your dense layers, such as tf.random_normal_initializer() or tf.glorot_uniform_initializer(). These initializers initialize the weights with random values, which allows the network to learn and produce more diverse outputs.\nTo fix the issue with your model, you can try changing the kernel_initializer for your dense layers to a different value, such as tf.random_normal_initializer() or tf.glorot_uniform_initializer(). This should allow your network to learn and avoid the issue where the loss becomes NaN and the model repeats the same action.\nYou can also try using a different optimizer, such as RMSProp or Adagrad, which may be better suited for this problem. Additionally, you can try adjusting the learning rate and other hyperparameters of the model to see if that improves its performance.\nIf the tf.zeros_initializer initializer is the only initializer that works for your network, but the performance is not good, there are several steps you can take to improve the performance of your network.\nFirst, you can try adjusting the parameters of the tf.zeros_initializer initializer to fine-tune the starting weights for your network. The tf.zeros_initializer initializer does not have any parameters, so you will need to use a different initializer and adjust its parameters to control the starting weights for your network.\nFor example, you can try using the tf.random_normal_initializer initializer, which will provide random starting weights for the network. You can adjust the mean and stddev parameters to control the distribution of the starting weights, and experiment with different values to see which provides the best performance for your network.\nAlternatively, you can try adjusting other hyperparameters, such as the learning rate or the optimizer, to improve the performance of your network. For example, you can try using a different optimizer, such as the Adam optimizer or the RMSprop optimizer, to see if it provides better performance for your network.\nYou can also try modifying the state, action, and reward definitions for your network to see if a different representation improves the performance of your network. For example, you can try using a different state representation, such as a different set of features or a different scaling or normalization method, to see if it improves the performance of your network.\nFinally, you can try using more data or more complex network architectures to improve the performance of your network. For example, you can try using a larger dataset, or a deeper or wider network, to see if it provides better performance for your network. For more information, see the TensorFlow documentation on training and evaluating neural networks. https://www.tensorflow.org/guide/keras/train_and_evaluate\n" ]

[ 0 ]

[]

[]

[ "actor_critics", "python", "random", "tensorflow" ]

[ "Some files init.py are missing\nI think that the problem is the missing of file __init__.py in your subfolders. Try to add this empty file in all your subfolders as I show you below:\ntest_lcm\n├── __init__.py\n├── src\n| └── __init__py\n│ └── arithmetics.py\n└── test\n └── __init__py\n └── lcm\n ├── __init__.py\n ├── test_lcm_exception.py\n └── test_lcm.py\n\nIf you see my tree folder I have create a folder test_lcm as root of the tree. You have to place inside it by cd command.\nSo in test_lcm folder execute:\n# go to test_lcm folder\ncd ~/test_lcm\n\n# execute test\npython3 -m unittest discover\n\nThe last part of the output is:\n----------------------------------------------------------------------\nRan 6 tests in 0.002s\n\nFAILED (failures=1, errors=2)\n\nThis show that are executed 6 tests with 2 errors (test_lcm_better_exception and test_lcm_exception fail).\n" ]

[ 1 ]

[]

[]

[ "python", "python_3.x", "python_unittest", "unit_testing" ]

[ "The issue was indeed that the data could only be accessed after logging in.\nI have managed to resolve it using Selenium and this answer.\nfrom io import StringIO \nimport pandas as pd\nimport requests\nfrom selenium import webdriver\n\n#start requests session with login from selenium driver\ns = requests.Session()\nselenium_user_agent = driver.execute_script(\"return navigator.userAgent;\")\ns.headers.update({\"user-agent\": selenium_user_agent})\n\n#copy cookies from selenium driver\nfor cookie in driver.get_cookies():\n s.cookies.set(cookie['name'], cookie['value'], domain=cookie['domain'])\n\n#read csv\nresponse = s.get(url)\nif response.ok:\n data = response.content.decode('utf8') \n df = pd.read_csv(StringIO(data))\n \n \n\n" ]

[ 0 ]

[]

[]

[ "csv", "pandas", "python" ]

[ "Weidong Zhu Robot!\nYou probably forgot to make function definition. Also @client.command() returns context, not message, so don't forget about this too.\nThere's your quick fix\[email protected]()\[email protected](1,30,commands.BucketType.user)\nasync def hunt_cmd(ctx):\n message = ctx.message\n if message.content.startswith('!sg hunt')\n await message.channel.send('You hunted a...')```\n\n" ]

[ 0 ]

[]

[]

[ "bots", "discord", "discord.py", "python" ]

[ "There's nothing easier and clearer than just running JavaScript.\nOpen new tab:\ndriver.execute_script(\"window.open('');\")\n", "open a new tab:\nbrowser.get('http://www.google.com')\n\nclose a tab:\nbrowser.close()\n\nswitch to a tab:\nbrowser.swith_to_window(window_name)\n\n", "You can choose which window you want to close：\nwindow_name = browser.window_handles[0]\n\nSwitch window：\nbrowser.switch_to.window(window_name=window_name)\n\nThen close it：\nbrowser.close()\n\n", "Just to combine the answers above for someone still curious. The below is based on Python 2.7 and a driver in Chrome.\nOpen new tab by: driver.execute_script(\"window.open('\"+URL+\"', '__blank__');\")\nwhere URL is a string such as \"http://www.google.com\".\nClose tab by:\ndriver.close() [Note, this also doubles as driver.quit() when you only have 1 tab open].\nNavigate between tabs by: driver.switch_to_window(driver.window_handles[0])\nand driver.switch_to_window(driver.window_handles[1]).\n", "Open new tab:\nbrowser.execute_script(\"window.open('\"+your url+\"', '_blank')\")\n\nSwitch to new tab:\nbrowser.switch_to.window(windows[1])\n\n", "IMHO all the above answers didn't exactly solve the original problem of closing a tab in a window and not closing the entire window or opening a blank tab.\nmy solution:\nbrowser.switch_to.window(\"tab1\") #change the tab1 to the id of the tab you want to close\nbrowser.execute_script(\"window.close('','_parent','');\")\n\n" ]

[ 14, 12, 11, 6, 2, 0 ]

[]

[]

[ "macos", "python", "selenium" ]

[ "Slice with boolean indexing before getting the min/max:\nB = A[A>1e-12]\nMin = np.min(B)\nMax = np.max(B)\nprint(Min, Max)\n\nOutput: 1.02453254e-06 9.49108487e-05\nB: array([9.49108487e-05, 1.02453254e-06, 1.02453254e-06])\n", "You can just select the values of the array greater than 1e-12 first and obtain the min and max of that:\n>>> A[A > 1e-12].min()\n1.02453254e-06\n>>> A[A > 1e-12].max()\n9.49108487e-05\n\n", "arr = np.array([9.49108487e-05,1.05634586e-19,5.68676707e-17,1.02453254e-06,2.48792902e-16,1.02453254e-06]) \nmask = arr > 1e-12 \nMin = np.min(arr[mask]) \nMax = np.max(arr[mask])\n\n" ]

[ 3, 2, 1 ]

[]

[]

[ "numpy", "python" ]

[ "The issue you have here is that your function getInputtemp is not getting fired. It only gets fired when button save_btn2 is clicked. Also, the if statement where the error is occuring will only get fired once. To fix this, you can either do what @Tkirishima have suggested.\nOr just move the if statement inside the getInputtemp function.\ndef getInputtemp():\n #global message \n #Then you would no longer need message as a global variable\n message = message_var2.get()\n message_var2.set(\"\")\n if message in ['1886', '2022']:\n page2(message)\n\nBut, if you do want the if statement outside the function (which I wouldn't recommend as stated earlier that it would execute when you start the script and never again):\ngetInputtemp() #The function is called to create message as global variable\nif message in ['1886', '2022']:\n page2(message)\n\n", "The problem that you have, is the usage of the global keyword.\nThe thing is that, you use global even tho the value doesn't exists in the global scope.\nIf you want your program to work, the best thing to do is to define message at the start of the program as None, with that process, message exists in the global scope with a value of None.\n...\nroot.geometry(\"600x400\")\nmessage_var2 = tk.StringVar()\nmessage = None # <<<<<<<<<<\n\ndef page2(message):\n print(f'test\\n{message}')\n...\n\nRelated:\n\nhttps://www.programiz.com/python-programming/global-keyword\nhttps://en.wikipedia.org/wiki/Scope_(computer_science)\n\n" ]

[ 2, 1 ]

[]

[]

[ "function", "global_variables", "python", "tkinter", "variables" ]

[ "If you're trying to avoid the overhead of creating an additional function, I'd say that list comprehensions are a Pythonic way of achieving what you need.\nAn approach like this would be idempotent, and elegant enough:\ndf[[col for col in df.columns if col != 'country']]\n\nThe upside with this method is that it's easily extensible if you want to drop a list of columns, by changing the != to a not in and passing your list of column names.\n" ]

[ 0 ]

[]

[]

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

[ "and simply returns either the first or the second operand, based on their truth value. If the first operand is considered false, it is returned, otherwise the other operand is returned.\nLists are considered true when not empty, so both lists are considered true. Their contents don't play a role here.\nBecause both lists are not empty, x and y simply returns the second list object; only if x was empty would it be returned instead:\n>>> [True, False] and ['foo', 'bar']\n['foo', 'bar']\n>>> [] and ['foo', 'bar']\n[]\n\nSee the Truth value testing section in the Python documentation:\n\nAny object can be tested for truth value, for use in an if or while condition or as operand of the Boolean operations below. The following values are considered false:\n[...]\n\nany empty sequence, for example, '', (), [].\n\n[...]\nAll other values are considered true — so objects of many types are always true.\n\n(emphasis mine), and the Boolean operations section right below that:\n\nx and y\n if x is false, then x, else y\nThis is a short-circuit operator, so it only evaluates the second argument if the first one is True.\n\nYou indeed need to test the values contained in the lists explicitly. You can do so with a list comprehension, as you discovered. You can rewrite it with the zip() function to pair up the values:\n[a and b for a, b in zip(x, y)]\n\n", "You could use numpy:\n>>> import numpy as np\n>>> x=np.array([True,True,False,False])\n>>> y=np.array([True,False,True,False])\n>>> x & y\narray([ True, False, False, False], dtype=bool)\n\nNumpy allows numerical and logical operations on arrays such as:\n>>> z=np.array([1,2,3,4])\n>>> z+1\narray([2, 3, 4, 5])\n\nYou can perform bitwise and with the & operator.\nInstead of a list comprehension, you can use numpy to generate the boolean array directly like so:\n>>> np.random.random(10)>.5\narray([ True, True, True, False, False, True, True, False, False, False], dtype=bool)\n\n", "Here is a simple solution:\nnp.logical_and(x,y)\n\n", "and is not necessarily a Boolean operator; it returns one of its two arguments, regardless of their type. If the first argument is false-ish (False, numeric zero, or an empty string/container), it returns that argument. Otherwise, it returns the second argument.\nIn your case, both x and y are non-empty lists, so the first argument is always true-ish, meaning x and y returns y and y and x returns x.\n", "This should do what you want:\nxy = [a and b for a, b in zip(x, y)]\n\nThe reason x and y returns y and y and x returns x is because boolean operators in python return the last value checked that determines the true-ness of the expression. Non-empty list's evaluate to True, and since and requires both operands to evaluate True, the last operand checked is the second operand. Contrast with x or y, which would return x because it doesn't need to check y to determine the true-ness of the expression.\n", "To generalize on the zip approach, use all and any for any number of lists.\nall for AND:\n[all(i) for i in zip(a, b, c)] # zip all lists\n\nand any for OR:\n[any(i) for i in zip(a, b, c)]\n\n", "Instead of using\n[a and b for a, b in zip(x, y)]\n\none could just use the possibility of numpy to multiply bool-values:\n(np.array(x)*np.array(y))\n>> array([ True, False, False, False], dtype=bool)\n\nOr do I overlook a special case?\n", "You can use the zip function\nx=[True,True,False,False]\ny=[True,False,True,False]\nz=[a and b for a,b in zip(x,y)]\n\n", "In addition to what @Martijn Pieters has answered, I would just add the following code to explain and and or operations in action.\nand returns the first falsy value encountered else the last evaluated argument. \nSimilarly or returns the first truthy value encountered else the last evaluated argument. \nnl1 = [3,3,3,3,0,0,0,0]\nnl2 = [2,2,0,0,2,2,0,0]\nnl3 = [1,0,1,0,1,0,1,0]\nand_list = [a and b and c for a,b,c in zip(nl1,nl2,nl3)]\nor_list = [a or b or c for a,b,c in zip(nl1,nl2,nl3)]\n\nValues are\nand_list = [1, 0, 0, 0, 0, 0, 0, 0]\nor_list = [3, 3, 3, 3, 2, 2, 1, 0]\n", "Thanks for the answer @Martijn Pieters and @Tony.\nI dig into the timing of the various options we have to make the AND of two lists and I would like to share my results, because I found them interesting.\nDespite liking a lot the pythonic way [a and b for a,b in zip(x,y) ], turns out really slow.\nI compare with a integer product of arrays (1*(array of bool)) * (1*(array of bool)) and it turns out to be more than 10x faster\nimport time\nimport numpy as np\narray_to_filter = np.linspace(1,1000000,1000000) # 1 million of integers :-)\nvalue_limit = 100\ncycles = 100\n\n# METHOD #1: [a and b for a,b in zip(x,y) ]\nt0=time.clock()\nfor jj in range(cycles):\n x = array_to_filter<np.max(array_to_filter)-value_limit # filter the values > MAX-value_limit\n y = array_to_filter>value_limit # filter the values < value_limit\n z= [a and b for a,b in zip(x,y) ] # AND\n filtered = array_to_filter[z]\nprint('METHOD #1 = %.2f s' % ( (time.clock()-t0)))\n\n\n\n# METHOD 1*(array of bool) AND 1*(array of bool)\nt0=time.clock()\nfor jj in range(cycles):\n x = 1*(array_to_filter<np.max(array_to_filter)-value_limit) # filter the values > MAX-value_limit\n y = 1*(array_to_filter>value_limit) # filter the values < value_limit\n z = x*y # AND\n z = z.astype(bool) # convert back to array of bool\n filtered = array_to_filter[z]\nprint('METHOD #2 = %.2f s' % ( (time.clock()-t0)))\n\nThe results are\nMETHOD #1 = 15.36 s\nMETHOD #2 = 1.85 s\n\nThe speed is almost affected equally by the size of the array or by the number of cycles.\nI hope I helped someone code to be faster. :-)\n" ]

[ 75, 20, 12, 7, 4, 2, 1, 0, 0, 0 ]

[ "The following works for me:\n([True,False,True]) and ([False,False,True])\n\noutput:\n[False, False, True]\n\n" ]

[ -1 ]

[ "boolean", "list", "operator_keyword", "python" ]

[ "You could do below:\na_positions = [idx + 1 for idx, el in enumerate(list1) if el == 'a']\na_repitition = len(a_positions)\n\nprint(a_positions):\n[1, 2, 4, 5, 7, 8, 10, 12, 13]\n\nprint(a_repitition):\n9\n\nIf you need repititions of each element you can also use collections.Counter\nfrom collections import Counter\ncounter = Counter(list1)\n\nprint(counter['a']):\n9\n\n", "If you want to get the indices and counts of all letters:\nlist1=['a','a','b','a','a','b','a','a','b','a','b','a','a']\npos = {}\nfor i,c in enumerate(list1, start=1): # 1-based indexing\n pos.setdefault(c, []).append(i)\npos\n# {'a': [1, 2, 4, 5, 7, 8, 10, 12, 13],\n# 'b': [3, 6, 9, 11]}\n\ncounts = {k: len(v) for k,v in pos.items()}\n# {'a': 9, 'b': 4}\n\n" ]

[ 2, 1 ]

[]

[]

[ "indexing", "python" ]

[ "import torch\nprint(torch.cat(tuple([torch.tensor(D[name]) for name in D.keys()]), dim=1))\n\nOutput:\ntensor([[0, 0, 0, 0, 0, 1, 1, 2, 2, 2, 2],\n [0, 0, 0, 0, 0, 1, 1, 2, 2, 2, 2],\n [0, 0, 0, 0, 0, 1, 1, 2, 2, 2, 2]])\n\n", "from itertools import chain \nl = []\nfor i in range(len(D)):\n t = [ D[k][i] for k in D ] \n l.append( list(chain.from_iterable(t)) )\n\nOutput:\n[[0, 0, 0, 0, 0, 1, 1, 2, 2, 2, 2],\n [0, 0, 0, 0, 0, 1, 1, 2, 2, 2, 2],\n [0, 0, 0, 0, 0, 1, 1, 2, 2, 2, 2]]\n\n" ]

[ 1, 0 ]

[]

[]

[ "dictionary", "python", "torch" ]

[ "This is the expected behavior I suppose. When the script exits with a non-zero exit code the StandardOut will not be returned. If you want to get the output irrespective of the status you can do something like this. The following will combine both STDOUT and STDERR and return while exiting the script with exit code 0. This will not move the execution to the catch block. So you will have to add a condition and check the returned message.\ntest_results = sh (\n script: \"python3 \\${WORKSPACE}/test.py notpass 2>&1 || echo \\\"status:\\$?\\\"; exit 0\",\n returnStdout: true\n )\n\n# Output\nTest results in passed test: tests did not pass\nDeployement interrupted by python.\nstatus:1\n\nAnother approach is to write the STDOUT to a file and read that in the catch block.\nstage('Test branch') {\n steps {\n script {\n test_results = 'position 1'\n try {\n test_results = sh (\n script: \"python3 \\${WORKSPACE}/test.py notpass > output\",\n returnStdout: true\n )\n echo \"Test results in passed test: ${test_results}\"\n } catch (err) {\n output = readFile(file: 'output')\n echo \"Test results in failed test numb 1: \" + output\n echo \"Test results in failed test numb 2: ${test_results}\"\n echo err.getMessage()\n println err.dump()\n }\n }\n }\n}\n\n" ]

[ 0 ]

[]

[]

[ "groovy", "jenkins", "jenkins_pipeline", "python", "try_catch" ]

[ "An alternative approach is to make a dictionary or dataframe in a format like CustomMonth | StartMonth | StartDay | EndMonth | EndDay\nwith one row for each of your custom months\nFrom there you can query your data using this\n", "As an option, I can offer if there is data for each month. Let's transform the 'period' column and create an auxiliary one with days.\nCheck if the day is greater than or equal to 10, then we copy the data from 'period' to the 'start' column. Next, fill in the gaps with the previous values. Create a 'finish' column using the 'start' column data + 30 days and change the number to 9.\nimport pandas as pd\n\ndf['period'] = df.date.dt.strftime('%Y-%m') \n\n\ndf['period'] = pd.to_datetime(df['period'])\ndf['day'] = df['date'].dt.day\nind_st = df['day'] >= 10\ndf.loc[ind_st, 'start'] = df.loc[ind_st, 'period'] + pd.to_timedelta(9, unit='D')\ndf['start'] = df['start'].fillna(method=\"ffill\")\ndf['finish'] = df['start'] + pd.to_timedelta(30, unit='D')\ndf['finish'] = df['finish'].apply(lambda dt: dt.replace(day=9))\n\n\nprint(df.loc[check_yes][['start', 'finish']].value_counts().sort_index())\n\nOutput\nstart finish \n2022-01-10 2022-02-09 1\n2022-02-10 2022-03-09 1\n2022-03-10 2022-04-09 1\n2022-05-10 2022-06-09 1\n2022-06-10 2022-07-09 1\n2022-08-10 2022-09-09 1\n2022-11-10 2022-12-09 2\n2022-12-10 2023-01-09 2\n\n" ]

[ 0, 0 ]

[]

[]

[ "pandas", "python" ]

[ "I managed to solve it like this:\ndf = df.withColumn(\n \"received\",\n SF.expr(\"\"\"transform(\n received, \n x -> struct(x.col1, x.col2, x.col3, x.col4, float(x.delay) as delay, x.col6))\"\"\"\n )\n )\n\n" ]

[ 0 ]

[]

[]

[ "pyspark", "python" ]

[ "open a shell in the workspace with ctrl-shift-s\non mac command-shift-s command prompt and run this command, it will install scikit\n\npip install scikit-learn\n\n" ]

[ 0 ]

[]

[]

[ "python", "replit", "replit_database" ]

[ "df3 = pd.concat([df1, df2[df2.id.isin(df1.id)]], axis = 0)\n" ]

[ 0 ]

[]

[]

[ "loops", "pandas", "python" ]

[ "Sry, I don't think dynamic choises are build into the discord api, but I usually use the following to add choises to a Slash command, maybee it can help you.\n@app_commands.choices(\nciclo=[\n app_commands.Choice(name=\"ASIX\", value=\"ASIX\"),\n app_commands.Choice(name=\"DAM\", value=\"DAM\")\n ],\nclase=[app_commands.Choice(name=\"A1\", value=\"A1\")]) \n\nI modified it to fit your example a bit to fit your example, but as said, as far as I know there is no way to make dynamic choices with the slash commands.\n", "This isn't possible from a Discord side of things. The choices have to be known beforehand and synced, so they can't dynamically change based on other values (also - you can fill them in in any order, so that wouldn't even work).\nYou'll have to do it some other way. You can't refer to the current value of that argument as the type of the Choice.\nOne option could be to use Views with Select menus, for example. A more hybrid approach could be to have a slash command with Choices, and let that one answer with a Select menu for the specific options they can choose depending on their argument. There's an example for Select menus: https://github.com/Rapptz/discord.py/blob/master/examples/views/dropdown.py\n" ]

[ 0, 0 ]

[]

[]

[ "discord", "discord.py", "python" ]

[ "If the idea is to delete the line with the given id, then gather up all of the non-matching lines (i.e. those that we want to keep) and then write them back to the original file.\ndef DelStudent(self, data):\n new_lines = []\n with open(data, \"r+\") as datafile:\n for line in datafile:\n datum = line.split()\n if datum[0] != id:\n new_lines.append(line)\n with open(data, \"w\") as datafile:\n for line in new_lines:\n datafile.write(line)\n\n" ]

[ 2 ]

[]

[]

[ "file", "python" ]

[ "Now resolved thanks to the owner of discord.py\n" ]

[ 0 ]

[]

[]

[ "bots", "discord", "discord.py", "python" ]

[ "Well, if we assume that your starting data has sort of groups of 3 rows of related data, we can maybe get away with adding a new grouping field to group them, and then we can pivot using that as well.\ndf['new_grouping_field'] = df.index // 3 # this gives the whole-number piece of the division,\n# so rows 0-2 will get a 0, rows 3-5 will get a 1, etc. Does require that the\n# index is numbers, which it should be\ndf.pivot(\n columns='GHG',\n values='GHG Conversion Factor 2022',\n index=['new_grouping_field','UOM','Unit']\n)\n\n" ]

[ 1 ]

[]

[]

[ "pandas", "python" ]

[ "What about :\nif tick == \"GOOG\"\n df.loc[:,\"CashFinancial\"] = None\n\nTo set an entire CashFinancial column to \"None\" only if your \"tick\" was GOOG, before writing it to csv.\nThe full code from the example you posted would he something like :\nimport pandas as pd\nfrom yahooquery import Ticker\nsymbols = ['AAPL','GOOG','MSFT']\nheader = [\"asOfDate\",\"CashAndCashEquivalents\",\"CashFinancial\",\"CurrentAssets\",\"TangibleBookValue\",\"CurrentLiabilities\",\"TotalLiabilitiesNetMinorityInterest\"]\n\nfor tick in symbols:\n faang = Ticker(tick)\n faang.balance_sheet(frequency='q')\n df = faang.balance_sheet(frequency='q')#,{\"symbol\":[1],\"asOfDate\":[2],\"CashAndCashEquivalents\":[3],\"CashFinancial\":[4],\"CurrentAssets\":[5],\"TangibleBookValue\":[6],\"CurrentLiabilities\":[7],\"TotalLiabilitiesNetMinorityInterest\":[8],\"marketCap\":[9]}\n for column_name in header :\n if not column_name in df.columns :\n #Here, if any column is missing from the names you defined \n #in your \"header\" variable, we add this column and set all \n #it's row values to None\n df.loc[:,column_name ] = None\n \n df.to_csv('output.csv', mode='a', index=True, header=False, columns=header)\n\n\n", "Load all dataframes into a list, then use pd.concat (it will create NaN in missing columns):\nimport pandas as pd\nfrom yahooquery import Ticker\n\nsymbols = [\"AAPL\", \"GOOG\", \"MSFT\"]\nheader = [\n \"asOfDate\",\n \"CashAndCashEquivalents\",\n \"CashFinancial\",\n \"CurrentAssets\",\n \"TangibleBookValue\",\n \"CurrentLiabilities\",\n \"TotalLiabilitiesNetMinorityInterest\",\n]\n\nall_dfs = []\nfor tick in symbols:\n faang = Ticker(tick)\n df = faang.balance_sheet(frequency=\"q\")\n all_dfs.append(df)\n\ndf = pd.concat(all_dfs)\n\nfor symbol, g in df.groupby(level=0):\n print(symbol)\n print(g[header])\n # to save to CSV:\n # g[header].to_csv('filename.csv')\n print(\"-\" * 80)\n\nPrints:\nAAPL\n asOfDate CashAndCashEquivalents CashFinancial CurrentAssets TangibleBookValue CurrentLiabilities TotalLiabilitiesNetMinorityInterest\nsymbol \nAAPL 2021-09-30 3.494000e+10 1.730500e+10 1.348360e+11 6.309000e+10 1.254810e+11 2.879120e+11\nAAPL 2021-12-31 3.711900e+10 1.799200e+10 1.531540e+11 7.193200e+10 1.475740e+11 3.092590e+11\nAAPL 2022-03-31 2.809800e+10 1.429800e+10 1.181800e+11 6.739900e+10 1.275080e+11 2.832630e+11\nAAPL 2022-06-30 2.750200e+10 1.285200e+10 1.122920e+11 5.810700e+10 1.298730e+11 2.782020e+11\nAAPL 2022-09-30 2.364600e+10 1.854600e+10 1.354050e+11 5.067200e+10 1.539820e+11 3.020830e+11\n--------------------------------------------------------------------------------\nGOOG\n asOfDate CashAndCashEquivalents CashFinancial CurrentAssets TangibleBookValue CurrentLiabilities TotalLiabilitiesNetMinorityInterest\nsymbol \nGOOG 2021-09-30 2.371900e+10 NaN 1.841100e+11 2.203950e+11 6.178200e+10 1.028360e+11\nGOOG 2021-12-31 2.094500e+10 NaN 1.881430e+11 2.272620e+11 6.425400e+10 1.076330e+11\nGOOG 2022-03-31 2.088600e+10 NaN 1.778530e+11 2.296810e+11 6.194800e+10 1.030920e+11\nGOOG 2022-06-30 1.793600e+10 NaN 1.723710e+11 2.300930e+11 6.135400e+10 9.976600e+10\nGOOG 2022-09-30 2.198400e+10 NaN 1.661090e+11 2.226000e+11 6.597900e+10 1.046290e+11\n--------------------------------------------------------------------------------\nMSFT\n asOfDate CashAndCashEquivalents CashFinancial CurrentAssets TangibleBookValue CurrentLiabilities TotalLiabilitiesNetMinorityInterest\nsymbol \nMSFT 2021-09-30 1.916500e+10 6.863000e+09 1.743260e+11 9.372900e+10 8.052800e+10 1.834400e+11\nMSFT 2021-12-31 2.060400e+10 6.255000e+09 1.741880e+11 1.016270e+11 7.751000e+10 1.803790e+11\nMSFT 2022-03-31 1.249800e+10 7.456000e+09 1.539220e+11 8.420500e+10 7.743900e+10 1.816830e+11\nMSFT 2022-06-30 1.393100e+10 8.258000e+09 1.696840e+11 8.772000e+10 9.508200e+10 1.982980e+11\nMSFT 2022-09-30 2.288400e+10 7.237000e+09 1.608120e+11 9.529900e+10 8.738900e+10 1.862180e+11\n--------------------------------------------------------------------------------\n\n" ]

[ 1, 1 ]

[]

[]

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

[ "I've just had the same problem.\nLooking back, the error is fairly clear: ModuleNotFoundError: No module named 'encodings' means Python is not able to find the encodings module, one of Python built-in modules. I had a look at my filesystem and found out that the encodings module is found at C:\\msys64\\mingw64\\lib\\python3.8\\encodings. This meant that Python could not find that folder!\nMy simple solution was to add C:\\msys64\\mingw64\\lib\\python3.8 to the PYTHONPATH environment variable. On Windows you can either set it from the Windows settings UI - the permanent way - or set it temporarily from the command line, using this command: set PYTHONPATH=C:\\msys64\\mingw64\\lib\\python3.8\\.\nThis kind of feels like a hack since, as a basic Python module, it should be found automatically and yet doesn't. However, at this point, I've had no other Python-related problems so I figure it really does the trick.\n" ]

[ 0 ]

[]

[]

[ "msys", "python", "python_3.x", "windows" ]