QuestionId
int64
74.8M
79.8M
UserId
int64
56
29.4M
QuestionTitle
stringlengths
15
150
QuestionBody
stringlengths
40
40.3k
Tags
stringlengths
8
101
CreationDate
stringdate
2022-12-10 09:42:47
2025-11-01 19:08:18
AnswerCount
int64
0
44
UserExpertiseLevel
int64
301
888k
UserDisplayName
stringlengths
3
30
78,330,264
14,501,369
How do I extract a list of conditional formatting rules from excel using the xml?
<p>I need to extract a list of all cells that have conditional formatting, where they might be highlighted yellow as a result of conditional formatting. I have tried using openxpyxl to do this, but it will only give me a list of all cells that have a conditional formatting rules, regardless of whether the rule might result in a yellow, red, blue, etc highlight. I am unable to us pywin32 or xlwings to do this as they run too slowly.</p> <p>Current openpyxl code:</p> <pre><code>wb_openpyxl = openpyxl.load_workbook(filename) ws_openpyxl = wb_openpyxl[&quot;Sheet1&quot;] for cf in ws_openpyxl.conditional_formatting: print(cf) #&lt;ConditionalFormatting A5:A9&gt; </code></pre> <p>Desired result:</p> <p>Conditional formatting A5:A9 if a5&gt;0, (hex color associated with yellow)</p>
<python><excel><xml><openpyxl>
2024-04-15 18:00:12
1
997
Hooded 0ne
78,330,155
14,501,369
How do I extract the color of an excel shape using the xml?
<p>I have the below code which works beautifully to extract the text from all shapes in an excel file. For a simple xlsx with 2 sheets and 1 shape per sheet it will return the following:</p> <p>['Text box 2'] ['Sheet2 ellipse text 3']</p> <p>However, I am needing it to also return the color of the shape. I cannot use xlwings or pywin32 as those take too long to process. I do not believe that openpyxl has this feature.</p> <pre><code>from zipfile import ZipFile from lxml import etree def main(): z = ZipFile('/home/luis/tmp/shapes.xlsx') drawings = [ drw for drw in z.filelist if &quot;drawings&quot; in drw.filename.split('/') ] for drw in drawings: with z.open(drw.filename) as sstr: ssdata = sstr.read() tree = etree.fromstring(ssdata) ns = {&quot;xdr&quot;: &quot;http://schemas.openxmlformats.org/drawingml/2006/spreadsheetDrawing&quot;, &quot;a&quot;: &quot;http://schemas.openxmlformats.org/drawingml/2006/main&quot;, &quot;r&quot;: &quot;http://schemas.openxmlformats.org/officeDocument/2006/relationships&quot; } shapes = tree.xpath('//xdr:txBody[preceding-sibling::xdr:spPr]/a:p/a:r/a:t/text()', namespaces = ns) print(shapes) if __name__ == '__main__': main() </code></pre>
<python><excel><xml>
2024-04-15 17:38:30
1
997
Hooded 0ne
78,330,095
4,398,966
dictionary of lists of dictionaries not filtering
<p>I tried the following:</p> <pre><code>positions = { 'IBM': [ {'ticker': 'IBM', 'start_date': '20240415', 'end_date': 99999999}, {'ticker': 'IBM', 'start_date': '20240416', 'end_date': 00000000}, {'ticker': 'IBM', 'start_date': '20240417', 'end_date': 99999999}], 'MRK': [ {'ticker': 'mmm', 'start_date': '20240415', 'end_date': 99999999}, {'ticker': 'mmm', 'start_date': '20240416', 'end_date': 00000000}, {'ticker': 'mmm', 'start_date': '20240417', 'end_date': 99999999}]} new_pos = [position for position in positions.values() if position[0].get('end_date') == 99999999] print(new_pos) </code></pre> <p>and got back:</p> <pre><code>[[{'ticker': 'IBM', 'start_date': '20240415', 'end_date': 99999999}, {'ticker': 'IBM', 'start_date': '20240416', 'end_date': 0}, {'ticker': 'IBM', 'start_date': '20240417', 'end_date': 99999999}], [{'ticker': 'mmm', 'start_date': '20240415', 'end_date': 99999999}, {'ticker': 'mmm', 'start_date': '20240416', 'end_date': 0}, {'ticker': 'mmm', 'start_date': '20240417', 'end_date': 99999999}]] </code></pre> <p>why is the if condition not working and why do we get back all records instead of just the first record in the list with <code>position[0]</code>?</p> <p>I think if the first dict of each list has end = 99999999 then we get back the entire list. How do we just get back the dicts where end = 99999999?</p>
<python><list><dictionary><conditional-statements>
2024-04-15 17:26:25
1
15,782
DCR
78,330,035
23,626,926
Python `regex` module - get unfuzzied match from fuzzy match
<p>I an writing a simple command interpreter for a project to allow the user to interact with a virtual world. In the commands the user can refer to objects by any number of different names, and the objects can appear and disappear. Also, to be nice to the user I am allowing some typos.</p> <p>To figure out which object the user is referring to, I am using the Python <a href="https://pypi.org/project/regex" rel="nofollow noreferrer">regex</a> module to do fuzzy matching. However, when there is a typo, the matched group no longer has the same name as the dictionary I am using to look up the actual object by its name, and the lookup fails.</p> <p>Pared down test case:</p> <pre class="lang-py prettyprint-override"><code># test.py from dataclasses import dataclass, field import regex @dataclass class Command: phrase: str arguments: list[str] argtypes: list[type] pattern: str = field(init=False) def __post_init__(self): assert (len(set(self.arguments)) == len(self.argtypes)) def fp(match): if (n := match.group(0)) in self.arguments: return rf&quot;(?:(?P&lt;{n}&gt;\b\L&lt;{n}&gt;\b)){{e&lt;=5}}&quot; return rf&quot;(?:\b{regex.escape(n)}\b){{e&lt;=3}}&quot; munged = regex.sub(r&quot;\b(\w+)\b&quot;, fp, self.phrase) munged = regex.sub(r&quot;\s+&quot;, r&quot;\\s+&quot;, munged) self.pattern = munged def match(self, string: str, candidate_captures: list) -&gt; tuple[list[str] | None, int | None]: &quot;&quot;&quot;Match self on the string, return the matched objects and the number of errors, or None, None if no match&quot;&quot;&quot; # assemble the candidates dict options = {x: [] for x in self.arguments} unmap = {} for c in candidate_captures: for a, t in zip(self.arguments, self.argtypes): if isinstance(c, t): s = str(c) options[a].append(s) unmap[s] = c match: regex.Match = regex.search(self.pattern, string, **options, flags=regex.BESTMATCH) if match: return ([unmap[match.group(g)] for g in self.arguments], #####&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt; sum(match.fuzzy_counts)) return None, None </code></pre> <pre><code>&gt;&gt;&gt; from test import Command &gt;&gt;&gt; x = Command(&quot;bar X foo Y&quot;, [&quot;X&quot;, &quot;Y&quot;], [int, float]) &gt;&gt;&gt; x Command(phrase='bar X foo Y', arguments=['X', 'Y'], argtypes=[&lt;class 'int'&gt;, &lt;class 'float'&gt;], pattern='(?:\\bbar\\b){e&lt;=3}\\s+(?:(?P&lt;X&gt;\\b\\L&lt;X&gt;\\b)){e&lt;=5}\\s+(?:\\bfoo\\b){e&lt;=3}\\s+(?:(?P&lt;Y&gt;\\b\\L&lt;Y&gt;\\b)){e&lt;=5}') &gt;&gt;&gt; x.match(&quot;baar 12 foo 345&quot;, [12, 34.5, 17, 65.9]) Traceback (most recent call last): File &quot;&lt;stdin&gt;&quot;, line 1, in &lt;module&gt; File &quot;test.py&quot;, line 39, in match return ([unmap[match.group(g)] for g in self.arguments], #####&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt; ~~~~~^^^^^^^^^^^^^^^^ KeyError: '345' &gt;&gt;&gt; # this is because it fuzzymatched &quot;34.5&quot; &gt;&gt;&gt; # how do I get the &quot;34.5&quot; out again? </code></pre> <p>What I am looking to do is, on the marked line that errors, to get the <em>original</em>, unfuzzied text, instead of the <em>actual</em> fuzzy text. I can't figure out how to do this.</p>
<python><regex><fuzzy-search><python-regex>
2024-04-15 17:12:22
0
360
dragoncoder047
78,329,987
5,312,606
numba dispatch on type
<p>I would like to dispatch on the type of the second argument in a function in numba and fail in doing so.</p> <p>If it is an integer then a vector should be returned, if it is itself an array of integers, then a matrix should be returned.</p> <p>The first code does not work</p> <pre class="lang-py prettyprint-override"><code>@njit def test_dispatch(X, indices): if isinstance(indices, nb.int64): ref_pos = np.empty(3, np.float64) ref_pos[:] = X[:, indices] return ref_pos elif isinstance(indices, nb.int64[:]): ref_pos = np.empty((3, len(indices)), np.float64) ref_pos[:, :] = X[:, indices] return ref_pos </code></pre> <p>while the second one, with an <code>else</code>, does.</p> <pre class="lang-py prettyprint-override"><code>@njit def test_dispatch(X, indices): if isinstance(indices, nb.int64): ref_pos = np.empty(3, np.float64) ref_pos[:] = X[:, indices] return ref_pos else: ref_pos = np.empty((3, len(indices)), np.float64) ref_pos[:, :] = X[:, indices] return ref_pos </code></pre> <p>I guess that the problem is the type declaration via <code>nb.int64[:]</code> but I don't get it to work in any other way. Do you have an idea?</p> <p>Note that this question applies to <code>numba&gt;=0.59</code>. <code>generated_jit</code> is deprecated in earlier versions and actually removed from versions 0.59 on.</p>
<python><types><numba>
2024-04-15 17:03:41
1
1,897
mcocdawc
78,329,750
12,800,206
dockerized python application takes a long time to trim a video with ffmpeg
<p>The project trims YouTube videos.</p> <p>When I ran the ffmpeg command on the terminal, it didn't take too long to respond. The code below returns the trimmed video to the front end but it takes too long to respond. A 10 mins trim length takes about 5mins to respond. I am missing something, but I can't pinpoint the issue.</p> <p>backend</p> <p>main.py</p> <pre><code>import os from flask import Flask, request, send_file from flask_cors import CORS, cross_origin app = Flask(__name__) cors = CORS(app) current_directory = os.getcwd() folder_name = &quot;youtube_videos&quot; save_path = os.path.join(current_directory, folder_name) output_file_path = os.path.join(save_path, 'video.mp4') os.makedirs(save_path, exist_ok=True) def convert_time_seconds(time_str): hours, minutes, seconds = map(int, time_str.split(':')) total_seconds = (hours * 3600) + (minutes * 60) + seconds return total_seconds def convert_seconds_time(total_seconds): new_hours = total_seconds // 3600 total_seconds %= 3600 new_minutes = total_seconds // 60 new_seconds = total_seconds % 60 new_time_str = f'{new_hours:02}:{new_minutes:02}:{new_seconds:02}' return new_time_str def add_seconds_to_time(time_str, seconds_to_add): total_seconds = convert_time_seconds(time_str) total_seconds -= seconds_to_add new_time_str = convert_seconds_time(total_seconds) return new_time_str def get_length(start_time, end_time): start_time_seconds = convert_time_seconds(start_time) end_time_seconds = convert_time_seconds(end_time) length = end_time_seconds - start_time_seconds length_str = convert_seconds_time(length) return length_str def download_url(url): command = [ &quot;yt-dlp&quot;, &quot;-g&quot;, url ] try: links = subprocess.run(command, capture_output=True, text=True, check=True) video, audio = links.stdout.strip().split(&quot;\n&quot;) return video, audio except subprocess.CalledProcessError as e: print(f&quot;Command failed with return code {e.returncode}.&quot;) print(f&quot;Error output: {e.stderr}&quot;) return None except ValueError: print(&quot;Error: Could not parse video and audio links.&quot;) return None def download_trimmed_video(video_link, audio_link, start_time, end_time): new_start_time = add_seconds_to_time(start_time, 30) new_end_time = get_length(start_time, end_time) if os.path.exists(output_file_path): os.remove(output_file_path) command = [ 'ffmpeg', '-ss', new_start_time + '.00', '-i', video_link, '-ss', new_start_time + '.00', '-i', audio_link, '-map', '0:v', '-map', '1:a', '-ss', '30', '-t', new_end_time + '.00', '-c:v', 'libx264', '-c:a', 'aac', output_file_path ] try: result = subprocess.run(command, capture_output=True, text=True, check=True) if result.returncode == 0: return &quot;Trimmed video downloaded successfully!&quot; else: return &quot;Error occurred while downloading trimmed video&quot; except subprocess.CalledProcessError as e: print(f&quot;Command failed with return code {e.returncode}.&quot;) print(f&quot;Error output: {e.stderr}&quot;) app = Flask(__name__) @app.route('/trimvideo', methods =[&quot;POST&quot;]) @cross_origin() def trim_video(): print(&quot;here&quot;) data = request.get_json() video_link, audio_link = download_url(data[&quot;url&quot;]) if video_link and audio_link: print(&quot;Downloading trimmed video...&quot;) download_trimmed_video(video_link, audio_link, data[&quot;start_time&quot;], data[&quot;end_time&quot;]) response = send_file(output_file_path, as_attachment=True, download_name='video.mp4') response.status_code = 200 return response else: return &quot;Error downloading video&quot;, 400 if __name__ == '__main__': app.run(debug=True, port=5000, host='0.0.0.0') </code></pre> <p>dockerfile</p> <pre><code>FROM ubuntu:latest # Update the package list and install wget and ffmpeg RUN apt-get update \ &amp;&amp; apt-get install -y wget ffmpeg python3 python3-pip \ &amp;&amp; rm -rf /var/lib/apt/lists/* # Download the latest version of yt-dlp and install it RUN wget https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp -O /usr/local/bin/yt-dlp \ &amp;&amp; chmod a+rx /usr/local/bin/yt-dlp WORKDIR /app COPY main.py /app/ COPY requirements.txt /app/ RUN pip install --no-cache-dir -r requirements.txt # Set the default command CMD [&quot;python3&quot;, &quot;main.py&quot;] </code></pre> <p>requirements.txt</p> <pre><code>blinker==1.7.0 click==8.1.7 colorama==0.4.6 Flask==3.0.3 Flask-Cors==4.0.0 itsdangerous==2.1.2 Jinja2==3.1.3 MarkupSafe==2.1.5 Werkzeug==3.0.2 </code></pre> <p>frontend</p> <p>App.js</p> <pre><code> import React, { useState } from 'react'; import './App.css'; import axios from 'axios'; async function handleSubmit(event, url, start_time, end_time, setVideoUrl, setIsSubmitted){ event.preventDefault(); if( url &amp;&amp; start_time &amp;&amp; end_time){ try { setIsSubmitted(true); const response = await axios.post('http://127.0.0.1:5000/trimvideo', { url: url, start_time: start_time, end_time: end_time }, { responseType: 'blob', headers: {'Content-Type': 'application/json'} } ) const blob = new Blob([response.data], { type: 'video/mp4' }); const newurl = URL.createObjectURL(blob); setVideoUrl(newurl); } catch (error) { console.error('Error trimming video:', error); } } else { alert('Please fill all the fields'); } } function App() { const [url, setUrl] = useState(''); const [startTime, setStartTime] = useState(''); const [endTime, setEndTime] = useState(''); const [videoUrl, setVideoUrl] = useState(''); const [isSubmitted, setIsSubmitted] = useState(false); return ( &lt;div className=&quot;App&quot;&gt; &lt;div className=&quot;app-header&quot;&gt;TRIM AND DOWNLOAD YOUR YOUTUBE VIDEO HERE&lt;/div&gt; &lt;input className = &quot;input-url&quot;placeholder='Enter url' value={url} onChange={e=&gt;setUrl(e.target.value)}/&gt; &lt;div className=&quot;input-container&quot;&gt; &lt;input className=&quot;start-time-url&quot; placeholder=&quot;start time&quot; value={startTime} onChange={(e)=&gt;setStartTime(e.target.value)}/&gt; &lt;input className=&quot;end-time-url&quot; placeholder=&quot;end time&quot; value={endTime} onChange={e=&gt;setEndTime(e.target.value)}/&gt; &lt;/div&gt; { !isSubmitted &amp;&amp; &lt;button onClick={(event) =&gt; handleSubmit(event, url, startTime, endTime, setVideoUrl, setIsSubmitted)} className='trim-button'&gt;Trim&lt;/button&gt; } { ( isSubmitted &amp;&amp; !videoUrl) &amp;&amp; &lt;div className=&quot;dot-pulse&quot;&gt;&lt;/div&gt; } { videoUrl &amp;&amp; &lt;video controls autoPlay width=&quot;500&quot; height=&quot;360&quot;&gt; &lt;source src={videoUrl} type='video/mp4'&gt;&lt;/source&gt; &lt;/video&gt; } &lt;/div&gt; ); } export default App; </code></pre>
<python><docker><flask><ffmpeg><yt-dlp>
2024-04-15 16:12:17
0
748
Ukpa Uchechi
78,329,714
2,687,317
Pandas - groupby string field and select by time-of-day range
<p>I have a dataset like this</p> <pre><code>index Date_Time Pass_ID El 0 3/30/23 05:12:36.36 A 1 1 3/30/23 05:12:38.38 A 2 1 3/30/23 05:12:40.40 A 3 1 3/30/23 05:12:42.42 A 4 1 3/30/23 05:12:44.44 A 4 1 3/30/23 12:12:50.50 B 3 1 3/30/23 12:12:52.52 B 4 1 3/30/23 12:12:54.54 B 5 1 3/30/23 12:12:56.56 B 6 1 3/30/23 12:12:58.58 B 7 1 3/30/23 12:13:00.00 B 8 1 3/30/23 12:13:02.02 B 9 1 3/31/23 20:02:02.02 C 3 1 3/31/23 20:02:05.05 C 4 </code></pre> <p>The Date_Time is pandas datetime object.</p> <p>I'd like to group the records by <code>Pass_ID</code>, and <em><strong>then select out only those unique Pass_IDs that occur between specific hours</strong></em> in the day: for instance, between 10:00 and 13:00 would return B.</p> <p>I don't know how to get groupby and 'between_time' to work in this case... which would seem to be the best way forward. I've also tried using a lambda function after setting the Date_Time as the index, but that didn't work. And using aggregate doesn't seem to allow me to pull out the dt.hour of the Date_Time field. Anyone know how to do this concisely?</p>
<python><pandas><group-by>
2024-04-15 16:05:19
2
533
earnric
78,329,495
4,909,242
what is the equivalent of numpy accumulate ufunc in pytorch
<p>In numpy, I can do the following:</p> <pre><code>&gt;&gt;&gt; x = np.array([[False, True, False], [True, False, True]]) &gt;&gt;&gt; z0 = np.logical_and.accumulate(x, axis=0) &gt;&gt;&gt; z1 = np.logical_and.accumulate(x, axis=1) </code></pre> <p>This returns the following:</p> <pre><code>&gt;&gt;&gt; z0 array([[False, True, False], [False, False, False]]) &gt;&gt;&gt; z1 array([[False, False, False], [ True, False, False]]) </code></pre> <p>What is the equivalent of this ufunc operation in pytorch?</p>
<python><numpy><pytorch><numpy-ufunc>
2024-04-15 15:26:35
2
709
Wei Li
78,329,445
15,452,168
How to Efficiently Process 6000 Requests in Python with Limited Time Constraints?
<p>I have a Python script that sends requests to an API to generate responses based on input prompts. Each request typically takes around 10 seconds to process. I have a CSV file containing 6000 rows, and I want to send the values from the 'text' column as input prompts to the API. Additionally, I aim to append the generated responses to a new column in the same CSV file.</p> <p>Given the large number of requests and the time it takes for each request to complete, I'm concerned about the overall processing time. Is there a way to optimize this process to avoid spending 6000*10 seconds sequentially?</p> <p>Here's the base code snippet:</p> <pre><code>import pandas as pd from requests_futures.sessions import FuturesSession import time # Define your existing function for sending a request def send_request(session, input_prompt): url = 'https://ai-chat-test.***.workers.dev/' consolidated_request = f&quot;&quot;&quot;From the input &quot;{input_prompt}&quot;, extract key descriptions, deduplicate them, and use these to create a compelling, concise description of the fashion item.&quot;&quot;&quot; return session.post(url, json={&quot;messages&quot;: [{&quot;role&quot;: &quot;user&quot;, &quot;content&quot;: consolidated_request}]}, headers={&quot;Content-Type&quot;: &quot;application/json&quot;}) # Function to apply send_request concurrently and save intermittently def apply_concurrent_requests(df, function): session = FuturesSession(max_workers=5) futures = {function(session, text): index for index, text in enumerate(df['text'])} results = [None] * len(df) batch_counter = 0 start_time = time.time() for future in as_completed(futures): index = futures[future] try: response = future.result() if response.status_code == 200: data = response.json() for message in data: if message['role'] == 'assistant': results[index] = message['content'] break else: results[index] = &quot;No assistant message found.&quot; else: results[index] = f&quot;Failed to get a response, status code: {response.status_code}&quot; except Exception as e: results[index] = f&quot;Exception occurred: {e}&quot; # Save after processing each batch of 100 or the last response batch_counter += 1 if batch_counter % 100 == 0 or index + 1 == len(df): # Save the responses so far to the DataFrame for j, res in enumerate(results[:index+1]): df.at[j, 'response'] = res # Save to CSV df.to_csv('intermediate_metadata.csv', index=False) print(f&quot;Processed {index+1} requests in {time.time() - start_time:.2f} seconds and saved to CSV.&quot;) start_time = time.time() # Reset the timer for the next batch return df df = pd.read_csv('metadata.csv') updated_df = apply_concurrent_requests(df, send_request) updated_df.to_csv('updated_metadata.csv', index=False) </code></pre> <p>How can I optimize this process to send multiple requests at once or utilize parallel processing to reduce the overall processing time?</p>
<python><python-3.x><multithreading><multiprocessing><python-multiprocessing>
2024-04-15 15:20:45
0
570
sdave
78,329,319
13,174,189
How to lattitude, longtitude and full address based on address?
<p>I have a dataframe with column address_1:</p> <pre><code>address_1 apple fifth avenue new york burj khalifa dubai microsoft office san francisco </code></pre> <p>I want to get longtitude, lattitude and full address based on address_1. How to di it? I tried geopandas library:</p> <pre><code>geolocator = Nominatim(user_agent=&quot;myGeocoder&quot;) location = geolocator.geocode(&quot;apple fifth avenue new york&quot;) address = location.address latitude = location.latitude longitude = location.longitude print(latitude, longitude) </code></pre> <p>but i get this error:</p> <pre><code>GeocoderInsufficientPrivileges Traceback (most recent call last) /usr/local/lib/python3.10/dist-packages/geopy/geocoders/base.py in _adapter_error_handler(self, error) 409 ) from error 410 else: --&gt; 411 raise exc_cls(str(error)) from error 412 else: 413 res = self._geocoder_exception_handler(error) GeocoderInsufficientPrivileges: Non-successful status code 403 </code></pre>
<python><python-3.x><google-maps><geolocation><geopandas>
2024-04-15 15:02:27
1
1,199
french_fries
78,329,226
447,426
PySpark where to find logs and how to log properly
<p>I have set up a local (docker) spark cluster as provided by <a href="https://github.com/bitnami/containers/blob/main/bitnami/spark/docker-compose.yml" rel="nofollow noreferrer">bitnami</a>. Everything is running fine. I can submit simple spark jobs and run them and i can interact with pyspark.</p> <p>Now i try to set up basic logging (my sources are <a href="https://stackoverflow.com/questions/25407550/how-do-i-log-from-my-python-spark-script">stackoverflow</a> and <a href="https://gist.github.com/smartnose/9f173b4c36dc31310e8efd27c3535a14" rel="nofollow noreferrer">github gist</a>) Mainly i see 2 ways first:</p> <pre><code>log4jref = sc._jvm.org.apache.log4j logger = log4jref.getLogger(__name__) </code></pre> <p>If i am using this in pyspark console i see at least the output in console:</p> <pre><code>&gt;&gt;&gt; logger.info(&quot;something&quot;) 24/04/15 14:24:31 INFO __main__: something </code></pre> <p>If i am using the &quot;cleaner solution&quot;:</p> <pre><code>import logging logger = logging.getLogger(__name__) logger.info(&quot;here we go&quot;) </code></pre> <p>i do not see anything returned in pyspark console.</p> <p>In both cases i do not see my outputs in <code>docker logs spark-spark-worker-1</code> nor on master <code>docker logs spark-spark-1</code> (other outputs are there). I also tried to look in <code>/opt/bitnami/spark/logs</code></p> <p>So my question is: <strong>where to find my log messages in general</strong>? And what of the 2 ways is better? In a first steps i just need a quick way to get feedback from my running code - making it production ready comes later (if not easily possible right from start)</p> <p><strong>little update</strong></p> <p>the job that i am running for test looks like this:</p> <pre><code># Create a SparkSession spark = SparkSession.builder \ .appName(&quot;My App&quot;) \ .getOrCreate() log = logging.getLogger(__name__) rdd = spark.sparkContext.parallelize(range(1, 100)) log.error(&quot;AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA&quot;) log.error(f&quot;THE SUM IS HERE: {rdd.sum()}&quot;) print(f&quot;THE SUM IS HERE: {rdd.sum()}&quot;) print(&quot;AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA&quot;) # Stop the SparkSession spark.stop() </code></pre> <p>While i see in logs that this job runs successfully i can't see my log entries nor the print. All this places seem to show the same log:</p> <ul> <li>docker ui (uses API to access logs?)</li> <li>docker logs (for master and worker)</li> <li>file /opt/bitnami/spark/work/app-20240416084653-0006/0/stderror on worker</li> </ul>
<python><docker><logging><pyspark>
2024-04-15 14:46:17
1
13,125
dermoritz
78,329,019
4,469,336
Django: Adding unique together errors to one of the involved fields
<p>Note: I asked this question on the <a href="https://forum.djangoproject.com/t/adding-unique-together-errors-to-one-of-the-involved-fields/30017/1" rel="nofollow noreferrer">Django forum</a>, but since I did not receive any answers there, I'm asking here, too.</p> <p>I’m building an application with multi-tenancy, meaning that all data is associated to a tenant model. When creating models, I often want to validate uniqueness of e.g. the <code>name</code> field. However, the uniqueness is only required per tenant, not globally, which is why I’m using <code>unique_together</code> (or <code>UniqueConstraint</code>, found that too late, but will migrate).</p> <p>However, when the uniqueness is violated, a <code>non_field_error</code> is added. I would however prefer to add this error to the <code>name</code> field, as the tenant field is implicit and not visible to the user (the tenant is always the tenant of the logged in user, and can’t be chosen).</p> <p>Any hints on how I could achieve that, preferably in the model or alternatively in a form? I thought about first validating the model, and then moving the error from the non_field_errors to the name field, but that feels kind of wrong.</p> <p>Starting point would be this code:</p> <pre class="lang-py prettyprint-override"><code>class Company(models.Model): tenant = models.ForeignKey(Tenant, on_delete=models.PROTECT) name = models.CharField(max_length=255) class Meta: unique_together = ('tenant', 'name') </code></pre>
<python><django><validation><unique-constraint>
2024-04-15 14:14:58
0
2,985
Timitry
78,328,963
3,042,018
PyCharm warns "Expected type 'int', got 'float'" for random.randint, but code still runs sometimes
<p>PyCharm (Python version 3.11.3 on Windows) is flagging <code>Expected type 'int', got 'float' instead</code> where I use <code>random.randint(x/y, 10)</code> (where x/y will be, say, 5.0), which makes sense. However the code still runs.</p> <p>Someone else using a different IDE tried to run the code and <code>TypeError</code> prevented the code from running. I don't know what system/version that was on, but it doesn't make sense to me.</p> <p>For minimal viable program:</p> <pre><code>import random x = 10 y = 2 print(random.randint(x/y, 10) </code></pre> <p>Why is there a discrepancy between the PyCharm warning and the runtime errors, and why is it not consistent across systems?</p>
<python><typeerror>
2024-04-15 14:04:42
1
3,842
Robin Andrews
78,328,935
547,231
Computing the Jacobian of an image: How to reshape the numpy array properly?
<p>I have a batch <code>x</code> of images of the shape <code>[k, width, height, channel_count]</code>. This batch is transformed by a function <code>f</code>. The result has the same shape and I need to compute the divergence (i.e. trace of the Jacobian) of this transformation. (To emphasize this: The transform is working on a batch; it's the output of a neural network and I cannot change it).</p> <p>I'm using <code>jax.jacfwd</code> to compute the Jacobian. The output has shape <code>[k, width, height, channel_count, k, width, height, channel_count]</code>. That is the first problem, I actually need the Jacobian per image. So the output should have shape <code>[k, width, height, channel_count, width, height, channel_count]</code> instead. I don't know how I can achieve this using <code>jax.jacfwd</code>, since I don't have a per image transform (only the per batch transform).</p> <p>And even if I have the desired output, how can I compute the trace (per image)? The output should have shape <code>[k]</code>. I think I need to reshape the jacobian output to <code>[k, width * height * channel_count, width * height * channel_count]</code>, but how can I do that?</p> <p><em>Remark</em>: Note that I also need to know the value of the transform itself. Since there is no <code>jax.val_and_jacfwd</code>, I'm return the actual value as an auxiliary variable, which should be fine. The solution to my question should still allow this.</p> <p><strong>EDIT</strong>:</p> <p>Here is a minimal reproducible example:</p> <pre><code>import jax import torch def f(x): # s = model.apply(params, x) ### The actual code queries a network model # for reproducibility, here is a simple transform: s = jax.numpy.empty(x.shape) for i in range(x.shape[0]): for u in range(x.shape[1]): for v in range(x.shape[2]): for c in range(x.shape[3]): s = s.at[i, u, v, c].set(x[i, u, v, c] * x[i, u, v, c]) return [s, s] jac = jax.jacfwd(f, has_aux = True) k = 3 width = 2 height = 2 channel_count = 1 x = torch.empty((k, width, height, channel_count)) ### The actual loads x from a batch it = 1 for i in range(x.shape[0]): for u in range(x.shape[1]): for v in range(x.shape[2]): x[i, u, v, 0] = it it += 1 f_jac, f_val = jac(x.numpy()) print(f_jac.shape) </code></pre> <p>The output is <code>(3, 2, 2, 1, 3, 2, 2, 1)</code>. This is clearly not what I want. I don't want to &quot;differentiate one image with respect to pixels of another&quot;. What I actually want is the &quot;per image&quot; Jacobian. So, the output should be of the shape <code>(3, 2, 2, 1, 2, 2, 1)</code> instead.</p> <p>But let me stress it again: I still <em>need</em> that <code>f</code> takes a batch of images, since invoking <code>model.apply</code> for each image in the batch separately would be terribly slow.</p> <p><strong>EDIT 2</strong>:</p> <p>BTW, if there is a direct way to compute the divergence of <code>f</code> - without the need of computing the whole Jacobian before - I would definitely prefer that.</p> <p><strong>EDIT 3</strong>:</p> <p>Regarding only the &quot;per batch Jacobian&quot; thing: Here is an even simpler example:</p> <pre><code>import jax import torch def f(x): s = jax.numpy.empty(x.shape) s = s.at[0].set(x[0] * x[0]) s = s.at[1].set(x[1] * x[1]) s = s.at[2].set(x[2] * x[2]) return [s, s] jac = jax.jacfwd(f, has_aux = True) x = torch.empty(3) x[0] = 1 x[1] = 2 x[2] = 3 f_jac, f_val = jac(x.numpy()) print(f_jac.shape) print(f_jac) print(f_val) </code></pre> <p>The goal would be that <code>f_jac</code> has shape <code>[3, 1, 1]</code>. (The Jacobian of a one-dimensional function is simply a scalar).</p>
<python><numpy><neural-network><jax><automatic-differentiation>
2024-04-15 13:58:44
1
18,343
0xbadf00d
78,328,798
3,324,415
Robot Framework Database Library calling Oracle stored procedure fails with character to number conversion error
<p>I have an <code>Oracle PL/SQL</code> procedure that I can directly call as follows without problem:</p> <pre><code>BEGIN example_package_name.example_procedure(p_item_no =&gt; 123456, p_send_now =&gt; true); END; </code></pre> <p>(Note: <code>p_item_no</code> expects a <code>NUMBER</code> and <code>p_send_now</code> expects a <code>BOOLEAN</code>)</p> <p>I am attempting to run this from within my <code>Robot Framework</code> test automation framework as shown below.</p> <p>First I have a small helper wrapper method for <code>robotframework-databaselibrary</code>:</p> <pre><code>Execute SQL stored procedure [Arguments] ... ${target_database} ... ${target_stored_procedure} ... ${stored_procedure_arguments} ... ${timeout}=5 minutes [Documentation] Small wrapper around DatabaseLibrary for the: Call stored procedure keyword. [Timeout] ${timeout} __Open connection to target database ${target_database} DatabaseLibrary.Set Auto Commit autoCommit=${True} DatabaseLibrary.Call Stored Procedure ${target_stored_procedure} ${stored_procedure_arguments} Close connection from the current database </code></pre> <p>Next from my test I am attempting something as follows:</p> <pre><code>${item_no_int}= Convert To Integer ${test_item_dictionary.item_no} ${example_procedure_argument_list}= Create List p_item_no =&gt; ${item_no_int} p_send_now =&gt; ${True} Execute SQL Stored Procedure target_database=test_db_name target_stored_procedure=example_package_name.example_procedure stored_procedure_arguments=${example_procedure_argument_list} </code></pre> <p>My error reads:</p> <blockquote> <p>[info (+0.10s)] Executing : Call Stored Procedure | example_package_name.example_procedure | ['p_item_no =&gt; 123456', 'p_send_now =&gt; True']</p> </blockquote> <blockquote> <p>[FAIL] DatabaseError:ORA-06502: PL/SQL: numeric or value error: character to number conversion error ORA-06512: at line 1 Help: <a href="https://docs.oracle.com/error-help/db/ora-06502/" rel="nofollow noreferrer">https://docs.oracle.com/error-help/db/ora-06502/</a></p> </blockquote> <p>Naturally I have been trying to ensure that my data is of the correct type when leaving Robot Framework, when reading the documentation on the Robot Framework: <code>DatabaseLibrary.Call Stored Procedure</code> keyword, I see:</p> <pre><code>def call_stored_procedure( self, spName: str, spParams: Optional[List[str]] = None, sansTran: bool = False, alias: Optional[str] = None ): </code></pre> <p>With description:</p> <blockquote> <p>Calls a stored procedure <code>spName</code> with the <code>spParams</code> - a <em>list</em> of parameters the procedure requires.</p> </blockquote> <p>Is it be possible that the Keyword <code>DatabaseLibrary.Call Stored Procedure</code> / <code>spParams: Optional[List[str]]</code> does not end up preserving one's intended data types? Or is something else perhaps missing on my part?</p> <p>I am running:</p> <ul> <li><code>robotframework&gt;=7.0.0</code></li> <li><code>robotframework-databaselibrary&gt;=1.4.3</code></li> <li><code>oracledb&gt;=2.1.0</code></li> </ul>
<python><oracle-database><plsql><robotframework>
2024-04-15 13:36:10
1
766
BernardV
78,328,713
2,401,856
Sqlite3 getting outdated value when selecting straight after inserting
<p>I'm having a very wiered behavior with sqlite3 in python.<br> I have a small flask service which has 2 endpoints <code>addTask</code> and <code>deleteTask</code>, and in my database I have <code>tasks_order</code> column which holds array that represents the tasks ordering. this order is changeable by the user.<br> <code>addTask</code> - pushes a new random id (task id) into tasks_order array and to the database, this is its code:<br></p> <p>Endpoint:</p> <pre><code>@tasks_bp.route(&quot;/addTask&quot;, methods=[&quot;POST&quot;]) def addTask(): hash = ''.join(secrets.choice(string.ascii_letters + string.digits + string.punctuation) for _ in range(10)) data = request.get_json() return tasksService.addTask(hash) </code></pre> <p>Service:</p> <pre><code>def addTask(hash): try: conn = db.getConn() new_task_id = random.randint(1, 999) tasks_order = dbm.getTasksOrderList() dbm.updateTasksOrderList(tasks_order + [new_task_id]) conn.commit() tasks_order = dbm.getTasksOrderList() print(hash, &quot;added&quot;, new_task_id, datetime.now().strftime(&quot;%M:%S.%f&quot;)) print(hash, tasks_order, datetime.now().strftime(&quot;%M:%S.%f&quot;)) except Exception as e: conn.rollback() raise e else: return {&quot;id&quot;: new_task_id} </code></pre> <p>dbm:</p> <pre><code>def getTasksOrderList(): cur = db.getConn().cursor() row = cur.execute(&quot;SELECT order_list FROM tasks_order&quot;).fetchone() order_list_str = row[&quot;order_list&quot;] return json.loads(order_list_str) def updateTasksOrderList(order_list): order_list_str = json.dumps(order_list) cur = db.getConn().cursor() cur.execute(&quot;UPDATE tasks_order SET order_list = ?&quot;, (order_list_str,)) </code></pre> <p>Notice that I'm addid <code>hash</code> and some prints for debugging manners only.<br> <code>deleteTask</code> - receives a task id, <strong>selects the tasks_order array from the database</strong>, removes this task id then updates the tasks_order in database. Here's its code:</p> <p>Endpoint:</p> <pre><code>@tasks_bp.route(&quot;/deleteTask&quot;, methods=[&quot;POST&quot;]) def deleteTask(): try: hash = ''.join(secrets.choice(string.ascii_letters + string.digits + string.punctuation) for _ in range(10)) data = request.get_json() return tasksService.deleteTask(hash, data[&quot;task_id&quot;]) except Exception as e: print(hash, &quot;ERROR OCCURED&quot;, data[&quot;task_id&quot;]) </code></pre> <p>Service:</p> <pre><code>def deleteTask(hash, task_id): print(hash, &quot;todelete&quot;, task_id, datetime.now().strftime(&quot;%M:%S.%f&quot;)) try: conn = db.getConn() tasks_order = dbm.getTasksOrderList() print(hash, tasks_order, task_id, datetime.now().strftime(&quot;%M:%S.%f&quot;)) tasks_order.remove(task_id) dbm.updateTasksOrderList(tasks_order) conn.commit() except Exception as e: conn.rollback() raise e else: return {}, 204 </code></pre> <p>I handle connections to the database as suggested in flask <a href="https://flask.palletsprojects.com/en/3.0.x/patterns/sqlite3/" rel="nofollow noreferrer">documentation</a>.</p> <pre><code>def getConn(): db = getattr(g, '_database', None) if db is None: db = g._database = sqlite3.connect(DATABASE) db.row_factory = sqlite3.Row return db @app.teardown_appcontext def close_connection(exception): db = getattr(g, '_database', None) if db is not None: db.close() </code></pre> <p>After running the service, I run a performance test in postman, which sends requests to <code>addTask</code> and <code>deleteTask</code> from 20 users. It proccesses few add and delete requests succesfully then fails.<br> I found out that it fails in <code>deleteTask</code> endpoint, when trying to remove a task_id from the array which doesn't exist. The weird thing that I see in logs that the insertion to the array (into the database) happened before calling <code>deleteTask</code> and I was able actually to see that in the logs. <br> But when the connection to the db was closed (which happens after proccessing each request) and a new connection established (which happens when receiving a new request), the new connection cannot see the changes in the database. Why?!</p> <p>logs:</p> <pre><code>?,PXYJ*pse added 237 02:51.805480 ?,PXYJ*pse [54, 950, 86, 678, 417, 404, 237] 02:51.805480 # here we can see that on 02:51.805480 when I selected the array from the database, it included 237 t&quot;qu5NI+bb todelete 237 02:51.903749 t&quot;qu5NI+bb [54, 950, 86, 678, 417, 151] 237 02:51.905749 # here we can see that AFTER 02:51.805480 when I selected the array from the database, it doesn't include 237 t&quot;qu5NI+bb ERROR OCCURED 237 </code></pre>
<python><sqlite><flask>
2024-04-15 13:23:39
1
620
user2401856
78,328,663
5,089,311
Python: swap position of vars expansion
<p>My function receives object <code>myenum</code> which can be either <code>set</code> or <code>dict</code>.<br /> <code>{'a', 'b', 'c'}</code><br /> or<br /> <code>{'a': 5, 'b' : 2, 'c' : 8}</code></p> <p>It represents enum like in C/C++.</p> <p>In case of <code>dict</code> I need enumerate over <code>.items()</code> in case of <code>set</code> I need use <code>enumerate(myenum)</code> like this</p> <pre><code>for k, i in myenum.items(): for i, k in enumerate(myenum): </code></pre> <p>I could use ternary <code>itr = myenum.items() if hasattr(myenum, 'keys') else enumerate(myenum)</code> And then simply have single <code>for</code>.</p> <p>Unfortunately <code>.items()</code> expands into <code>key, value</code> and <code>enumerate()</code> expands into <code>index, key</code>. <code>index</code> is actually the desired value for the enum.</p> <p>Is there way to swap places in the expansion? So that <code>enumerate()</code> also expands into <code>key, index</code>?</p> <p>Right now I have naive solution of 2 <code>for</code> cycles that are identical, just the order of <code>k, i</code> is different, which is ugly.</p>
<python>
2024-04-15 13:15:31
4
408
Noob
78,328,539
13,819,714
Huggingface pipeline available models
<p>I'm working with Huggingface in Python to make inference with specific LLM text generation models. So far I used pipelines like this to initialize the model, and then insert input from a user and retrieve the response:</p> <pre><code>import torch from transformers import pipeline print(torch.cuda.is_available()) generator = pipeline('text-generation', model='gpt2', device=&quot;cuda&quot;) #Inference code </code></pre> <p>However, when I change <code>gpt2</code> with <code>google/gemma-2b-it</code> or some other models, it might ask for authentication or directly it thwors an error indicating it´s not available from <code>pipeline()</code>.</p> <p>I know some models need specific tokenizers and dependencies, but, is there any way to list all available models from <code>pipeline()</code>? And is there any way I can use other models inside <code>pipeline()</code> with all its dependencies without needing to import or use them inside the script?</p>
<python><python-3.x><machine-learning><pytorch><large-language-model>
2024-04-15 12:54:43
2
5,253
Cardstdani
78,328,474
12,094,039
FastAPI Error: HTTPBasic.__call__() missing 1 required positional argument: 'request'
<p>I am trying to create a simple WebSocket app with authentication using FastAPI, but after adding the authentication I face issues,</p> <p>The <code>/{id}</code> request passes, but the WebSocket connection <code>/ws/{client_id}</code> gets failed due to the following error</p> <pre><code>ERROR: Exception in ASGI application Traceback (most recent call last): File &quot;D:\Installations\envs\chat-app\lib\site-packages\uvicorn\protocols\websockets\websockets_impl.py&quot;, line 240, in run_asgi result = await self.app(self.scope, self.asgi_receive, self.asgi_send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\uvicorn\middleware\proxy_headers.py&quot;, line 69, in __call__ return await self.app(scope, receive, send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\fastapi\applications.py&quot;, line 1054, in __call__ await super().__call__(scope, receive, send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\applications.py&quot;, line 123, in __call__ await self.middleware_stack(scope, receive, send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\middleware\errors.py&quot;, line 151, in __call__ await self.app(scope, receive, send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\middleware\exceptions.py&quot;, line 65, in __call__ await wrap_app_handling_exceptions(self.app, conn)(scope, receive, send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\_exception_handler.py&quot;, line 64, in wrapped_app raise exc File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\_exception_handler.py&quot;, line 53, in wrapped_app await app(scope, receive, sender) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\routing.py&quot;, line 756, in __call__ await self.middleware_stack(scope, receive, send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\routing.py&quot;, line 776, in app await route.handle(scope, receive, send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\routing.py&quot;, line 373, in handle await self.app(scope, receive, send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\routing.py&quot;, line 96, in app await wrap_app_handling_exceptions(app, session)(scope, receive, send) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\_exception_handler.py&quot;, line 64, in wrapped_app raise exc File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\_exception_handler.py&quot;, line 53, in wrapped_app await app(scope, receive, sender) File &quot;D:\Installations\envs\chat-app\lib\site-packages\starlette\routing.py&quot;, line 94, in app await func(session) File &quot;D:\Installations\envs\chat-app\lib\site-packages\fastapi\routing.py&quot;, line 338, in app solved_result = await solve_dependencies( File &quot;D:\Installations\envs\chat-app\lib\site-packages\fastapi\dependencies\utils.py&quot;, line 572, in solve_dependencies solved_result = await solve_dependencies( File &quot;D:\Installations\envs\chat-app\lib\site-packages\fastapi\dependencies\utils.py&quot;, line 600, in solve_dependencies solved = await call(**sub_values) TypeError: HTTPBasic.__call__() missing 1 required positional argument: 'request' INFO: connection open INFO: connection closed </code></pre> <p><strong>Code:</strong></p> <pre><code>app = FastAPI() security = HTTPBasic() templates = Jinja2Templates(directory=&quot;templates&quot;) class ConnectionManager: def __init__(self): self.active_connections: List[WebSocket] = [] async def connect(self, websocket: WebSocket): await websocket.accept() self.active_connections.append(websocket) def disconnect(self, websocket: WebSocket): self.active_connections.remove(websocket) async def send_personal_message(self, message: str, websocket: WebSocket): await websocket.send_text(message) async def broadcast(self, message: str, websocket: WebSocket) : for connection in self.active_connections: if (connection == websocket): continue await connection.send_text(message) connectionmanager = ConnectionManager() def get_current_username( credentials: Annotated[HTTPBasicCredentials, Depends(security)], ): is_correct_username, is_correct_password= validate_user(credentials.username.encode(&quot;utf8&quot;),credentials.password.encode(&quot;utf8&quot;)) if not (is_correct_username and is_correct_password): raise HTTPException( status_code=status.HTTP_401_UNAUTHORIZED, detail=&quot;Incorrect username or password&quot;, headers={&quot;WWW-Authenticate&quot;: &quot;Basic&quot;}, ) return credentials.username @app.get(&quot;/{id}&quot;, response_class=HTMLResponse) def read_user(id: int, username: Annotated[str, Depends(get_current_username)], request: Request): # Render the HTML template return templates.TemplateResponse(&quot;index.html&quot;, {&quot;request&quot;: request}) @app.websocket(&quot;/ws/{client_id}&quot;) async def websocket_endpoint( websocket: WebSocket, client_id: int, username: Annotated[str, Depends(get_current_username)], request: Request): # accept connections await connectionmanager.connect(websocket) try: while True: data = await websocket.receive_text() await connectionmanager.send_personal_message(f&quot;You : {data}&quot;, websocket) await connectionmanager.broadcast(f&quot;Client #{client_id}: {data}&quot;, websocket) except WebSocketDisconnect: connectionmanager.disconnect(websocket) await connectionmanager.broadcast(f&quot;Client #{client_id} left the chat&quot;) if __name__ == &quot;__main__&quot;: uvicorn.run(app, host='localhost', port=8000) </code></pre> <p>Before adding the authentication mechanism it was working fine.</p> <p>Even though I added the Request parameter, still I am getting this error.</p>
<python><python-3.x><fastapi><fastapi-middleware>
2024-04-15 12:42:43
1
411
Aravindan vaithialingam
78,328,203
3,603,546
How to get python out of the "raw" terminal mode?
<p>(In Linux context) Entering terminal raw mode (to read mouse, Function keys etc) in python is easy:</p> <pre><code>import sys, tty, termios fd=sys.stdin.fileno() #tty.setraw(fd) tty.setcbreak(fd) </code></pre> <p>But then I want to get back and use <code>input()</code> again. Unfortunately, there is no <code>tty.setcooked()</code> function. What to do?</p>
<python><terminal><stty>
2024-04-15 11:56:27
1
316
user3603546
78,327,891
4,046,235
Use enum fields to define higher level aggregate
<p>In the project I am working on, it's a common pattern to have a function like <code>MachineType.get_external</code> to group the fields:</p> <pre><code>from enum import IntEnum class MachineType(IntEnum): Cloud = 1 OnPrem = 2 Saas = 3 @classmethod def get_external(cls): return [ MachineType.Cloud, MachineType.Saas ] </code></pre> <p>I don't like it and would rather prefer having something more efficient, like:</p> <pre><code>class MachineAggregateType: EXTERNAL = [MachineType.Cloud, MachineType.Saas] </code></pre> <p>i.e. less function calls and less list constructions. (It is used in event processing service, doing 1M events/min, and there are many classes like this which are used for every given event.)</p> <p>Is there a way to have such grouping in the original class as a class field? In my opinion it's better for maintainability and structure than another class.</p>
<python><python-3.x><enums>
2024-04-15 10:59:31
1
3,201
Yuriy Vasylenko
78,327,887
17,530,552
How can I integrate values of a color channel with the image dimensions using cv2?
<p>Using <code>cv2</code>, one can load an image into Python and directly transfer the image into a grayscale version, as shown below.</p> <pre><code>import cv2 image = cv2.imread(&quot;/path-to-image&quot;, cv2.IMREAD_GRAYSCALE) </code></pre> <p>Then, the image is a <em>two dimensional</em> numpy array. Hence, the intensity of the gray colors for each pixel are integrated into the height and width dimensions of the image.</p> <pre><code>image.shape (720, 1024) </code></pre> <p>As we can see above, the image or numpy array has only two dimensions, height (<code>720</code>) and width (<code>1024</code>). Each value or pixel in this matrix consists of a specific value, where the value corresponds to the brightness or gray color of the respective pixel.</p> <p><strong>Selecting single color channels:</strong> Now, I am interested in the three color channels of the image, that is, in the red, green, and blue channels. It is possible to pick the three color channels as follows when we don't load the image in a grayscaled format.</p> <pre><code>import cv2 image = cv2.imread(&quot;/path-to-image&quot;, 1) # the value 1 means to load the original color image for color in [&quot;red&quot;, &quot;green&quot;, &quot;blue&quot;]: if color == &quot;red&quot;: red[:,:,0] = 0 red[:,:,1] = 0 elif color == &quot;green&quot;: green[:,:,0] = 0 green[:,:,2] = 0 elif color == &quot;blue&quot;: blue[:,:,1] = 0 blue[:,:,2] = 0 </code></pre> <p>Paradigmatically, plotting the red image then works as follows:</p> <pre><code>cv2.imshow(&quot;red image plot&quot;, red) cv2.waitKey(0) </code></pre> <p>So far so good.</p> <p><strong>My problem and question:</strong> I would like to integrate the color values for each pixel into a <em>two dimensional</em> numpy array, exactly as it is automatically done when I directly load the image as grayscale format via <code>image = cv2.imread(&quot;/path-to-image&quot;, cv2.IMREAD_GRAYSCALE)</code>.</p> <p>When I individually select the color channels I get a <em>three dimensional</em> numpy array, respectively for each color channel. Here is an example for the red color channel:</p> <pre><code>red.shape (720, 1024, 3) </code></pre> <p>How can I now &quot;integrate&quot; the color channel values, such as in the case of the red image, from the three dimensional into a two dimensional numpy array, as automatically performed when I load the image as grayscale? Therefore, my aim is to get a two dimensional numpy array or matrix, where each pixel in that matrix is the color value for the red color.</p>
<python><arrays><numpy><opencv>
2024-04-15 10:58:24
0
415
Philipp
78,327,822
7,132,482
Python API JAMA: Put attachment to item
<p>I'm facing issues with API Jama to attach an attachment to an item.</p> <p>I need to automatise my jobs and use python to create items and attach zipfile to them. With the console I can create my attachment with my zip file and attach to it to my item.</p> <p>But with the API Jama I don't know how I can create my attachment..</p> <p>In the documentation it seems that I must use this function:</p> <pre><code>def put_attachments_file(self, attachment_id, file_path): &quot;&quot;&quot; Upload a file to a jama attachment :param attachment_id: the integer ID of the attachment item to which we are uploading the file :param file_path: the file path of the file to be uploaded :return: returns the status code of the call &quot;&quot;&quot; resource_path = 'attachments/' + str(attachment_id) + '/file' with open(file_path, 'rb') as f: files = {'file': f} try: response = self.__core.put(resource_path, files=files) except CoreException as err: py_jama_rest_client_logger.error(err) raise APIException(str(err)) self.__handle_response_status(response) return response.status_code </code></pre> <p>But I don't know how I can create the attachment_id then associate my zip file to it.</p> <p>Thanks for your help !</p>
<python><rest><jama>
2024-04-15 10:46:33
1
335
Laurent Cesaro
78,327,635
1,791,983
In Python, want to find all the placemarks in a kml file, but the list is returning empty
<p>Trying to find all the placemarks in a kml file, so I can alter them. But findall doesnt find any.</p> <pre><code>def scan (filename): from lxml import etree tree = etree.parse(open(filename,encoding='utf-8')) root = tree.getroot() namespaces = {'kml': 'http://www.opengis.net/kml/2.2'} placemarks = root.findall(&quot;.//{kml}Placemark&quot;, namespaces) for placemark in placemarks: ##LIST IS EMPTY!!! name_text = placemark.findtext('.//name') return #file structure is &lt;kml&gt;&lt;Document&gt;&lt;Placemark&gt;&lt;...&gt;&lt;/Placemark&gt;&lt;/Document&gt;&lt;/kml&gt; I seem unable gto post the actual kml file, keeps complaining its not proper code </code></pre>
<python><kml>
2024-04-15 10:13:04
1
381
user1791983
78,327,547
6,197,439
pyqt5 change QMessageBox when using multiprocessing (Cannot set parent, new parent is in a different thread)?
<p>Here is an example, based on the example here <a href="https://stackoverflow.com/questions/48262966/terminate-a-long-running-python-command-within-a-pyqt-application/78325636#78325636">Terminate a long-running Python command within a PyQt application</a> - but with QMessageBox:</p> <pre class="lang-python prettyprint-override"><code>import sys import time from PyQt5.QtWidgets import (QApplication, QMainWindow, QLabel, QDialog, QVBoxLayout, QPushButton, QDialogButtonBox, QMessageBox) from multiprocessing import Pool def mp_long_task(): for x in range(2): print('long task:', x) time.sleep(1) return 'finished' class MainWindow(QMainWindow): def __init__(self, pool): super().__init__() button = QPushButton(&quot;Start&quot;, self) button.clicked.connect(self.long_task) self.setGeometry(300, 300, 300, 200) self.show() self.pool = pool def long_task(self): msgBox = QMessageBox( QMessageBox.Information, &quot;Long task&quot;, &quot;long task...&quot;, QMessageBox.Cancel, parent=None ) orig_ipmsize = msgBox.iconPixmap().size() orig_ipm = msgBox.iconPixmap() msgBox.setIcon( QMessageBox.NoIcon ) def callback(msg): #print(msg) msgBox.setText(msg) msgBox.setIconPixmap(orig_ipm) # QObject::setParent: Cannot set parent, new parent is in a different thread self.pool.apply_async(mp_long_task, callback=callback) if msgBox.exec_() == QMessageBox.Cancel: pool.terminate() msgBox.done(0) print('terminated') if __name__ == '__main__': pool = Pool() app = QApplication(sys.argv) main = MainWindow(pool) app.exec_() </code></pre> <p>The <code>mp_long_task</code> is started when you click the Start button; terminating the <code>mp_long_task</code> with Cancel button actually works fine - but if you wait for the <code>mp_long_task</code> to complete on its own, then <code>callback</code> runs, and in it:</p> <ul> <li>The <code>msgBox.setText(msg)</code>, if its on its own (that is, the <code>msgBox.setIconPixmap...</code> line is commented), applies the text and the QMessageBox stays on screen - which is what I expect the behavior to be</li> <li>If the <code>msgBox.setIconPixmap(orig_ipm)</code> is uncommented and it runs, then a sort of a segfault occurs, and the whole app crashes with:</li> </ul> <pre class="lang-none prettyprint-override"><code>QObject::setParent: Cannot set parent, new parent is in a different thread QCoreApplication::postEvent: Unexpected null receiver </code></pre> <p>How should I set up this code so it works: that is, when the multiprocessing <code>.apply_async</code> called task completes on its own, how can I both change the text and change the icon pixmap, keeping the QMessageBox, without the application crashing?</p>
<python><pyqt5><python-multiprocessing>
2024-04-15 09:56:38
1
5,938
sdbbs
78,327,535
23,461,455
scikit-learn 1.1.3. import cannot import name 'METRIC_MAPPING64' in python
<p>I am trying to <code>import linear_model</code> from scikit-learn into my python code in vscode and get an unexpected error message.</p> <pre><code>import sklearn from sklearn import linear_model </code></pre> <p>the error:</p> <pre><code>cannot import name 'METRIC_MAPPING64' from 'sklearn.metrics._dist_metrics' </code></pre> <p>I am not trying to import these metrics, how to solve this?</p> <p>The scikit-learn version used is 1.1.3.</p>
<python><machine-learning><scikit-learn><linear-regression><scikits>
2024-04-15 09:54:33
1
1,284
Bending Rodriguez
78,327,520
188,331
Parameter 'function' of the transform datasets.arrow_dataset.Dataset._map_single couldn't be hashed properly, a random hash was used instead
<p>I have a function to pre-process the dataset before tokenization. Here is the code:</p> <pre><code>source_lang = &quot;en&quot; target_lang = &quot;fr&quot; def preprocess_dataset(examples): inputs = [example[source_lang] for example in examples[&quot;translation&quot;]] targets = [example[target_lang] for example in examples[&quot;translation&quot;]] model_inputs = tokenizer(inputs, return_tensors=&quot;pt&quot;, text_target=targets, padding=&quot;max_length&quot;, max_length=200, truncation=True) return model_inputs </code></pre> <p>and to call the function using <code>.map()</code>:</p> <pre><code>tokenized_dataset = dataset.map(preprocess_dataset, batched=True) </code></pre> <p>The warning message is shown:</p> <pre><code>Parameter 'function'=&lt;function preprocess_dataset at 0x7f968c4d4790&gt; of the transform datasets.arrow_dataset.Dataset._map_single couldn't be hashed properly, a random hash was used instead. Make sure your transforms and parameters are serializable with pickle or dill for the dataset fingerprinting and caching to work. If you reuse this transform, the caching mechanism will consider it to be different from the previous calls and recompute everything. This warning is only showed once. Subsequent hashing failures won't be showed. </code></pre> <p>What have I done wrong? How can I avoid this warning?</p> <p>p.s. the subsequent training process can be done even if the warning appears.</p> <p>p.s. the dataset is in JSON strings format, example data is as follows:</p> <pre><code>{&quot;translation&quot;: {&quot;en&quot;: &quot;Hello&quot;, &quot;fr&quot;: &quot;Bonjour&quot;}} {&quot;translation&quot;: {&quot;en&quot;: &quot;Enjoy your meal&quot;, &quot;fr&quot;: &quot;bon appétit&quot;}} {&quot;translation&quot;: {&quot;en&quot;: &quot;Please&quot;, &quot;fr&quot;: &quot;s'il te plaît&quot;}} </code></pre>
<python><huggingface-tokenizers>
2024-04-15 09:51:30
0
54,395
Raptor
78,327,471
6,775,670
How to convert pydantic model to python dataclass?
<p>There the opposite is very popular question. Why not to convert a ready pydantic model into a standard python library dataclass?</p>
<python><pydantic><python-dataclasses>
2024-04-15 09:43:59
1
1,312
Nikolay Prokopyev
78,327,423
10,357,604
Naming of file while saving with extension
<p>How can I save a file with the name as: oldname+&quot;_new&quot;+extension in an elegant way?</p> <p>I currently do:</p> <pre><code>ext = os.path.splitext(file)[1] output_file = (root+'/'+ os.path.splitext(file)[0]+&quot;_new&quot;+ext) #or output_file = (os.path.join(root, os.path.splitext(file)[0])+'_new'+ext) with open(output_file, 'w', encoding='utf-8') as file: file.write(text) </code></pre> <p>(with the help of <a href="https://stackoverflow.com/questions/38596511/python-how-to-retain-the-file-extension-when-renaming-files-with-os">Python: how to retain the file extension when renaming files with os?</a> , but not duplicate)</p> <p>-- I use Python 3.12.2</p>
<python><file><os.path><code-readability>
2024-04-15 09:35:21
1
1,355
thestruggleisreal
78,327,204
18,380,493
Changing the input layout of an ONNX model from NCHW to NHWC
<p>I have an ONNX model with an input shape of <code>(1, 1, 28, 28)</code>, which is in an <code>NCHW</code> layout. Is there any way that I can convert the model to take input data in the <code>NHWC</code> layout, i.e., as <code>(1, 28, 28, 1)</code>?</p> <p>The model is loaded from a file using the following code: <code>model_obj = onnx.load(model_path)</code></p> <p>A method to modify and convert the ONNX file can also be helpful.</p>
<python><deep-learning><onnx><onnxruntime>
2024-04-15 08:56:06
1
326
moriaz
78,327,110
13,942,929
Install GMP library on Mac OS and PyCharm
<p>I'm trying to run my Cython project. And one of the header is <code>gmpxx.h</code>.</p> <p>Even though I already installed the gmp library using <code>brew install gmp</code>. I could not run my cython file with <code> python3 setup.py build_ext --inplace</code>.</p> <pre><code>fatal error: 'gmpxx.h' file not found #include &lt;gmpxx.h&gt; ^~~~~~~~~ 1 error generated. error: command '/usr/bin/clang' failed with exit code 1 </code></pre> <p>So I use <code>brew list gmp</code> to check the location of the <code> gmpxx.h</code> header. So it is actually inside <code>/opt/homebrew/Cellar/gmp/6.3.0/include/</code> folder.</p> <p>With <code>Xcode</code>, I can just add the location into the <code>header search paths</code>. But I'm trying to do the same thing with <code>Pycharm</code>. How do I add the location of my gmpxx.h header to pycharm?</p> <p>I need a little help. Please kindly give me your take. Thank you.</p>
<python><c++><pycharm><cython><gmp>
2024-04-15 08:39:57
2
3,779
Punreach Rany
78,327,057
9,381,746
Euler's identity in python
<p>I am trying to calculate the Euler's in python with the following code:</p> <pre><code>import numpy as np import cmath x = 0 y = 1 z=complex(x,y) out=complex(np.e**(np.pi*z.imag)) print(out) </code></pre> <p>But what I am getting is the following (and I should get -1 or something close to it with floating point errors)</p> <pre><code>$ python code.py (23.140692632779263+0j) </code></pre>
<python><math><complex-numbers>
2024-04-15 08:31:11
1
5,557
ecjb
78,327,045
3,042,018
Appending to Path to Access Parent Directory in PyCharm vs Terminal
<p>Why is it that</p> <pre><code>import sys sys.path.append(&quot;..&quot;) from file_in_parent_dir import a_function </code></pre> <p>works in PyCharm, but not in when I run the file from CMD or PowerShell?</p> <p>(Windows 10, Python 11.3.3)</p> <p><code>ModuleNotFoundError: No module named 'file_in_parent_dir'</code></p> <p>Instead I have to use</p> <pre><code>import sys sys.path.insert(0, '..') from trace_recursion import trace </code></pre> <p>(which also works on Linux, with <code>&quot;./&quot;</code>)</p> <p>Could someone please explain the difference between the two approaches, and why PyCharm allows the first?</p>
<python><path><pycharm>
2024-04-15 08:29:01
2
3,842
Robin Andrews
78,326,981
447,738
Pandas to find a streak of sales but with a tolerance for streak interruption
<p>I am trying to use Panda's to filter out a DataFrame to find a streak of sales that shall be of variant length. I am looking to identify periods of consecutive sales, but also allow for some days with no sale in between. In the df below, I am looking to select rows 12 to 19.</p> <pre><code>data = [['2023-11-16', 1], ['2023-11-17', 0], ['2023-11-20', 1], ['2023-11-21', 0], ['2023-11-22', 1], ['2023-11-24', 0], ['2023-11-27', 0], ['2023-11-28', 0], ['2023-11-29', 0], ['2023-11-30', 0], ['2023-12-01', 0], ['2023-12-04', 0], ['2023-12-05', 1], ['2023-12-06', 1] , ['2023-12-07', 1], ['2023-12-08', 1], ['2023-12-11', 0], ['2023-12-12', 0], ['2023-12-13', 1], ['2023-12-14', 1], ['2023-12-15', 0], ['2023-12-18', 0], ['2023-12-19', 0], ['2023-12-20', 0]] df = pd.DataFrame(data, columns=['date', 'sold']) </code></pre> <p><a href="https://i.sstatic.net/rA7Aw.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/rA7Aw.png" alt="enter image description here" /></a></p>
<python><pandas>
2024-04-15 08:16:51
1
2,357
cksrc
78,326,962
5,043,301
Using UserCreationForm
<p>I am learning Django. I would like to understand below code.</p> <pre><code>class RegisterPage(FormView): template_name = 'base/register.html' form_class = UserCreationForm redirect_authenticated_user = True success_url = reverse_lazy('tasks') def form_valid( self, form ): user = form.save() if user is not None: login( self.request, user ) return super( RegisterPage, self ).form_valid( form ) def get( self, *args, **kwargs ): if self.request.user.is_authenticated: return redirect('tasks') return super(RegisterPage, self).get(*args, **kwargs) </code></pre> <p>Can I use <code> def register( self, form ):</code> instead of <code> def form_valid( self, form ):</code>?</p> <p>How <code>self</code> and <code>form</code> are passing here?</p> <p>Can I use <code> def save( self, *args, **kwargs ):</code> instead of <code> def get( self, *args, **kwargs ):</code>?</p>
<python><django>
2024-04-15 08:12:54
1
7,102
abu abu
78,326,924
4,429,265
Qdrant takes 25 second to retrieve 1000 single products
<p>We have a single-product API that is causing us problems when called in large numbers (not even very large; 1000 requests take 25 seconds to finish).</p> <h2>What Are We Using</h2> <p>So, we are using Qdrant for our product database because of the semantic vector search capabilities. But also, we want to use it for the single-product API, which does nothing other than taking the ID and giving back the payload. When called with a single ID, it takes 70-90 ms to finish. But calling 1000 concurrent requests causes it to slow and return responses to all of them in 25-30 seconds. And while these 1000 requests are being processed, if we call it with one single ID, this time it takes 2500 ms (from 70-90 went to 2500 ms).</p> <h2>The API</h2> <p>This is the single product api:</p> <pre><code>def retrieve(self, request, pk): search_lang = request.query_params.get(&quot;language&quot;, &quot;tr&quot;) country = request.query_params.get(&quot;country&quot;, &quot;tr&quot;) payload = search.get_single_product(search_lang=search_lang, point_id=int(pk)) payload[&quot;extra&quot;] = functions.generate_extra_fields( product_id=payload[&quot;id&quot;], total_rating_count=payload[&quot;stats&quot;][&quot;total_rating_count&quot;], lang=search_lang, country=country, ) return Response( {&quot;data&quot;: payload}, status=status.HTTP_200_OK, ) </code></pre> <h2>The search function</h2> <p>And this is the <code>get_single_product</code> function:</p> <pre><code> def get_single_product(search_lang: str, point_id: int): &quot;&quot;&quot; Retrieves a single product based on the provided language and product ID. Parameters: search_lang (str): The language code for the product. point_id (int): The ID of the product to retrieve. Returns: dict: A dictionary containing the details of the retrieved product. Note: - This function retrieves the language code from the environment variable LANGUAGE_KEY_MAP. - It uses the Qdrant client to retrieve the product from the appropriate language collection. - If the provided search language is not supported, it returns an empty dictionary. &quot;&quot;&quot; lang = int(LANGUAGE_KEY_MAP.get(search_lang)) qd_client = qdrant_client() if not lang: return 0 search_results = qd_client.retrieve( collection_name=f&quot;lang{lang}_products&quot;, ids=[point_id], ) qd_client.close() payloads = [] for search_result in search_results: payloads.append(search_result.payload) return payloads[0] if len(payloads) &gt; 0 else [] </code></pre> <h2>How I tested with 1000 requests:</h2> <p>For testing, I first need to have the ids of 1000 products. Our main relational DB is pg, so I first retrieve the ids, then create async requests; as below:</p> <p><code>test.py</code>:</p> <pre><code>import aiohttp import asyncio from utils.functions import pg_connect import time pg_start = time.time() pg_connection, pg_cursor = pg_connect() query = &quot;SELECT id from public.products limit 1000;&quot; # Execute the query pg_cursor.execute(query) # Fetch IDs ids = [row[0] for row in pg_cursor.fetchall()] pg_connection.close() pg_cursor.close() print(f&quot;it took {time.time() - pg_start} seconds to get ids from pg.\n&quot;) async def fetch(session, url): async with session.get(url) as response: return await response.text() async def fetch_all(urls): async with aiohttp.ClientSession() as session: task_start = time.time() tasks = [fetch(session, url) for url in urls] print(f&quot;tasks created at {time.time() -task_start} seconds. \n&quot;) return await asyncio.gather(*tasks) async def main(ids): base_url = &quot;https://api2.markabu.com/v1/product/&quot; url_start = time.time() urls = [f&quot;{base_url}{id}/&quot; for id in ids] print(f&quot;urls created in {time.time() - url_start} seconds.\n&quot;) request_start = time.time() responses = await fetch_all(urls) print(f&quot;responses came in {time.time() - request_start} seconds.\n&quot;) if __name__ == &quot;__main__&quot;: asyncio.run(main(ids)) </code></pre> <h2>The results</h2> <p>I ran the code a few times, the results are within 5% of each other, one example:</p> <pre><code>it took 0.00798797607421875 seconds to get ids from pg. urls created in 0.00010538101196289062 seconds. tasks created at 0.00011730194091796875 seconds. responses came in 26.443119525909424 seconds. </code></pre> <h2>What is the problem?</h2> <p>So, there are different areas which may cause the problem. Before blaming Qdrant or my own data structure, I thought about web server issues which may have problems distributing a large (fairly large) number of requests. That is why I created a simple API that just returns a simple message. Then called it even more than 100,000 times, and for almost 130,000 calls, it took only 8-9 seconds.</p> <p>So, the problem is not with the web server, nor the Django backend.</p> <h2>What about data structure?</h2> <p>One other thing to consider is that our products have 30 columns (sorry for using 'column' to refer to JSON structured data in Qdrant, but you know what I mean), some of which are fairly large JSON fields (such as <code>variants</code>, which for some products may be a list of 10-20 variants, each containing data about different variants of the product).</p> <h2>Can Qdrant actually handle large amounts of concurrent query requests?</h2> <p>Another possibility to consider is Qdrant's limitations. I could not find any note about concurrent requests in Qdrant. Or, perhaps the problem lies in how I handle the <code>qdrant_client</code>. I open the client using a simple function called in the <code>get_single_product</code> function above:</p> <pre><code>qdrant_client = QdrantClient( host=kwargs.get(&quot;host&quot;), port=kwargs.get(&quot;port&quot;), timeout=kwargs.get(&quot;timeout&quot;), ) return qdrant_client </code></pre> <p>And after using it to get the product, I close the connection (refer to <code>get_single_product</code> function above).</p> <h1>UPDATE</h1> <h2>Details about Qdrant setup</h2> <p>I am hosting Qdrant on a docker container on my server, with 30 cores of CPU and 50 GBs of RAM. This is the docker-compose for the container:</p> <pre><code> qdrant: image: qdrant/qdrant:v1.8.2 restart: always ports: - &quot;6333:6333&quot; - &quot;6334:6334&quot; volumes: - qdrant_data:/qdrant/storage </code></pre> <p>Also my qdrant collections are created in this manner:</p> <pre><code>client.recreate_collection( collection_name=collection_name, vectors_config=models.VectorParams( size=encoder.get_sentence_embedding_dimension(), distance=models.Distance.DOT, ), ) </code></pre> <p>Which I do not think is relevant to the problem, but included here if needed. If any more details on the structure of the project is needed, please ask.</p>
<python><database><qdrant>
2024-04-15 08:07:03
1
417
Vahid
78,326,802
18,730,707
"DevTools listening on ws://127.0.0.1" message does not go away in python selenium
<p>I know this question is a duplicate. However, no matter how much I write, the following phrase will not disappear.</p> <blockquote> <p>DevTools listening on ws://127.0.0.1:62784/devtools/browser/9f06f86e-f98b-4896-9f35-8c5c73317c7a</p> </blockquote> <p>There is a reason why the above phrase should be deleted. You need to import usd-to-krw, kospi, and nasdaq indices from external sources and print them to the console. And I copy that console and use it. But the above correctly linked statement is stuck in the middle of my entire text. I don't want the above text to be printed because it's inconvenient to erase it every time.</p> <p>Here is the full code I used:</p> <pre class="lang-py prettyprint-override"><code>from selenium import webdriver from selenium.webdriver.chrome.service import Service from selenium.webdriver.chrome.options import Options from webdriver_manager.chrome import ChromeDriverManager from bs4 import BeautifulSoup import os def get_usd_to_krw_exchange_rate(): service = Service( executable_path = ChromeDriverManager().install(), log_path=os.devnull ) options = Options() options.add_experimental_option('excludeSwitches', ['enable-logging']) options.add_argument('--headless') options.add_argument('--log-level=3') options.add_argument('--disable-logging') options.add_argument('--disable-dev-shm-usage') driver = webdriver.Chrome( service = service, options = options ) driver.get('https://www.google.com/search?q=usd+to+krw') html = driver.page_source soup = BeautifulSoup(html, 'html.parser') rate = soup.select_one('#knowledge-currency__updatable-data-column &gt; div.b1hJbf &gt; div.dDoNo.ikb4Bb.gsrt &gt; span.DFlfde.SwHCTb').get_text() print(rate) get_usd_to_krw_exchange_rate() </code></pre> <p>Please let me know what more I should do or what I'm doing wrong.</p>
<python><google-chrome><selenium-webdriver>
2024-04-15 07:41:25
0
878
na_sacc
78,326,717
2,695,082
ImportError: tools/python/lib/python3.9/lib-dynload/math.so: undefined symbol: PyFloat_Type
<p>I am using Python 3.9 on Linux. The python installation structure looks like:</p> <pre><code>/tools/python/bin - containing python executable /tools/python/lib/libpython3.9.so /tools/python/lib/python3.9/lib-dynload/* </code></pre> <p>I am linking a x.so with libpython.so and trying to run it. And when I run a simple python script test.py(present in /share/python) that looks like:</p> <pre><code>import unittest import math class Test_2(unittest.TestCase): def test_distance(self): self.assertAlmostEqual(box1.distance(box2), math.sqrt(2)) def suite() : suite = unittest.TestSuite() suite.addTest(unittest.makeSuite(Test_2)) return suite </code></pre> <p>It simply gives the following error in the log:</p> <pre><code>ImportError: Failed to import test module: test Traceback (most recent call last): File &quot;/share/python/test.py&quot;, line 3, in &lt;module&gt; import math ImportError: /tools/python/lib/python3.9/lib-dynload/math.so: undefined symbol: PyFloat_Type </code></pre> <p>I have set PYTHONPATH as : /share/python and PYTHONHOME as : /tools/python And appended LD_LIBRARYPATH with /tools/python/lib so that it finds libpython3.9.so But still get the above import error: &quot;math.so: undefined symbol: PyFloat_Type&quot; Can someone pls help me in this issue.</p> <p>Following is the output:</p> <pre><code>[abc@xyz01 ~]$ ldd /tools/python/bin/python linux-vdso.so.1 (0x00007ffc281ce000) libpython3.9.so.1.0 =&gt; /tools/python/lib/libpython3.9.so.1.0 (0x00007f28887d8000) libcrypt.so.1 =&gt; /lib64/libcrypt.so.1 (0x00007f28885af000) libpthread.so.0 =&gt; /lib64/libpthread.so.0 (0x00007f288838f000) libdl.so.2 =&gt; /lib64/libdl.so.2 (0x00007f288818b000) libutil.so.1 =&gt; /lib64/libutil.so.1 (0x00007f2887f87000) libm.so.6 =&gt; /lib64/libm.so.6 (0x00007f2887c05000) libc.so.6 =&gt; /lib64/libc.so.6 (0x00007f2887840000) /lib64/ld-linux-x86-64.so.2 (0x00007f2888b6a000) [abc@xyz01 ~]$ echo $PYTHONPATH /share/python [abc@xyz01 ~]$ echo $PYTHONHOME /tools/python [abc@xyz01 bin]$ echo $LD_LIBRARY_PATH tools/python/lib:/build/tools/libstdc+/lib64:/build/tools/lib/64bit:/build/tools/lib:/build81/tools/Qt/v5/64bit/lib:/build/tools/boost/lib/64bit/:/usr/lib64:/build/tools/lib/64bit/RHEL/RHEL8:/build/tools/lib/64bit/RHEL/RHEL7:/build81/tools/libstdc++/lib64:/build81/tools/lib/64bit:/usr/X11R6/lib:/lib:/usr/lib </code></pre>
<python><linux><python-3.9>
2024-04-15 07:26:37
0
329
user2695082
78,326,712
3,405,291
No module named pip
<h1>Conda environment</h1> <p>I'm creating the following conda environment by <code>conda env create -f environment.yml</code>.</p> <p>The <code>environment.yml</code> file content is:</p> <pre class="lang-yaml prettyprint-override"><code>name: deep3d_pytorch channels: - pytorch - conda-forge - defaults dependencies: - python=3.6 - pytorch=1.6.0 - torchvision=0.7.0 - numpy=1.18.1 - scikit-image=0.16.2 - scipy=1.4.1 - pillow=6.2.1 - pip=20.0.2 - ipython=7.13.0 - yaml=0.1.7 - pip: - matplotlib==2.2.5 - opencv-python==3.4.9.33 - tensorboard==1.15.0 - tensorflow==1.15.0 - kornia==0.5.5 - dominate==2.6.0 - trimesh==3.9.20 </code></pre> <h1>Error</h1> <p>At the end of the environment creation process, these errors are thrown:</p> <pre><code>Installing pip dependencies: \ Ran pip subprocess with arguments: ['/usr/local/envs/deep3d_pytorch/bin/python', '-m', 'pip', 'install', '-U', '-r', '/content/Deep3DFaceRecon_pytorch/condaenv.i1gomfsb.requirements.txt', '--exists-action=b'] Pip subprocess output: Pip subprocess error: /usr/local/envs/deep3d_pytorch/bin/python: No module named pip failed CondaEnvException: Pip failed </code></pre> <p>In addition to <code>conda env create -f environment.yml</code>, updating the conda environment by <code>conda env update -f environment.yml</code> would throw similar errors.</p> <h1>Note</h1> <p>The errors occur on my local machine and also on the Google Colab. I'm just following the instructions. Does anybody have any clue or hint?</p> <p>I have looked at this, but couldn't figure it out: <a href="https://stackoverflow.com/questions/41060382/using-pip-to-install-packages-to-anaconda-environment">Using Pip to install packages to Anaconda Environment</a></p>
<python><pip><anaconda><conda><google-colaboratory>
2024-04-15 07:25:35
1
8,185
Megidd
78,326,697
10,722,752
Trend and Residue plots in seasonal decompose not spanning entire duration of data
<p>I am performing SARIMAX forecasting and visualizing the seasonal decompose plots. I have monthly data running from Jan 1, 2022 to March 1, 2024. When I plot my data I am getting a &quot;partial&quot; trend and residue plots.</p> <p>Sample Data:</p> <pre><code>from statsmodels.tsa.seasonal import seasonal_decompose from matplotlib import pyplot np.random.seed(0) df = pd.DataFrame(index=pd.date_range('2022-01-01', end='2024-03-01', freq='MS'), data = {'rev' : np.random.randint(10, high = 1000, size = len(pd.date_range('2022-01-01', end='2024-03-01', freq='MS')))}) seasonal_decompose(df).plot(); </code></pre> <p><a href="https://i.sstatic.net/hHpCq.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/hHpCq.png" alt="enter image description here" /></a></p> <p>But when I do the same for data with longer time span, the plot seems to better span the time line.</p> <pre><code>df1 = pd.DataFrame(index=pd.date_range('2015-01-01', end='2024-03-01', freq='MS'), data = {'rev' : np.random.randint(10, high = 1000, size = len(pd.date_range('2015-01-01', end='2024-03-01', freq='MS')))}) seasonal_decompose(df1).plot(); </code></pre> <p><a href="https://i.sstatic.net/HLhB5.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/HLhB5.png" alt="enter image description here" /></a></p> <p>Could someone please help me understand if my understanding is right, if so, what would be the way to understand the trend plot where we don't have data for sufficiently long duration.</p>
<python><pandas><statsmodels><sarimax>
2024-04-15 07:22:39
1
11,560
Karthik S
78,326,633
11,503,237
Issue with generating poisson arrivals in a system
<p>I am simulating the poisson arrivals of requests in the system. I am using numpy poisson for this purpose. The code is given below:</p> <pre><code>no_of_req=500 timesteps =500 rate = no_of_req / timesteps id=1 poisson=[] idx=[] for time_step in range(1, timesteps + 1): number_of_new_agents = np.random.poisson(rate) time=update_time(start_time, time_step).time() poisson.append(number_of_new_agents) idx.append(time_step) plt.bar(idx, poisson) </code></pre> <p>When I plot it I get the following bar</p> <p><a href="https://i.sstatic.net/H5iRQ.jpg" rel="nofollow noreferrer"><img src="https://i.sstatic.net/H5iRQ.jpg" alt="enter image description here" /></a></p> <p>This doesnot resemble a Poisson Distribution. Where is the problem?</p>
<python><numpy><poisson>
2024-04-15 07:11:32
1
483
Hamsa
78,326,625
13,687,718
Handle continuous parallel requests without blocking in Asyncio
<p>I am new to python asyncio and have a slightly convoluted requirement. I have been going through the documentation for asyncio but have not found the right solution yet as I am unable to understand some aspects.</p> <p>I have a script <strong>KafkaScript.py</strong> that reads from a kafka topic (streaming) and calls an API endpoint. The API will call a method <strong>method_A</strong> in another script <strong>ProcessingScript.py</strong> in <em>parallel</em> for each incoming request/kafka record.</p> <p><em>method_A</em> calls an asynchronous <em>method_B</em> which uses a globally defined ThreadPoolExecutor to call another asynchronous <em>method_C</em> using <em>run_in_executor()</em>. <em>method_C</em> return a string back to <em>method_B</em> which in turn must respond back to <em>method_A</em>.</p> <p>Here is the script:</p> <p>KafkaScript.py:</p> <pre><code>import requests if __name__ == '__main__': for i in Kafka-Topic: # Use a new thread to call make API request for each kafka record response = requests.get(&lt;url generated&gt;) # i is passed as param </code></pre> <p>The API:</p> <pre><code>import ProcessingScript @app.route('/v1/generate', methods = ['GET']) def generate(): data = request.args.get('data', type = str) response = ProcessingScript.method_A(data) return response </code></pre> <p>ProcessingScript.py:</p> <pre><code>executor = ThreadPoolExecutor(max_workers=5) async def method_C(val): # Calls playwright to get content from a page which is a blocking call per thread return &quot;processed:&quot; + val async def method_B(val): loop = asyncio.get_event_loop() # For each parallel thread, call method_C, get the response without blocking other threads response = loop.run_in_executor(self.executor, asyncio.run, method_C(val)) return response def method_A(val): response = asyncio.run(method_B(val)) return response </code></pre> <p>As I understand, <em>method_B</em> can receive the string from <em>method_C</em> only when I make a call such as this.</p> <pre><code>response = await loop.run_in_executor(self.executor, asyncio.run, method_C(val)) </code></pre> <p>The problem with the <em>await</em> here is, although multiple threads are calling <em>method_A</em>, the <em>await</em> makes all threads wait for the first thread to have a response.</p> <p>How do I ensure this in a thread safe manner where none of the parallel requests to <em>method_A</em> are blocked by others threads waiting for the response?</p>
<python><multithreading><python-asyncio>
2024-04-15 07:09:50
1
832
mang4521
78,326,591
11,770,390
pd.date_range with enddate included
<p>Using pandas 2.2.2 and python 3.11, why doesn't this give me the daterange containing the end date:</p> <pre><code>import pandas as pd start_date = pd.to_datetime('2023-04-05T04:01:40Z') end_date = pd.to_datetime('2024-04-15T00:00:00Z') full_date_range = pd.date_range(start=start_date, end=end_date, freq='1M', inclusive='both') for date in full_date_range.to_list(): print(date) </code></pre> <p>It gives me:</p> <pre><code>2023-04-30 04:01:40+00:00 2023-05-31 04:01:40+00:00 2023-06-30 04:01:40+00:00 2023-07-31 04:01:40+00:00 2023-08-31 04:01:40+00:00 2023-09-30 04:01:40+00:00 2023-10-31 04:01:40+00:00 2023-11-30 04:01:40+00:00 2023-12-31 04:01:40+00:00 2024-01-31 04:01:40+00:00 2024-02-29 04:01:40+00:00 2024-03-31 04:01:40+00:00 </code></pre> <p>and it stops at the end of march. I thought inclusive would give me an added entry of</p> <pre><code>2024-04-30 04:01:40+00:00 </code></pre> <p>as this would include my enddate as well (at least this is how I understand <a href="https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.date_range.html" rel="nofollow noreferrer">the <code>inclusive</code> argument</a>)</p>
<python><pandas><dataframe><timestamp>
2024-04-15 07:03:30
3
5,344
glades
78,326,573
2,508,672
Convert any format of date string to datetime
<p>I am reading csv file and in date column it has different format some rows using m/d/yyyy some using d-m-yyyy and others</p> <p>I am using below code to convert</p> <pre><code>if(datetime.strptime(row['Date'],&quot;%m/%d/%Y&quot;).astimezone() &lt;= datetime.strptime(&quot;2024-03-03&quot;,&quot;%Y-%m-%d&quot;).astimezone()): ## do the things </code></pre> <p>if the format matches it works, but it throws error when it does not match</p> <blockquote> <p>time data '2022-05-16' does not match format '%m/%d/%Y'</p> </blockquote> <p>How can I convert string date of any format to a valid date</p> <p>Thanks</p>
<python>
2024-04-15 06:59:07
2
4,608
Md. Parvez Alam
78,326,265
2,739,700
Azure alerting Creation using Python SDK
<p>We are creating Azure alert using python SDK for Custom KQL query and below is the code which we are trying</p> <pre><code>from azure.mgmt.monitor import MonitorManagementClient from azure.identity import DefaultAzureCredential credentials = DefaultAzureCredential() sub_id= &quot;xxxx&quot; resource_group = &quot;rgname&quot; client = MonitorManagementClient(credential=credentials,subscription_id=sub_id) # Define the KQL query kql_query = &quot;&quot;&quot; ServiceMonitor | where Computer contains &quot;xyz-host&quot; | where SvcName contains &quot;linux-service&quot; | where SvcState != &quot;Running&quot; &quot;&quot;&quot; # Create a unique name for the metric alert metric_alert_name = &quot;KQL_Metric_alert&quot; # Create the alert based on the KQL query Alert = client.metric_alerts.create_or_update(resource_group, metric_alert_name, # Use the unique metric alert name { &quot;location&quot;: &quot;global&quot;, &quot;description&quot;: &quot;Alert triggered when the service is not running on xyz_host&quot;, &quot;severity&quot;: &quot;3&quot;, &quot;enabled&quot;: True, &quot;scopes&quot;: [ f&quot;/subscriptions/{sub_id}/resourceGroups/{resource_group}&quot; ], &quot;evaluation_frequency&quot;: &quot;PT1M&quot;, &quot;window_size&quot;: &quot;PT15M&quot;, &quot;target_resource_type&quot;: &quot;Microsoft.Insights/components&quot;, &quot;target_resource_region&quot;: &quot;global&quot;, &quot;criteria&quot;: { &quot;odata.type&quot;: &quot;Microsoft.Azure.Monitor.MultipleResourceMultipleMetricCriteria&quot;, &quot;all_of&quot;: [ { &quot;criterion_type&quot;: &quot;StaticThresholdCriterion&quot;, &quot;metric_name&quot;: &quot;KQLQueryExecutionResult&quot;, &quot;metric_namespace&quot;: &quot;&quot;, &quot;operator&quot;: &quot;GreaterThan&quot;, &quot;threshold&quot;: 0, &quot;aggregation&quot;: &quot;Count&quot;, &quot;dimensions&quot;: [], &quot;metric_namespace&quot;: &quot;microsoft.insights/components&quot;, &quot;additional_properties&quot;: { &quot;query&quot;: kql_query } } ] }, &quot;auto_mitigate&quot;: False, &quot;actions&quot;: [ { &quot;action_group_id&quot;: f&quot;/subscriptions/{sub_id}/resourceGroups/rg-name/providers/microsoft.insights/actionGroups/AlertTrigger&quot;, &quot;webhook_properties&quot;: {} } ] } ) print(&quot;Alert created Successfully.&quot;) </code></pre> <p>We get below error:</p> <pre><code>Traceback (most recent call last): File &quot;/Users/testuser/Documents/test/alerting/service_check.py&quot;, line 21, in &lt;module&gt; Alert = client.metric_alerts.create_or_update(resource_group, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;/Users/vcimalap/Documents/test/alerting/.venv/lib/python3.11/site-packages/azure/core/tracing/decorator.py&quot;, line 78, in wrapper_use_tracer return func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File &quot;/Users/vcimalap/Documents/test/alerting/.venv/lib/python3.11/site-packages/azure/mgmt/monitor/v2018_03_01/operations/_metric_alerts_operations.py&quot;, line 609, in create_or_update raise HttpResponseError(response=response, model=error, error_format=ARMErrorFormat) azure.core.exceptions.HttpResponseError: (BadRequest) Required property 'name' not found in JSON. Path '', line 1, position 512. Activity ID: df779cd2-198b-4098-be4f-c68d95c27f00. Code: BadRequest Message: Required property 'name' not found in JSON. Path '', line 1, position 512. Activity ID: df779cd2-198b-4098-be4f-c68d95c27f00. </code></pre> <p>We are not sure whats the wrong with code and anybody fixing issue will be great</p> <p>Python version: 3.11</p> <p>package versions:</p> <pre><code>azure-common==1.1.28 azure-core==1.30.1 azure-identity==1.16.0 azure-mgmt-core==1.4.0 azure-mgmt-monitor==6.0.2 certifi==2024.2.2 cffi==1.16.0 charset-normalizer==3.3.2 cryptography==42.0.5 idna==3.7 isodate==0.6.1 msal==1.28.0 msal-extensions==1.1.0 packaging==24.0 portalocker==2.8.2 pycparser==2.22 PyJWT==2.8.0 requests==2.31.0 six==1.16.0 typing_extensions==4.11.0 urllib3==2.2.1 </code></pre>
<python><azure><azure-monitoring><azure-alerts>
2024-04-15 05:41:23
1
404
GoneCase123
78,325,944
7,584,138
How to consolidate slowing changing dimension tables using sql, python or r?
<p>I have below input table:</p> <div class="s-table-container"><table class="s-table"> <thead> <tr> <th>id</th> <th>type</th> <th>value</th> <th>date_from</th> <th>date_to</th> </tr> </thead> <tbody> <tr> <td>1</td> <td>department</td> <td>finance</td> <td>2020-01-01</td> <td>9999-12-31</td> </tr> <tr> <td>1</td> <td>headcount</td> <td>10</td> <td>2020-01-01</td> <td>2020-02-03</td> </tr> <tr> <td>1</td> <td>headcount</td> <td>15</td> <td>2020-02-04</td> <td>9999-12-31</td> </tr> <tr> <td>1</td> <td>location</td> <td>DC</td> <td>2020-01-01</td> <td>2020-01-21</td> </tr> <tr> <td>1</td> <td>location</td> <td>NY</td> <td>2020-01-22</td> <td>9999-12-31</td> </tr> </tbody> </table></div> <p>I want to convert it to a wide table containing all &quot;type&quot; fields as columns:</p> <div class="s-table-container"><table class="s-table"> <thead> <tr> <th>id</th> <th>department</th> <th>headcount</th> <th>location</th> <th>date_from</th> <th>date_to</th> </tr> </thead> <tbody> <tr> <td>1</td> <td>finance</td> <td>10</td> <td>DC</td> <td>2020-01-01</td> <td>2020-01-21</td> </tr> <tr> <td>1</td> <td>finance</td> <td>10</td> <td>NY</td> <td>2020-01-22</td> <td>2020-02-03</td> </tr> <tr> <td>1</td> <td>finance</td> <td>15</td> <td>NY</td> <td>2020-02-04</td> <td>9999-12-31</td> </tr> </tbody> </table></div> <p>Note that my actual data has multiple &quot;id&quot; and unknown &quot;type&quot; values. How can I achieve this effectively using either <code>sql</code>, <code>python</code> or <code>r</code>?</p> <p><em>Edit</em></p> <p>The final output represents a historic table which will enable me to go back to any given day to view the status on that day.</p> <p>In my example, &quot;department&quot; is always &quot;finance&quot;. &quot;headcount&quot; and &quot;location&quot; have changed, so that there are two new rows.</p>
<python><sql><pandas><dplyr><tidyverse>
2024-04-15 03:24:01
1
1,688
Frank Zhang
78,325,675
1,202,417
Serving a webpage content by running a php/python script
<p>I'm trying to set up a RSS for my site. So I would like to make a link that takes in a keyword and produces a RSS feed.</p> <p>I have a python script (<code>script.py</code>) to generate this xml, but I don't know how to run it and serve the text to the user when my page is called.</p> <p>Essentially I would like to have someone visit <code>mysite.com/&lt;keyword&gt;</code> and be served the text generated in <code>script.py</code></p> <p>I can make the text appear to the user by simply running javascript, but this isn't being picked up by the rss.</p> <pre><code>&lt;html&gt; &lt;head&gt; &lt;title&gt;Run Python Script&lt;/title&gt; &lt;script src=&quot;https://cdnjs.cloudflare.com/ajax/libs/jquery/3.5.1/jquery.min.js&quot;&gt;&lt;/script&gt; &lt;/head&gt; &lt;body&gt; &lt;script&gt; $.ajax({ method: &quot;POST&quot;, url: &quot;/script.py&quot;, }) .done(function( response ) { var newXML = document.open(&quot;text/xml&quot;, &quot;replace&quot;); newXML.write(response); newXML.close(); }); &lt;/script&gt; &lt;/body&gt; &lt;/html&gt; </code></pre> <p>It seams like there should be some way of generating this text and serving it to the user, but I feel like I'm just missing something obvious</p> <p>I'm using godaddy and cpanel if that helps</p>
<javascript><python><rss>
2024-04-15 00:55:23
0
411
Ben Fishbein
78,325,531
11,930,602
What is the standard way to insert out-of-order elements in a list of a possibly smaller size?
<p>What I have:</p> <pre class="lang-py prettyprint-override"><code>msg: list = [] </code></pre> <p>expected behaviour:</p> <pre class="lang-py prettyprint-override"><code>msg.insert(2,&quot;two&quot;) # msg = [None, None, &quot;two&quot;] msg.insert(10,&quot;ten&quot;) # msg = [None, None, &quot;two&quot;, None, None, None, None, None, None, None, &quot;ten&quot;] </code></pre> <p>Current code:</p> <pre class="lang-py prettyprint-override"><code>def insert_into_list(orig_list: list, index: int, element: str)-&gt;None: # does not need to return the list if index &gt;= len(orig_list): # sanity checks removed for MRE for _ in range(index - len(orig_list) + 1): orig_list.append(None) orig_list.insert(index, element) </code></pre> <p>Is there a better (or a more standard and possibly shorter than writing another method) way to do this?</p>
<python><list><insert>
2024-04-14 23:21:46
2
2,322
kesarling
78,325,356
6,197,439
How to setup an instance method monkeypatch in this PyQt5 code?
<p>Yes, I have seen:</p> <ul> <li><a href="https://stackoverflow.com/questions/28127874/monkey-patching-python-an-instance-method">monkey-patching python an instance method</a></li> <li><a href="https://stackoverflow.com/questions/8726238/how-to-call-the-original-method-when-it-is-monkey-patched">How to call the original method when it is monkey-patched?</a></li> </ul> <p>... but none of these approaches seem to work in my example:</p> <p>Basically, this is an example that tries to use <a href="https://pypi.org/project/QtAwesome/" rel="nofollow noreferrer">QtAwesome</a> to provide a spinning icon to a QMessageBox; there is a class <code>Spin</code> in QtAwesome that does animations, which has a method <code>_update</code>: I would like to monkeypatch this method, so first the original method is called, and then a message is printed - so I have this:</p> <pre class="lang-python prettyprint-override"><code>import sys from PyQt5.QtWidgets import QApplication, QWidget, QMessageBox import qtawesome as qta class Example(QWidget): def __init__(self): super().__init__() self.initUI() def initUI(self): self.setGeometry(300, 300, 300, 220) self.setWindowTitle('Hello World') self.show() msgBox = QMessageBox( QMessageBox.Information, &quot;Title&quot;, &quot;Content ...&quot;, QMessageBox.Cancel ) orig_ipmsize = msgBox.iconPixmap().size() print(orig_ipmsize.width(), orig_ipmsize.height()) # 0 0 for QMessageBox.NoIcon; 32 32 for QMessageBox.Information animation = qta.Spin(msgBox, autostart=True) DO_MONKEYPATCH = 2 # 1 or 2 if DO_MONKEYPATCH == 1: old_anim_update = animation._update def new_anim_update(self): old_anim_update(self) # TypeError: Spin._update() takes 1 positional argument but 2 were given print(&quot;new_anim_update&quot;) animation._update = new_anim_update.__get__(animation, qta.Spin) # https://stackoverflow.com/a/28127947 elif DO_MONKEYPATCH == 2: def update_decorator(method): def decorate_update(self=None): method(self) # TypeError: Spin._update() takes 1 positional argument but 2 were given print(&quot;decorate_update&quot;) return decorate_update animation._update = update_decorator(animation._update) # https://stackoverflow.com/a/8726680 #print(animation._update) spin_icon = qta.icon('fa5s.spinner', color='red', animation=animation) msgBox.setIconPixmap(spin_icon.pixmap(orig_ipmsize)) #msgBox.setIcon(spin_icon) #animation.start() returnValue = msgBox.exec() if __name__ == '__main__': app = QApplication(sys.argv) ex = Example() sys.exit(app.exec_()) </code></pre> <p>However, no matter which <code>DO_MONKEYPATCH</code> method I choose, I get <code>TypeError: Spin._update() takes 1 positional argument but 2 were given</code>?</p> <hr /> <p>OK, just noticed how both errors apply while writing the question, and found if I change the &quot;old method&quot; calls to NOT use a <code>(self)</code> argument -- i.e. I change <code>old_anim_update(self)</code> / <code>method(self)</code>, to <code>old_anim_update()</code> / <code>method()</code> -- then both <code>DO_MONKEYPATCH</code> methods allow for running without the positional argument error - however only <code>DO_MONKEYPATCH</code> method 1 seems to preserve <code>self</code>:</p> <pre class="lang-python prettyprint-override"><code>import sys from PyQt5.QtWidgets import QApplication, QWidget, QMessageBox import qtawesome as qta class Example(QWidget): def __init__(self): super().__init__() self.initUI() def initUI(self): self.setGeometry(300, 300, 300, 220) self.setWindowTitle('Hello World') self.show() msgBox = QMessageBox( QMessageBox.Information, &quot;Title&quot;, &quot;Content ...&quot;, QMessageBox.Cancel ) orig_ipmsize = msgBox.iconPixmap().size() print(orig_ipmsize.width(), orig_ipmsize.height()) # 0 0 for QMessageBox.NoIcon; 32 32 for QMessageBox.Information animation = qta.Spin(msgBox, autostart=True) DO_MONKEYPATCH = 1 # 1 or 2 if DO_MONKEYPATCH == 1: old_anim_update = animation._update def new_anim_update(self): old_anim_update() # no error print(&quot;new_anim_update {}&quot;.format(self)) # self is &lt;qtawesome.animation.Spin object at 0x00000238f1d45f10&gt; animation._update = new_anim_update.__get__(animation, qta.Spin) # https://stackoverflow.com/a/28127947 elif DO_MONKEYPATCH == 2: def update_decorator(method): def decorate_update(self=None): method() # TypeError: Spin._update() takes 1 positional argument but 2 were given print(&quot;decorate_update {}&quot;.format(self)) # self is None return decorate_update animation._update = update_decorator(animation._update) # https://stackoverflow.com/a/8726680 #print(animation._update) spin_icon = qta.icon('fa5s.spinner', color='red', animation=animation) msgBox.setIconPixmap(spin_icon.pixmap(orig_ipmsize)) #msgBox.setIcon(spin_icon) #animation.start() returnValue = msgBox.exec() if __name__ == '__main__': app = QApplication(sys.argv) ex = Example() sys.exit(app.exec_()) </code></pre> <p>So, it seems that <code>DO_MONKEYPATCH == 1</code> method is the answer to this question as originally posed - but I am still worried, how does my call to <code>old_anim_update()</code>, without any original references to <code>self</code>, call the old method correctly? Or is there a more correct method to do this kind of monkeypatch?</p>
<python><python-3.x><pyqt5><monkeypatching>
2024-04-14 21:57:33
1
5,938
sdbbs
78,325,210
777,304
maping of memory into structure using ctypes in python
<p>I am trying to map memory into structure but unable to get desired results. My memory is list of 32 bit values i.e.</p> <pre><code>[0x7E008000, 0x1234AAAA, 0xBBBBFFFF] </code></pre> <p>and I am trying to get this output.</p> <pre><code>ns: 0x7d008000 us: 0x1234 zs: 0xaaaabbbb crc: 0xffff </code></pre> <p>with following code I am not getting correct &quot;zs&quot; value, and if I enable &quot;crc&quot; field I get error. This is the output of below mentioned code</p> <pre><code>&gt; ns: 0x7d008000 &gt; us: 0x1234 &gt; zs: 0xbbbbffff </code></pre> <p>Here is the code</p> <pre><code> import struct import ctypes class MemoryParser: @classmethod def parse_memory(cls, memory): memory_bytes = bytearray() for word in memory: # Convert the word to little-endian bytes word_bytes = struct.pack(&quot;&gt;I&quot;, word) # Extend memory_bytes with the word bytes in little-endian order memory_bytes.extend(word_bytes) # Create a ctypes pointer from the memory_bytes ubuffer = (ctypes.c_ubyte * len(memory_bytes)).from_buffer(memory_bytes) # Return the class instance created from the ctypes pointer return cls.from_buffer(ubuffer) class Data(MemoryParser, ctypes.BigEndianStructure): _fields_ = [(&quot;ns&quot;, ctypes.c_uint32), (&quot;us&quot;, ctypes.c_uint16), (&quot;zs&quot;, ctypes.c_uint32)] #(&quot;crc&quot;, ctypes.c_uint16)] memory = [0x7E008000, 0x1234AAAA, 0xBBBBFFFF] data = Data.parse_memory(memory) print(&quot;ns:&quot;, hex(data.ns)) print(&quot;us:&quot;, hex(data.us)) print(&quot;zs:&quot;, hex(data.zs)) #print(&quot;crc:&quot;, hex(data.crc)) </code></pre> <p>Update 1:</p> <p>So I collected another data set and created fields but it fails again.</p> <p>data set, example</p> <pre><code>[0x7E008000, 0x1234AAAA, 0xBBBBFFFC, 0xCCCDDEEE] </code></pre> <p>what will be the new class fields structure for expected data like below</p> <pre><code>v1 = 0x7E008000 v2 = 0x1234 v3 = 0xAAAABBBB v4 = 0xFFF v5 = 0xCCCC v6 = 0xDD v7 = 0xEEEE </code></pre>
<python><ctypes>
2024-04-14 20:45:17
2
455
user777304
78,324,897
1,493,192
Python ConnectionRefusedError using a Docker compose with jupyter notebook and spring boot
<p>I wrote a local application using spring boot in which I have three containers: mongodb, mongo-express and jupyter noteboot. Using a python script I can access the data without errors</p> <pre><code>import requests import json url = &quot;http://localhost:8080/api/v1/metadata/sites&quot; response = requests.get(url) print(response) # print json content print(response.json()) &gt;&gt;&gt; &lt;Response [200]&gt; </code></pre> <p>However, when I repeat the same code in the notebook (of the container) I get this error</p> <p><a href="https://i.sstatic.net/BAZTq.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/BAZTq.png" alt="enter image description here" /></a></p> <pre><code>FROM python:3.9 LABEL authors=&quot;&quot; # Install any necessary packages RUN apt-get update &amp;&amp; \ apt-get install -y --no-install-recommends \ build-essential \ git &amp;&amp; \ apt-get clean &amp;&amp; \ rm -rf /var/lib/apt/lists/* # Set the working directory WORKDIR /app # Copy the requirements file and install the dependencies RUN pip install --upgrade pip #RUN pip install --no-cache-dir jupyterlab numpy==1.26.4 &amp;&amp; \ # pandas==2.2.2 &amp;&amp; \ # requests==2.31.0 RUN pip install --no-cache-dir numpy==1.26.4 \ jupyter \ pandas==2.2.2 \ requests==2.31.0 # Copy the notebooks to the container COPY . /app EXPOSE 8888 # Set the default command to run Jupyter Notebook CMD [&quot;jupyter&quot;, &quot;notebook&quot;, &quot;--ip=0.0.0.0&quot;, &quot;--port=8888&quot;, &quot;--no-browser&quot;, &quot;--allow-root&quot;, &quot;--NotebookApp.token=''&quot;] version: '3.9' services: # Database - Mongo DB mongodb: image: mongo:latest container_name: mongodb restart: always ports: - &quot;27017:27017&quot; env_file: .env environment: - MONGO_INITDB_ROOT_USERNAME=${MONGO_INITDB_ROOT_USERNAME} - MONGO_INITDB_ROOT_PASSWORD=${MONGO_INITDB_ROOT_PASSWORD} - MONGO_INITDB_DATABASE=${MONGO_INITDB_DATABASE} volumes: # - data:/data - mongo-volume:/data/db2 - ./mongo-init.js:/docker-entrypoint-initdb.d/mongo-init.js:ro # - mongo-data:/data/db # Database Manager mongo-express: image: mongo-express:latest container_name: mongo-express restart: always ports: - &quot;8081:8081&quot; depends_on: - mongodb env_file: .env environment: - ME_CONFIG_MONGODB_ADMINUSERNAME=${ME_CONFIG_MONGODB_ADMINUSERNAME} - ME_CONFIG_MONGODB_ADMINPASSWORD=${ME_CONFIG_MONGODB_ADMINPASSWORD} - ME_CONFIG_MONGODB_SERVER=mongodb # same name of &quot;mongodb&quot; in services jupyter: build: context: . dockerfile: ./jupyter/Dockerfile container_name: jupyter-metadata-itineris ports: - '8888:8888' volumes: - ./notebooks:/app # Define named volumes volumes: mongo-volume: {} networks: default: name: mongodb_network </code></pre>
<python><spring-boot><docker-compose><jupyter-notebook>
2024-04-14 18:44:19
0
8,048
Gianni Spear
78,324,819
7,199,629
Optimizing Lat/Lon Extraction from Astropy's GeocentricTrueEcliptic SkyCoord
<p>I am facing a challenge that should be straightforward but has proven to be quite complex with the Astropy library. Specifically, I need to frequently compute the longitude and latitude from a <code>SkyCoord</code> object that has been transformed to the <code>GeocentricTrueEcliptic</code> frame. Here is the relevant section of my code:</p> <pre class="lang-py prettyprint-override"><code>from astropy.coordinates import SkyCoord, GeocentricTrueEcliptic, ICRS import astropy.units as u from astropy.time import Time coo_icrs = SkyCoord(150*u.deg, 19*u.deg, frame=ICRS(), obstime=Time(2460791., format=&quot;jd&quot;)) coo = coo_icrs.geocentrictrueecliptic some_function(coo.lon, coo.lat) </code></pre> <p>Accessing <code>coo.lon</code> and <code>coo.lat</code> proves to be time-consuming due to unnecessary computations that don’t align with my use case. The issue is exacerbated as this computation needs to be repeated millions of times during each testing phase of my module, significantly impacting performance.</p> <p>Upon investigation, I found that <code>coo.lon</code> and <code>coo.lat</code> may be slow because <code>SkyCoord</code> performs a transformation to spherical coordinates. I discovered a much faster approach by directly accessing internal attributes (<code>coo_icrs._sky_coord_frame._data</code>):</p> <pre class="lang-py prettyprint-override"><code>%timeit coo_icrs._sky_coord_frame._data._lon %timeit coo_icrs.ra # 46.1 ns ± 0.203 ns per loop (mean ± std. dev. of 7 runs, 10,000,000 loops each) # 6.21 µs ± 69.8 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each) </code></pre> <p>However, this shortcut doesn’t apply to <code>coo</code> since its coordinates are only available in XYZ Cartesian format:</p> <pre class="lang-py prettyprint-override"><code>coo._sky_coord_frame._data # &lt;CartesianRepresentation (x, y, z) [dimensionless] # (-0.83586289, 0.53450713, 0.12504145)&gt; </code></pre> <p>Extracting values directly (<code>._x.value</code>, <code>._y.value</code>, and <code>._z.value</code>) and converting them manually to longitude and latitude is still faster than using <code>coo.lon</code> and <code>coo.lat</code>, but this manual conversion is also computationally intensive.</p> <p>I am searching for a method to more efficiently retrieve longitude and latitude values from a <code>SkyCoord</code> object in the <code>GeocentricTrueEcliptic</code> frame, avoiding the extra computational overhead. Is there a faster, more direct way to obtain these values?</p>
<python><numpy><numba><jit><astropy>
2024-04-14 18:16:23
1
361
ysBach
78,324,721
1,182,299
BERT MLM model fine-tune on small data bad results
<p>I want to fine tune a BERT MLM model from Hugging Face. I have only a small amount of data (train.csv) like this:</p> <pre><code>text בראשית ברא אלהים את השמים ואת הארץ והארץ היתה תהו ובהו וחשך על פני תהום ורוח אלהים מרחפת על פני המים ויאמר אלהים יהי אור ויהי אור וירא אלהים את האור כי טוב ויבדל אלהים בין האור ובין החשך ויקרא אלהים לאור יום ולחשך קרא לילה ויהי ערב ויהי בקר יום אחד ויאמר אלהים יהי רקיע בתוך המים ויהי מבדיל בין מים למים ויעש אלהים את הרקיע ויבדל בין המים אשר מתחת לרקיע ובין המים אשר מעל לרקיע ויהי כן ויקרא אלהים לרקיע שמים ויהי ערב ויהי בקר יום שני ויאמר אלהים יקוו המים מתחת השמים אל מקום אחד ותראה היבשה ויהי כן ויקרא אלהים ליבשה ארץ ולמקוה המים קרא ימים וירא אלהים כי טוב </code></pre> <p>Below is my script in doing the fine-tuning:</p> <pre><code>from huggingface_hub import login from transformers import AutoModelForMaskedLM from transformers import AutoTokenizer import torch from datasets import load_dataset from transformers import DataCollatorForLanguageModeling import collections import numpy as np from transformers import default_data_collator from transformers import TrainingArguments from transformers import Trainer import math from transformers import EarlyStoppingCallback, IntervalStrategy access_token_write = &quot;hf_XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX&quot; login(token = access_token_write) model_checkpoint = &quot;dicta-il/BEREL_2.0&quot; model = AutoModelForMaskedLM.from_pretrained(model_checkpoint) tokenizer = AutoTokenizer.from_pretrained(model_checkpoint) sam_dataset = load_dataset(&quot;johnlockejrr/samv2&quot;) def tokenize_function(examples): result = tokenizer(examples[&quot;text&quot;]) if tokenizer.is_fast: result[&quot;word_ids&quot;] = [result.word_ids(i) for i in range(len(result[&quot;input_ids&quot;]))] return result tokenized_datasets = sam_dataset.map( tokenize_function, batched=True, remove_columns=[&quot;text&quot;] ) chunk_size = 128 tokenized_samples = tokenized_datasets[&quot;train&quot;][:3] concatenated_examples = { k: sum(tokenized_samples[k], []) for k in tokenized_samples.keys() } def group_texts(examples): # Concatenate all texts concatenated_examples = {k: sum(examples[k], []) for k in examples.keys()} # Compute length of concatenated texts total_length = len(concatenated_examples[list(examples.keys())[0]]) # We drop the last chunk if it's smaller than chunk_size total_length = (total_length // chunk_size) * chunk_size # Split by chunks of max_len result = { k: [t[i : i + chunk_size] for i in range(0, total_length, chunk_size)] for k, t in concatenated_examples.items() } # Create a new labels column result[&quot;text&quot;] = result[&quot;input_ids&quot;].copy() return result lm_datasets = tokenized_datasets.map(group_texts, batched=True) data_collator = DataCollatorForLanguageModeling(tokenizer=tokenizer, mlm_probability=0.15) samples = [lm_datasets[&quot;train&quot;][i] for i in range(2)] train_size = 1_000 test_size = int(0.1 * train_size) downsampled_dataset = lm_datasets[&quot;train&quot;].train_test_split( train_size=train_size, test_size=test_size, seed=42 ) batch_size = 64 logging_steps = len(downsampled_dataset[&quot;train&quot;]) // batch_size model_name = model_checkpoint.split(&quot;/&quot;)[-1] training_args = TrainingArguments( output_dir=f&quot;{model_name}-finetuned-sam-v1&quot;, overwrite_output_dir=True, evaluation_strategy=IntervalStrategy.STEPS, # &quot;steps&quot; eval_steps = 50, # Evaluation and Save happens every 50 steps save_total_limit = 5, # Only last 5 models are saved. Older ones are deleted. learning_rate=2e-5, weight_decay=0.01, per_device_train_batch_size=batch_size, per_device_eval_batch_size=batch_size, num_train_epochs=5, push_to_hub=True, fp16=True, logging_steps=logging_steps, metric_for_best_model = 'f1', load_best_model_at_end=True ) trainer = Trainer( model=model, args=training_args, train_dataset=downsampled_dataset[&quot;train&quot;], eval_dataset=downsampled_dataset[&quot;test&quot;], data_collator=data_collator, tokenizer=tokenizer, ) trainer.train() eval_results = trainer.evaluate() print(f&quot;&gt;&gt;&gt; Perplexity: {math.exp(eval_results['eval_loss']):.2f}&quot;) trainer.push_to_hub() </code></pre> <p>All went well, model trained, pushed to Hugging Face. Problems:</p> <ol> <li>Loss: 6.2156</li> <li>Model work for <code>Fill-Mask</code> but with my new data seems something happened in the process and some words are masked as tokens and not full words:</li> </ol> <p>Here is a script as an example of <code>Fill-Mask</code> and the output:</p> <pre><code>from transformers import pipeline mask_filler = pipeline( &quot;fill-mask&quot;, model=&quot;johnlockejrr/BEREL_2.0-finetuned-sam-v1&quot; ) text = &quot;ואלין שמהת בני [MASK] דעלו למצרים עם יעקב גבר וביתה עלו&quot; preds = mask_filler(text) for pred in preds: print(f&quot;&gt;&gt;&gt; {pred['sequence']}&quot;) </code></pre> <p>Output:</p> <pre><code>ואלין שמה ##ת בני יעקב דעלו למצרים עם יעקב גבר ובית ##ה עלו ואלין שמה ##ת בני ישראל דעלו למצרים עם יעקב גבר ובית ##ה עלו ואלין שמה ##ת בני לאה דעלו למצרים עם יעקב גבר ובית ##ה עלו ואלין שמה ##ת בני ראובן דעלו למצרים עם יעקב גבר ובית ##ה עלו ואלין שמה ##ת בני דן דעלו למצרים עם יעקב גבר ובית ##ה עלו </code></pre> <p>I was expecting this output:</p> <pre><code>ואלין שמהת בני יעקב דעלו למצרים עם יעקב גבר וביתה עלו ואלין שמהת בני ישראל דעלו למצרים עם יעקב גבר וביתה עלו ואלין שמהת בני לאה דעלו למצרים עם יעקב גבר וביתה עלו ואלין שמהת בני ראובן דעלו למצרים עם יעקב גבר וביתה עלו ואלין שמהת בני דן דעלו למצרים עם יעקב גבר וביתה עלו </code></pre> <p>What am I doing wrong?</p>
<python><huggingface-transformers><bert-language-model>
2024-04-14 17:44:46
1
1,791
bsteo
78,324,619
1,802,483
Problem installing `apache-flink:1.19.0` using python docker 3.9 - 3.12
<p>I am having trouble installing <code>PyFlink</code>/<code>apache-flink 1.19.0</code> using python docker 3.9 - 3.12 following <a href="https://nightlies.apache.org/flink/flink-docs-master/docs/dev/python/installation/" rel="nofollow noreferrer">the official tutorial</a>.</p> <p>It seems that the error is about a path that return <code>NoneType</code>, but I am not sure how to fix it.</p> <p>I did not post <code>python:3.12</code> as the error is the same as others.</p> <p>Can you please help me on this?</p> <p>Thank you very much.</p> <h2>Python 3.9</h2> <p><code>docker run - it python:3.9 python - m pip install apache-flink</code></p> <pre><code> × Getting requirements to build wheel did not run successfully. │ exit code: 1 ╰─&gt; [22 lines of output] Traceback (most recent call last): File &quot;/usr/local/lib/python3.9/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 353, in &lt;module&gt; main() File &quot;/usr/local/lib/python3.9/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 335, in main json_out['return_val'] = hook(**hook_input['kwargs']) File &quot;/usr/local/lib/python3.9/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 118, in get_requires_for_build_wheel return hook(config_settings) File &quot;/tmp/pip-build-env-rdoi1tww/overlay/lib/python3.9/site-packages/setuptools/build_meta.py&quot;, line 162, in get_requires_for_build_wheel return self._get_build_requires( File &quot;/tmp/pip-build-env-rdoi1tww/overlay/lib/python3.9/site-packages/setuptools/build_meta.py&quot;, line 143, in _get_build_requires self.run_setup() File &quot;/tmp/pip-build-env-rdoi1tww/overlay/lib/python3.9/site-packages/setuptools/build_meta.py&quot;, line 267, in run_setup super(_BuildMetaLegacyBackend, File &quot;/tmp/pip-build-env-rdoi1tww/overlay/lib/python3.9/site-packages/setuptools/build_meta.py&quot;, line 158, in run_setup exec(compile(code, __file__, 'exec'), locals()) File &quot;setup.py&quot;, line 169, in &lt;module&gt; include_dirs=get_java_include() + ['src/main/c/pemja/core/include'], File &quot;setup.py&quot;, line 111, in get_java_include inc = os.path.join(get_java_home(), inc_name) File &quot;/usr/local/lib/python3.9/posixpath.py&quot;, line 76, in join a = os.fspath(a) TypeError: expected str, bytes or os.PathLike object, not NoneType [end of output] note: This error originates from a subprocess, and is likely not a problem with pip. error: subprocess-exited-with-error × Getting requirements to build wheel did not run successfully. │ exit code: 1 ╰─&gt; See above for output. </code></pre> <h2>Python 3.10</h2> <p><code>docker run - it python:3.10 python - m pip install apache-flink</code></p> <pre><code> × Getting requirements to build wheel did not run successfully. │ exit code: 1 ╰─&gt; [22 lines of output] Traceback (most recent call last): File &quot;/usr/local/lib/python3.10/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 353, in &lt;module&gt; main() File &quot;/usr/local/lib/python3.10/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 335, in main json_out['return_val'] = hook(**hook_input['kwargs']) File &quot;/usr/local/lib/python3.10/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 118, in get_requires_for_build_wheel return hook(config_settings) File &quot;/tmp/pip-build-env-ptcni4h1/overlay/lib/python3.10/site-packages/setuptools/build_meta.py&quot;, line 162, in get_requires_for_build_wheel return self._get_build_requires( File &quot;/tmp/pip-build-env-ptcni4h1/overlay/lib/python3.10/site-packages/setuptools/build_meta.py&quot;, line 143, in _get_build_requires self.run_setup() File &quot;/tmp/pip-build-env-ptcni4h1/overlay/lib/python3.10/site-packages/setuptools/build_meta.py&quot;, line 267, in run_setup super(_BuildMetaLegacyBackend, File &quot;/tmp/pip-build-env-ptcni4h1/overlay/lib/python3.10/site-packages/setuptools/build_meta.py&quot;, line 158, in run_setup exec(compile(code, __file__, 'exec'), locals()) File &quot;setup.py&quot;, line 169, in &lt;module&gt; include_dirs=get_java_include() + ['src/main/c/pemja/core/include'], File &quot;setup.py&quot;, line 111, in get_java_include inc = os.path.join(get_java_home(), inc_name) File &quot;/usr/local/lib/python3.10/posixpath.py&quot;, line 76, in join a = os.fspath(a) TypeError: expected str, bytes or os.PathLike object, not NoneType [end of output] note: This error originates from a subprocess, and is likely not a problem with pip. error: subprocess-exited-with-error × Getting requirements to build wheel did not run successfully. │ exit code: 1 ╰─&gt; See above for output. note: This error originates from a subprocess, and is likely not a problem with pip. </code></pre> <h2>Python 3.11</h2> <p><code>docker run - it python:3.11 python - m pip install apache-flink</code></p> <pre><code> × Getting requirements to build wheel did not run successfully. │ exit code: 1 ╰─&gt; [27 lines of output] Traceback (most recent call last): File &quot;/usr/local/lib/python3.11/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 353, in &lt;module&gt; main() File &quot;/usr/local/lib/python3.11/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 335, in main json_out['return_val'] = hook(**hook_input['kwargs']) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;/usr/local/lib/python3.11/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 118, in get_requires_for_build_wheel return hook(config_settings) ^^^^^^^^^^^^^^^^^^^^^ File &quot;/tmp/pip-build-env-jpt4cjc2/overlay/lib/python3.11/site-packages/setuptools/build_meta.py&quot;, line 162, in get_requires_for_build_wheel return self._get_build_requires( ^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;/tmp/pip-build-env-jpt4cjc2/overlay/lib/python3.11/site-packages/setuptools/build_meta.py&quot;, line 143, in _get_build_requires self.run_setup() File &quot;/tmp/pip-build-env-jpt4cjc2/overlay/lib/python3.11/site-packages/setuptools/build_meta.py&quot;, line 268, in run_setup self).run_setup(setup_script=setup_script) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;/tmp/pip-build-env-jpt4cjc2/overlay/lib/python3.11/site-packages/setuptools/build_meta.py&quot;, line 158, in run_setup exec(compile(code, __file__, 'exec'), locals()) File &quot;setup.py&quot;, line 169, in &lt;module&gt; include_dirs=get_java_include() + ['src/main/c/pemja/core/include'], ^^^^^^^^^^^^^^^^^^ File &quot;setup.py&quot;, line 111, in get_java_include inc = os.path.join(get_java_home(), inc_name) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;&lt;frozen posixpath&gt;&quot;, line 76, in join TypeError: expected str, bytes or os.PathLike object, not NoneType [end of output] note: This error originates from a subprocess, and is likely not a problem with pip. error: subprocess-exited-with-error × Getting requirements to build wheel did not run successfully. │ exit code: 1 ╰─&gt; See above for output. </code></pre> <h3>UPDATE 15/4/2024</h3> <p>It works in Windows. But it does not work in Mac when using this command: Is it a permission problem?</p> <p><code>docker run -it python:3.11 pip install apache-flink</code></p> <p>My Mac Config:</p> <ul> <li>M1 Max</li> <li>Sonoma 14.4.1</li> </ul> <p>Message below:</p> <pre><code>× Getting requirements to build wheel did not run successfully. │ exit code: 1 ╰─&gt; [27 lines of output] Traceback (most recent call last): File &quot;/usr/local/lib/python3.11/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 353, in &lt;module&gt; main() File &quot;/usr/local/lib/python3.11/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 335, in main json_out['return_val'] = hook(**hook_input['kwargs']) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;/usr/local/lib/python3.11/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py&quot;, line 118, in get_requires_for_build_wheel return hook(config_settings) ^^^^^^^^^^^^^^^^^^^^^ File &quot;/tmp/pip-build-env-r8xb1epe/overlay/lib/python3.11/site-packages/setuptools/build_meta.py&quot;, line 162, in get_requires_for_build_wheel return self._get_build_requires( ^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;/tmp/pip-build-env-r8xb1epe/overlay/lib/python3.11/site-packages/setuptools/build_meta.py&quot;, line 143, in _get_build_requires self.run_setup() File &quot;/tmp/pip-build-env-r8xb1epe/overlay/lib/python3.11/site-packages/setuptools/build_meta.py&quot;, line 268, in run_setup self).run_setup(setup_script=setup_script) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;/tmp/pip-build-env-r8xb1epe/overlay/lib/python3.11/site-packages/setuptools/build_meta.py&quot;, line 158, in run_setup exec(compile(code, __file__, 'exec'), locals()) File &quot;setup.py&quot;, line 169, in &lt;module&gt; include_dirs=get_java_include() + ['src/main/c/pemja/core/include'], ^^^^^^^^^^^^^^^^^^ File &quot;setup.py&quot;, line 111, in get_java_include inc = os.path.join(get_java_home(), inc_name) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;&lt;frozen posixpath&gt;&quot;, line 76, in join TypeError: expected str, bytes or os.PathLike object, not NoneType [end of output] note: This error originates from a subprocess, and is likely not a problem with pip. </code></pre>
<python><docker><apache-flink>
2024-04-14 17:07:58
2
705
Ellery Leung
78,324,332
1,848,244
How to instantiate a large number of NotRequired arguments in a TypedDict?
<p>Consider this contrived example:</p> <pre class="lang-py prettyprint-override"><code>from typing import Mapping, Union, MutableMapping from typing_extensions import TypedDict, NotRequired class Pet(TypedDict): softness: NotRequired[int] name: NotRequired[str] # **IMPORTANT**: Assume these are only known at run time. softness_exists = False name_exists = True optargs: MutableMapping[str, Union[int, str]] = dict() if softness_exists: optargs['softness'] = 999999 if name_exists: optargs['name'] = 'David' p = Pet( type='Dog', #Unsupported type &quot;MutableMapping[str, Union[Food, int, str]]&quot; for ** expansion in TypedDict **optargs ) print(p) </code></pre> <p>In my real world use case, I have a relatively large number of optional arguments. Enough that conditionally populating optargs based on the run-time input is the only wieldy way to accomplish the construction of the TypedDict.</p> <p>But this does not appear to be allowed. What is the recommended way to construct a TypedDict with a large number of <code>NotRequired</code> fields, whose applicability is decided at run time?</p> <p>I am suspecting that the decision on which NotRequired fields present in an instantiation of a TypedDict cannot be decided at run time.</p>
<python><mypy><python-typing>
2024-04-14 15:41:39
1
437
user1848244
78,324,285
10,200,497
How can I find the first row that meets conditions of a mask for each group?
<p>This is my DataFrame:</p> <pre><code>import pandas as pd df = pd.DataFrame( { 'a': ['x', 'x', 'x', 'x', 'x', 'y', 'y', 'y', 'y', 'y', 'y', 'y'], 'b': [1, 1, 1, 2, 2, 1, 1, 1, 2, 2, 2, 2], 'c': [9, 8, 11, 13, 14, 3, 104, 106, 11, 100, 70, 7] } ) </code></pre> <p>Expected output: Creating column <code>out</code>:</p> <pre><code> a b c out 0 x 1 9 NaN 1 x 1 8 NaN 2 x 1 11 NaN 3 x 2 13 found 4 x 2 14 NaN 5 y 1 3 NaN 6 y 1 104 found 7 y 1 106 NaN 8 y 2 11 NaN 9 y 2 100 NaN 10 y 2 70 NaN 11 y 2 7 NaN </code></pre> <p>The mask is:</p> <pre><code>mask = (df.c &gt; 10) </code></pre> <p>The process: Grouping is by column <code>a</code>:</p> <p><strong>a)</strong> For each group, finding the first row that meets the conditions of the <code>mask</code>.</p> <p><strong>b)</strong> For group <code>x</code> this condition only applies when <code>b == 2</code>. That is why row <code>3</code> is selected.</p> <p>And this is my attempt. It is getting close but it feels like this is not the way:</p> <pre><code>def func(g): mask = (g.c &gt; 10) g.loc[mask.cumsum().eq(1) &amp; mask, 'out'] = 'found' return g df = df.groupby('a').apply(func) </code></pre>
<python><pandas><dataframe><group-by>
2024-04-14 15:24:24
3
2,679
AmirX
78,323,995
7,633,739
`BaseSettings` has been moved to the `pydantic-settings` package
<p>I am struggling with this error from long now, tried other solutions but no luck, here is my overview about the issue.</p> <p>Firstly i am not at all using this package anywhere in my project, seems like python is executing this internally, but still i am facing this issue, here are my python configurations</p> <pre><code>Python 3.9.6 pip 21.2.4 </code></pre> <p>I tried other version of python too which is <code>3.10.9</code></p> <p>Below is the error i am facing.</p> <pre><code> File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/langchain/vectorstores/chroma.py&quot;, line 81, in __init__ import chromadb File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/chromadb/__init__.py&quot;, line 1, in &lt;module&gt; import chromadb.config File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/chromadb/config.py&quot;, line 1, in &lt;module&gt; from pydantic import BaseSettings, Field File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/pydantic/__init__.py&quot;, line 380, in __getattr__ return _getattr_migration(attr_name) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/pydantic/_migration.py&quot;, line 296, in wrapper raise PydanticImportError( pydantic.errors.PydanticImportError: `BaseSettings` has been moved to the `pydantic-settings` package. See https://docs.pydantic.dev/2.7/migration/#basesettings-has-moved-to-pydantic-settings for more details. For further information visit https://errors.pydantic.dev/2.7/u/import-error During handling of the above exception, another exception occurred: Traceback (most recent call last): File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask/app.py&quot;, line 2213, in __call__ return self.wsgi_app(environ, start_response) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask/app.py&quot;, line 2193, in wsgi_app response = self.handle_exception(e) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask_cors/extension.py&quot;, line 176, in wrapped_function return cors_after_request(app.make_response(f(*args, **kwargs))) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask_cors/extension.py&quot;, line 176, in wrapped_function return cors_after_request(app.make_response(f(*args, **kwargs))) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask/app.py&quot;, line 2190, in wsgi_app response = self.full_dispatch_request() File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask/app.py&quot;, line 1486, in full_dispatch_request rv = self.handle_user_exception(e) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask_cors/extension.py&quot;, line 176, in wrapped_function return cors_after_request(app.make_response(f(*args, **kwargs))) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask_cors/extension.py&quot;, line 176, in wrapped_function return cors_after_request(app.make_response(f(*args, **kwargs))) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask/app.py&quot;, line 1484, in full_dispatch_request rv = self.dispatch_request() File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/flask/app.py&quot;, line 1469, in dispatch_request return self.ensure_sync(self.view_functions[rule.endpoint])(**view_args) File &quot;/Users/pradeepyenkuwale/Work/Repositories/TripAdvisor/setup.py&quot;, line 53, in userQuery resp = ProcessConsumerQuery(query) File &quot;/Users/pradeepyenkuwale/Work/Repositories/TripAdvisor/lib/process_query.py&quot;, line 30, in ProcessConsumerQuery index = VectorstoreIndexCreator().from_loaders([loader]) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/langchain/indexes/vectorstore.py&quot;, line 83, in from_loaders return self.from_documents(docs) File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/langchain/indexes/vectorstore.py&quot;, line 88, in from_documents vectorstore = self.vectorstore_cls.from_documents( File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/langchain/vectorstores/chroma.py&quot;, line 771, in from_documents return cls.from_texts( File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/langchain/vectorstores/chroma.py&quot;, line 707, in from_texts chroma_collection = cls( File &quot;/Users/pradeepyenkuwale/Library/Python/3.9/lib/python/site-packages/langchain/vectorstores/chroma.py&quot;, line 84, in __init__ raise ImportError( ImportError: Could not import chromadb python package. Please install it with pip install chromadb.` </code></pre> <p>FYI, I am using <code>LangChain</code> AI framework with <code>ChatGPT</code>. The same code works fine in windows machine, i am trying to run the repository in MacBook Pro, I tried same version as in windows as well but still i am facing this issue.</p>
<python><langchain>
2024-04-14 13:48:48
0
320
Pradeep Yenkuwale
78,323,943
5,257,430
Statistic values of Fleiss Kappa using statsmodels.stats.inter_rater
<p>I use statsmodels.stats.inter_rater.fleiss_kappa to calculate my inter-rater reliability. I only get the kappa value. What if I need the z-value, p-value, and range?</p>
<python><pandas><statsmodels>
2024-04-14 13:31:15
1
621
pill45
78,323,859
9,111,293
Broadcast pytorch array across channels based on another array
<p>I have two arrays, <code>x</code> and <code>y</code>, with the same shape (<code>B 1 N</code>).</p> <p><code>x</code> represents data and <code>y</code> represents which class (from <code>1</code> to <code>C</code>) each datapoint in <code>x</code> belongs to.</p> <p>I want to create a new tensor <code>z</code> (with shape <code>B C</code>) where</p> <ol> <li>the data in <code>x</code> are partitioned into channels based on their classes in <code>y</code></li> <li>and summed over <code>N</code></li> </ol> <p>I can accomplish this if I use a one-hot encoding. However, for large tensors (especially with a large number of classes), PyTorch's one-hot encoding quickly uses up all memory on the GPU.</p> <p>Is there a more memory-efficient way to do this broadcasting without explicitly allocating a <code>B C N</code> tensor?</p> <p>Here's an MWE of what I'm after:</p> <pre class="lang-py prettyprint-override"><code>import torch B, C, N = 2, 10, 1000 x = torch.randn(B, 1, N) y = torch.randint(low=0, high=C, size=(B, 1, N)) one_hot = torch.nn.functional.one_hot(y, C) # B 1 N C one_hot = one_hot.squeeze().permute(0, -1, 1) # B C N z = x * one_hot # B C N z = z.sum(-1) # B C </code></pre>
<python><pytorch><tensor>
2024-04-14 12:59:45
1
579
Vivek
78,323,749
6,915,206
AWS EC2 Server not serving some pages and static files propperly
<p>I just Deployed a <a href="http://3.17.142.65/" rel="nofollow noreferrer">website</a> on AWS EC2 from github clone. When i visit to <a href="http://3.17.142.65/influencers/" rel="nofollow noreferrer">Influencer Marketing</a> &amp; <a href="http://3.17.142.65/career/" rel="nofollow noreferrer">Career</a> pages the server serving the static files from <strong>S3 Bucket</strong> correctly. But when i am visiting my <a href="http://3.17.142.65/" rel="nofollow noreferrer">home page</a> &amp; <a href="http://3.17.142.65/COE/" rel="nofollow noreferrer">Who are We</a> pages its not serving the static files also i am not getting these pages contents (raw data). I assume that its not serving static files its ok but <strong>where is the content of both pages gone</strong>. What i am missing here. I am new to AWS and Website deployment So if i did any mistake please let me now i will correct it. If you require any additional information i will add it.</p> <p>Here you can access both home &amp; Who we are pages -<a href="https://github.com/rahul6612/testing-repo/tree/main" rel="nofollow noreferrer">link</a></p> <p><strong>Configure Nginx to Proxy Pass to Gunicorn</strong></p> <p><div class="snippet" data-lang="js" data-hide="false" data-console="true" data-babel="false"> <div class="snippet-code"> <pre class="snippet-code-html lang-html prettyprint-override"><code>[Unit] Description=gunicorn daemon Requires=gunicorn.socket After=network.target [Service] User=ubuntu Group=www-data WorkingDirectory=/home/ubuntu/try-django-digital-marketing/try-django-digital-marketing ExecStart=/home/ubuntu/try-django-digital-marketing/try-django-digital-marketing/env/bin/gunicorn \ --access-logfile - \ --workers 3 \ --bind unix:/run/gunicorn.sock \ BE.wsgi:application server { listen 80; server_name 3.17.142.65; location = /favicon.ico { access_log off; log_not_found off; } location / { include proxy_params; proxy_pass http://unix:/run/gunicorn.sock; } }</code></pre> </div> </div> </p> <p><strong>Bucket Policy, Cross-origin resource sharing (CORS) &amp; User Policy</strong></p> <p><div class="snippet" data-lang="js" data-hide="false" data-console="true" data-babel="false"> <div class="snippet-code"> <pre class="snippet-code-html lang-html prettyprint-override"><code>{ "Version": "2008-10-17", "Statement": [ { "Sid": "PublicReadGetObject", "Effect": "Allow", "Principal": { "AWS": "*" }, "Action": [ "s3:GetObject", "s3:PutObject" ], "Resource": "arn:aws:s3:::try-marketing/*" } ] } __________________________________________________ [ { "AllowedHeaders": [ "*" ], "AllowedMethods": [ "POST", "GET", "PUT" ], "AllowedOrigins": [ "*" ], "ExposeHeaders": [] } ] ___________________________________________________ { "Version": "2012-10-17", "Statement": [ { "Action": [ "s3:PutObject", "s3:PutObjectAcl", "s3:GetObject", "s3:GetObjectAcl", "s3:AbortMultipartUpload" ], "Resource": [ "arn:aws:s3:::try-marketing", "arn:aws:s3:::try-marketing/*" ], "Effect": "Allow" } ] }</code></pre> </div> </div> </p> <p>AWS Conf file</p> <pre><code>AWS_USERNAME = 'user11111' AWS_ACCESS_KEY_ID = 'xxxxxxxxxx' AWS_SECRET_ACCESS_KEY = 'xxxxxxxxxxxx' AWS_PRELOAD_METADATA = True AWS_QUERYSTRING_AUTH = False AWS_S3_SIGNATURE_VERSION = &quot;s3v4&quot; AWS_S3_REGION_NAME = 'us-east-2' DEFAULT_FILE_STORAGE = 'BE.aws.utils.MediaRootS3BotoStorage' STATICFILES_STORAGE = 'BE.aws.utils.StaticRootS3BotoStorage' AWS_STORAGE_BUCKET_NAME = 'try-marketing' S3DIRECT_REGION = 'us-east-2' S3_URL = '//%s.s3.amazonaws.com/' % AWS_STORAGE_BUCKET_NAME MEDIA_URL = '//%s.s3.amazonaws.com/media/' % AWS_STORAGE_BUCKET_NAME MEDIA_ROOT = MEDIA_URL STATIC_URL = S3_URL + 'static/' ADMIN_MEDIA_PREFIX = STATIC_URL + 'admin/' AWS_DEFAULT_ACL = None </code></pre>
<python><django><amazon-web-services><amazon-s3><django-staticfiles>
2024-04-14 12:13:05
2
563
Rahul Verma
78,323,737
1,440,299
Python+Selenium. Page wan't load in headless mode
<p>I'm trying load page with that code</p> <pre><code> from selenium import webdriver from selenium.webdriver.chrome.options import Options from selenium.webdriver.support.ui import WebDriverWait from selenium.webdriver.support import expected_conditions as EC from selenium.webdriver.common.by import By from selenium.common.exceptions import TimeoutException chrome_options = Options() chrome_options.add_argument(&quot;--headless=new&quot;) driver = webdriver.Chrome(options=chrome_options) driver.get('https://www.chabad.org/calendar/zmanim_cdo/aid/143790/locationid/790/locationtype/1/save/1/tdate/4-15-2024/jewish/Zmanim-Halachic-Times.htm') try: myElem = WebDriverWait(driver, 20).until(EC.presence_of_element_located((By.ID, 'group_4/15/2024'))) print(&quot;Page is ready!&quot;) except TimeoutException: print(&quot;Loading took too much time!&quot;) </code></pre> <p>and I get a <code>TimeoutException</code>. But if I run browser in standard mode (not headless) it works fine.</p> <p>Because I need get information from page on the server, I want run browser in headless mode. So how can I load information from that page?</p>
<python><selenium-webdriver><selenium-chromedriver><google-chrome-headless>
2024-04-14 12:09:41
1
355
Ishayahu
78,323,703
2,913,106
What sparse object should be used for indexing via coordinates and efficient summing across dimensions?
<p>I'd like to find some container object to contain scalar values (doesn't really matter whether integral or fractional or floating point types). The following snippet roughly outlines the usecases and the criteria it needs to fulfill. In this snipped I'm using a numpy array, but I'd like to avoid that as it needs a lot of memory, because it will only be sparsely filled. But other than that, it fulfills my requirements.</p> <ul> <li>The object should hold scalar values, that are indexed by a <code>d</code>-dimensional index with values of <code>0,1,2,...,n-1</code>. (Imagine something like <code>5 &lt;= d &lt;= 100</code>, <code>20 &lt;= n &lt;= 200</code>).</li> <li><s>The object should be mutable, i.e. the values need to be updatable.</s> (Edit: Not actually necessary, see discussion below.)</li> <li>All possible indices should be initialized with zero at the start. That is, all indices that have not been accessed should implicitly assumed to hold zero.</li> <li>It should be efficient to sum across one or multiple dimensions.</li> </ul> <p>Is there built in python object that satisfies this, or is there a data structure that can effciently implemented in python?</p> <p>So far I have found the scipy <a href="https://docs.scipy.org/doc/scipy/reference/generated/scipy.sparse.coo_array.html#scipy.sparse.coo_array" rel="nofollow noreferrer">COO arrays</a>, that satisfy some of the criteria, but they only support 1- and 2-d indexing.</p> <p>For context: The idea is building a frequency table of certain objects, and use this then as a distribution to sample from, marginalize, etc.</p> <pre><code>import numpy as np # just some placeholder data data = [((1,0,5),6),((2,6,5),100),((5,3,1),1),((2,0,5),4),((2,6,5),100)] # structure mapping [0,n]^d coordinates to scalars data_structure = np.zeros((10, 10, 10)) # needs to be mutable #EDIT: doesn't need to be, see discussion below for coords, value in data: data_structure[coords] += value #EDIT: can be done as preprocessing step # needs to be able to efficiently sum across dimensions for k in range(100): x = data_structure.sum(axis=(1,2)) y = data_structure.sum(axis=(0,)) z = data_structure.sum(axis=(0,2)) </code></pre>
<python><arrays><numpy><data-structures><probability-distribution>
2024-04-14 11:55:38
1
11,728
flawr
78,323,611
1,725,836
OpenCV play video without waitKey
<p>I'm trying to play a video with OpenCV in Python. For that I use this code:</p> <pre class="lang-py prettyprint-override"><code>i = 0 while True: frame = frames[i] cv2.imshow(&quot;video&quot;, frame) cv2.waitKey(int(1000 / frame_rate)) i += 1 </code></pre> <p>But once the user uses his keyboard, the playback speed changes and the video plays in a &quot;fast-forward&quot; mode. I don't need key detection, so I tried to replace the <code>waitKey</code> with <code>time.sleep</code> but the video is not playing at all.</p> <p>How can I prevent the keyboard from changing the playback speed?</p>
<python><opencv><user-interface>
2024-04-14 11:24:06
1
9,887
nrofis
78,323,579
6,709,460
FastAPI TrustedHostMiddleware refuses my host
<p>When I was developing the application on my local machine and I tried to reach <code>/docs</code> I had to set up up <code>127.0.0.1</code> as my trusted host. This worked fine. But now I have put the app on the server and when I try to reach <code>/docs</code> with the IP of my computer, I am denied access.</p> <p>I have checked my IP on several websites, I have also checked <code>request.client.host</code> to confirm the IP. But I can not access the <code>/docs</code>.</p> <p>The only way I can access the route is, if I set <code>trusted_hosts = [&quot;*&quot;]</code>. But this doesn't make sense.</p> <p>Here is my code:</p> <pre><code>from fastapi import FastAPI from fastapi.middleware.trustedhost import TrustedHostMiddleware app = FastAPI() # Define a list of trusted hosts trusted_hosts = [ &quot;my_ip_address&quot; ] # Add TrustedHostMiddleware to the app with the list of trusted hosts app.add_middleware(TrustedHostMiddleware, allowed_hosts=trusted_hosts) </code></pre>
<python><fastapi><middleware><starlette>
2024-04-14 11:14:46
1
741
Testing man
78,323,390
10,480,181
How to install wheel file dependency with extras on Databricks?
<p>One of my workflows has a python wheel file dependency. However that wheel file also has extra dependencies that need to be installed. How can I do that on databricks?</p> <p>I tried using a init script but it doesn't work. I want to do something analogous to:</p> <pre><code>pip install &quot;path/to/wheel/file/mywheel.whl[extras]&quot; </code></pre> <p>How to achieve this in a Databricks workflow? The entry point to the task is a python function defined in the wheel file.</p>
<python><azure-databricks><python-packaging><python-wheel>
2024-04-14 10:04:10
0
883
Vandit Goel
78,323,238
7,611,838
Timeout before the position for partition could be determined in one topic in apache beam ReadFromKafka
<p>I'm having a Google dataflow job, I'm using apache beam <code>ReadFromKafka</code> to consume topic messages. I'm consuming 4 topics. The pipeline used to work fine, after we added a new broker to our kafka cluster and triggered a rebalancing, the consumer started failing on one specific topic but successfully kept working for the other 3 topics.</p> <pre><code>The error is: org.apache.kafka.common.errors.TimeoutException: Timeout of 60000ms expired before the position for partition topic_name-0 could be determined </code></pre> <p>Here is my code:</p> <pre><code>( pcoll | f&quot;Read From Kafka {self.transf_label}&quot; &gt;&gt; ReadFromKafka( consumer_config=self.kafka_consumer_config, topics=_topic_list, with_metadata=True, ) | f&quot;Extract KVT {self.transf_label}&quot; &gt;&gt; Map(extract_key_value_topic) | f&quot;Decode messages {self.transf_label}&quot; &gt;&gt; ParDo(DecodeKafkaMessage()) ) </code></pre> <p>In the logs I can see</p> <pre><code>[Consumer clientId=consumer-group-dummy-4, groupId=consumer-group-dummy] Subscribed to partition(s): topic_name-0 </code></pre> <p>But after few secs it fails with the timeout error, while continuously pulling messages of other topics</p> <p>Here is the config:</p> <pre><code>{ &quot;bootstrap.servers&quot;: BROKERS, &quot;security.protocol&quot;: SECURITY_PROTOCOL, &quot;sasl.mechanism&quot;: SASL_MECHANISM, &quot;group.id&quot;: &quot;consumer-group-dummy&quot;, &quot;session.timeout.ms&quot;: &quot;60000&quot;, &quot;sasl.jaas.config&quot;: f'org.apache.kafka.common.security.scram.ScramLoginModule required username=&quot;{SASL_USERNAME}&quot; password=&quot;{SASL_PASSWORD}&quot;;', &quot;auto.offset.reset&quot;: &quot;latest&quot;, } </code></pre> <p>One thing to consider is our other Kafka consumer apps are not throwing any issue, this behavior is noticed only with the apache beam pipeline. Also when I tried to trigger locally a consumer with the same config and same consumer group using Kafka confluent library and simple python script, it seems to be working fine and pulling the messages.</p> <p>I found this <a href="https://github.com/apache/flink-connector-kafka/pull/91" rel="nofollow noreferrer">bug</a> in Apache flink, can it be the same in the beam Kafka connector in Java?</p>
<python><java><apache-kafka><google-cloud-dataflow><apache-beam>
2024-04-14 08:56:13
1
974
Idhem
78,323,213
13,443,114
Environment variable set as password is appearing as empty string in Python Environ library
<p>I just noticed that when I set an environment variable as <code>PASSWORD</code> for my script in Python, I am using the <a href="https://pypi.org/project/python-environ/" rel="nofollow noreferrer">python-environ</a> library to read and access environment variables, the value for the environment variable set to password appears as an empty string, setting the environment variable to another other thing but the word PASSWORD works, I do not understand why this is the case, and I can not seem to find a documented explanation or justification for this.</p> <p><strong>.env</strong></p> <pre><code>USERNAME=testuser PASSWORD=sdfnsidufbsdufbsiudf PASSWORDU=sdfnsidufbsdufbsiudf </code></pre> <p><strong>myscript.py</strong></p> <pre><code>import environ env = environ.Env() # reading .env file environ.Env.read_env('.env') USERNAME = env.str(&quot;USERNAME&quot;) PASSWORD = env.str(&quot;PASSWORD&quot;) PASSWORDU = env.str(&quot;PASSWORDU&quot;) print(f&quot;Usermame {USERNAME}&quot;) # testuser print(f&quot;Password {PASSWORD}&quot;) # print(f&quot;Passwordu {PASSWORDU}&quot;) # sdfnsidufbsdufbsiudf </code></pre>
<python><python-3.x><environment-variables>
2024-04-14 08:43:38
2
428
Brian Obot
78,323,033
12,454,639
Im not sure I understand why Django wont let me do asynchronous database transactions
<p>I am trying to create a function that will get_or_create a user record for my discord bot in my django User model as soon as someone runs the <code>!home</code> command - but I am encountering an issue that (as far as I understand) is a normal part of django's existing model wherein it disallows asynchronous database transactions.</p> <p>I've tried using the built in sync_to_async method but it raises similiar issues. Is there something I can change in my settings file or something that I havent been able to find online that would be able achieve this functionality here?</p> <p>Here is my home command that is invoked by the command:</p> <pre><code>@bot.command(name=&quot;home&quot;, description=&quot;Displays the character sheet of your character in Oblivion After you have already created your guy&quot;) async def home_async(ctx): user_instance, created = await sync_to_async(User.objects.get_or_create( discord_id=discord_user_id, defaults={'username': ctx.author.name} )) if created: character = Character.objects.create(name=f&quot;New Character for {ctx.author.name}&quot;) user_instance.character = character user_instance.save() view = HomePageView(user=user_instance) await ctx.send(f&quot;Welcome {ctx.author.display_name}! Manage your character and inventory here.&quot;, view=view) </code></pre> <p>Currently this is the issue that I am encountering this error:</p> <pre><code>Traceback (most recent call last): File &quot;/mnt/m/ZocPy/OblivionAlchemy/OblivionAlchemy/venv/oblivion/lib/python3.10/site-packages/discord/ext/commands/bot.py&quot;, line 1350, in invoke await ctx.command.invoke(ctx) File &quot;/mnt/m/ZocPy/OblivionAlchemy/OblivionAlchemy/venv/oblivion/lib/python3.10/site-packages/discord/ext/commands/core.py&quot;, line 1029, in invoke await injected(*ctx.args, **ctx.kwargs) # type: ignore File &quot;/mnt/m/ZocPy/OblivionAlchemy/OblivionAlchemy/venv/oblivion/lib/python3.10/site-packages/discord/ext/commands/core.py&quot;, line 244, in wrapped raise CommandInvokeError(exc) from exc discord.ext.commands.errors.CommandInvokeError: Command raised an exception: SynchronousOnlyOperation: You cannot call this from an async context - use a thread or sync_to_async </code></pre>
<python><django><python-asyncio>
2024-04-14 07:17:46
1
314
Syllogism
78,322,897
3,224,483
Why does this code use more and more memory over time?
<p>Python: 3.11 Saxonche: 12.4.2</p> <p>My website keeps consuming more and more memory until the server runs out of memory and crashes. I isolated the problematic code to the following script:</p> <pre class="lang-python prettyprint-override"><code>import gc from time import sleep from saxonche import PySaxonProcessor xml_str = &quot;&quot;&quot; &lt;root&gt; &lt;stuff&gt;Lorem ipsum dolor sit amet, consectetur adipiscing elit. Vestibulum ac auctor ex. Nunc in tincidunt urna. Sed tincidunt eros lacus, sed pulvinar sem venenatis et. Donec euismod orci quis pellentesque sagittis. Donec at tortor in dui mattis facilisis. Pellentesque vel varius lectus. Nunc sed gravida risus, ac finibus elit. Etiam sollicitudin nunc a velit efficitur molestie in ac lectus. Donec vulputate orci odio, sit amet hendrerit odio rhoncus commodo.&lt;/stuff&gt; &lt;stuff&gt;Lorem ipsum dolor sit amet, consectetur adipiscing elit. Vestibulum ac auctor ex. Nunc in tincidunt urna. Sed tincidunt eros lacus, sed pulvinar sem venenatis et. Donec euismod orci quis pellentesque sagittis. Donec at tortor in dui mattis facilisis. Pellentesque vel varius lectus. Nunc sed gravida risus, ac finibus elit. Etiam sollicitudin nunc a velit efficitur molestie in ac lectus. Donec vulputate orci odio, sit amet hendrerit odio rhoncus commodo.&lt;/stuff&gt; &lt;stuff&gt;Lorem ipsum dolor sit amet, consectetur adipiscing elit. Vestibulum ac auctor ex. Nunc in tincidunt urna. Sed tincidunt eros lacus, sed pulvinar sem venenatis et. Donec euismod orci quis pellentesque sagittis. Donec at tortor in dui mattis facilisis. Pellentesque vel varius lectus. Nunc sed gravida risus, ac finibus elit. Etiam sollicitudin nunc a velit efficitur molestie in ac lectus. Donec vulputate orci odio, sit amet hendrerit odio rhoncus commodo.&lt;/stuff&gt; &lt;/root&gt; &quot;&quot;&quot; while True: print('Running once...') with PySaxonProcessor(license=False) as proc: proc.parse_xml(xml_text=xml_str) gc.collect() sleep(1) </code></pre> <p>This script consumes memory at a rate of about 0.5 MB per second. The memory usage does not plateau after a while. I have logs showing that memory usage continues to grow for hours until the server runs out of memory and crashes.</p> <p>Other things I tried that aren't shown above:</p> <ul> <li>Using a PyDocumentBuilder to parse the XML instead of a PySaxonProcessor. It didn't appear to change anything.</li> <li>Deleting the Saxon processor and the return value of <code>parse_xml()</code> using the <code>del</code> Python keyword. No change.</li> </ul> <p>I have to use Saxon instead of lxml because I need XPath 3.0 support.</p> <p>What am I doing wrong? How do I parse XML using Saxon in a way that doesn't leak?</p> <hr /> <p>A few folks have suggested that instantiating the PySaxonProcessor once before the loop will fix the leak. It doesn't. This still leaks:</p> <pre class="lang-python prettyprint-override"><code>with PySaxonProcessor(license=False) as proc: while True: print('Running once...') proc.parse_xml(xml_text=xml_str) gc.collect() sleep(1) </code></pre>
<python><saxon><saxon-c>
2024-04-14 06:14:07
3
3,659
Rainbolt
78,322,641
13,142,245
SqlModel - specify index type (B-Tree, Hash, etc.)
<p>I'm looking at documentation for SQLModel. The python library is written by the author of FastAPI, so I'm eager to leverage the seamless db integration via SqlModel.</p> <p>In the <a href="https://sqlmodel.tiangolo.com/tutorial/indexes/" rel="nofollow noreferrer">docs</a>, he illustrates what an index is and how it works by graphically depicting binary search. So I think that it's reasonable to conclude that SQLModel can index on a given variable via B-Tree by default.</p> <p>B-Trees (and binary search) make sense for range and inequality based queries. But equality or identity queries (<code>name=='John'</code>) can be greatly sped up through hash-based indices. But if you could only support one index type, B-Tree obviously makes more sense than hash.</p> <p>Here's an example</p> <pre><code>from sqlmodel import Field, Session, SQLModel, create_engine, select class Hero(SQLModel, table=True): id: int | None = Field(default=None, primary_key=True) name: str = Field(index=True) secret_name: str age: int | None = Field(default=None, index=True) </code></pre> <p>But suppose that I want to use a Hash index, can this be done through some SQLModel configuration?</p> <p>Under the hood, SQLModel is built on top of SQLAlchemy. Perhaps as a backup, one should drop down a level and define the index type through SQLAlchemy. Ideally, building Db-provider specific solutions (ex: Postgres) can be avoided.</p>
<python><sqlalchemy><orm><sqlmodel>
2024-04-14 03:38:53
1
1,238
jbuddy_13
78,322,538
4,117,496
ModuleNotFoundError: No module named 'torch' on AWS Batch GPU instance
<p>I have a job that runs on AWS Batch on a GPU instance, my application uses torch, i.e.</p> <pre><code>import torch </code></pre> <p>The Compute Environment has only one GPU instance, I was able to confirm that <code>torch</code> is available there by connecting to the instance via AWS Console and ran:</p> <pre><code>sh-4.2$ python3 Python 3.7.16 (default, Aug 30 2023, 20:37:53) [GCC 7.3.1 20180712 (Red Hat 7.3.1-15)] on linux Type &quot;help&quot;, &quot;copyright&quot;, &quot;credits&quot; or &quot;license&quot; for more information. &gt;&gt;&gt; import torch &gt;&gt;&gt; print(torch.__version__) 1.13.1+cu117 &gt;&gt;&gt; </code></pre> <p>However when I submit my Batch job, it fails saying</p> <pre><code> ModuleNotFoundError: No module named 'torch' </code></pre> <p>I looked into this and found that: My AWS Batch job definition has this Command:</p> <pre><code> CoolGPUJobDefinition: DependsOn: ComputeRole Type: AWS::Batch::JobDefinition Properties: Type: container ContainerProperties: Command: - &quot;/opt/prod/bin/python3&quot; - &quot;/opt/prod/bin/start.py&quot; </code></pre> <p>From the stacktrace of my application, it shows:</p> <pre><code>File &quot;/opt/prod/lib/python3.8/site-packages/cool_service/slowfast/utils/distributed.py&quot;, line 9, in &lt;module&gt; import torch ModuleNotFoundError: No module named 'torch' </code></pre> <p>But when I tried to <code>ls -a</code> in the <code>/opt</code> directory of this GPU instance, it didn't even have <code>prod</code>:</p> <pre><code>sh-4.2$ pwd /opt sh-4.2$ ls -a . .. aws containerd nvidia </code></pre> <p>Somehow there's a disconnect from what I can see in the GPU instance and how AWS Batch runs my application on this GPU instance.</p> <p>I'd like to understand:</p> <ol> <li>Why this disconnect/discrepancy?</li> <li>How could I resolve this module not found error in this case?</li> </ol> <p>Thanks!</p>
<python><amazon-web-services><tensorflow><pytorch><aws-batch>
2024-04-14 02:31:57
1
3,648
Fisher Coder
78,322,530
25,891
How to use OpenCV estimateAffinePartial2D
<p>I need to align two (actually hundreds of) images and I'm at a loss on how to do that in Python with OpenCV (which I have never heard of before). It looks like I should first estimate the transformation to apply as follows, and then apply it to one of the image (rinse and repeat hundreds of times). However even the simplest</p> <pre><code>import cv2 img1 = cv2.imread(&quot;img1.jpg&quot;, cv2.IMREAD_COLOR) img2 = cv2.imread(&quot;img2.jpg&quot;, cv2.IMREAD_COLOR) cv2.estimateAffinePartial2D(img1, img2) </code></pre> <p>fails with</p> <pre><code>cv2.error: OpenCV(4.9.0) /io/opencv/modules/calib3d/src/ptsetreg.cpp:1108: error: (-215:Assertion failed) count &gt;= 0 &amp;&amp; to.checkVector(2) == count in function 'estimateAffinePartial2D' </code></pre> <p>Stack overflow and OpenCV forum has a few questions about this problem but no solution, other than the one mentioned at <a href="https://stackoverflow.com/questions/35428739/opencv-estimateaffine3d-failes-with-cryptic-error-message">OpenCV estimateAffine3D failes with cryptic error message</a> (which is even more cryptic than the error message itself).</p> <p>How to do that estimation <strong>in Python</strong>?</p> <p>EDIT:</p> <pre><code>$ python Python 3.8.18 (default, Aug 25 2023, 13:20:30) [GCC 11.4.0] on linux Type &quot;help&quot;, &quot;copyright&quot;, &quot;credits&quot; or &quot;license&quot; for more information. &gt;&gt;&gt; import cv2 &gt;&gt;&gt; cv2.__version__ '4.9.0' </code></pre>
<python><opencv>
2024-04-14 02:29:49
1
17,904
Davide
78,322,515
5,264,310
Is there a way to make a long for loop to run faster?
<p>Im making a simple montecarlo simulator that takes in a 3*4 matrix of probabilities, and the number of iterations you want to simulate for. And the output is a table with all the results. Each row of the table is a list that contains: [iter no, random no, result1, result2] The logic is simple, just generate a random number and compare it with the cumulative probability. So when i try 1000000 iterations it takes 16 seconds or more and im trying to find out if its possible to get a better time. So far I've tried:</p> <ul> <li>Using memoization which i guess it doesnt make sense since the return data depends on a random number.</li> <li>Using numpy arrays and the method cumsum() to get cumulative probabilities.</li> <li>The best data structured i could come up with is a list of tuples of the possible outcomes and a 2 dimensional list to store the probabilities.</li> </ul> <pre><code>from random import random import numpy as np matrix = [ [0.2, 0.2, 0.05, 0.1], [0.7, 0.5, 0.6, 0.4], [0.2, 0.25, 0.2, 0.04], ] estados = [ (&quot;buena&quot;, &quot;buena&quot;), (&quot;buena&quot;, &quot;regular&quot;), (&quot;buena&quot;, &quot;critica&quot;), (&quot;buena&quot;, &quot;alta&quot;), (&quot;regular&quot;, &quot;buena&quot;), (&quot;regular&quot;, &quot;regular&quot;), (&quot;regular&quot;, &quot;critica&quot;), (&quot;regular&quot;, &quot;alta&quot;), (&quot;critica&quot;, &quot;buena&quot;), (&quot;critica&quot;, &quot;regular&quot;), (&quot;critica&quot;, &quot;critica&quot;), (&quot;critica&quot;, &quot;alta&quot;), ] estado_siguiente = {&quot;buena&quot;: 0, &quot;regular&quot;: 1, &quot;critica&quot;: 2, &quot;alta&quot;: 3} def get_result(rnd, probabilities, sig_estado=None): probabilities= ( probabilities[0] if (sig_estado == &quot;&quot; or sig_estado == &quot;alta&quot;) else probabilities[estado_siguiente[sig_estado] + 1] ) for i in range(len(probabilities[0])): if rnd &lt; probabilities[0][i]: return probabilities[1][i] return probabilities[1][len(probabilities) - 1] def start_simulation(probabilities, iter): vector_estado = [0, 0, &quot;&quot;, &quot;&quot;, 0, &quot;&quot;] table_final = list() #Here i have 4 accumulators for the probabilities of the 3 rows plus one for the whole table because a new result1 depends on the previous result2 meaning sometimes i just need the probs of just a row to_np_arr = np.array(probabilities) all_acc = (np.cumsum(to_np_arr), estados[:]) buena_acc = (all_acc[0][0:4], estados[0:4]) regular_acc = (all_acc[0][4:8], estados[4:8]) critico_acc = (all_acc[0][8:], estados[8:]) for _ in range(iter): rnd1 = random() estado1, estado2 = definir_condicion( rnd1, probabilities=[all_acc, buena_acc, regular_acc, critico_acc], sig_estado=vector_estado[3], ) #Add the list to the rest of the rows tabla_final.append(vector_estado) #Here I replace the old list with the new one vector_estado = [ vector_estado[0] + 1, &quot;%.3f&quot; % (rnd1), estado1, estado2, &quot;%.3f&quot; % (rnd2) if rnd2 else &quot;-&quot;, condicion_alta, ] return tabla_final </code></pre>
<python><list><numpy><loops>
2024-04-14 02:23:11
1
1,147
Xcecution
78,322,507
10,483,893
sqlalchemy session "Cannot use autocommit mode with future=True."
<p>Since which version sqlalchemy session <em>&quot;Cannot use autocommit mode with future=True.&quot;</em>? <a href="https://rattailproject.org/docs/wuttjamaican/_modules/sqlalchemy/orm/session.html" rel="nofollow noreferrer">https://rattailproject.org/docs/wuttjamaican/_modules/sqlalchemy/orm/session.html</a></p> <pre><code> if autocommit: if future: &lt;-- I don't really get this... raise sa_exc.ArgumentError( &quot;Cannot use autocommit mode with future=True.&quot; ) self.autocommit = True else: self.autocommit = False </code></pre> <p>This is breaking changes, I found out only after I had to reinstall Anaconda.</p> <pre><code>Python 3.9.19 (main, Mar 21 2024, 17:21:27) [MSC v.1916 64 bit (AMD64)] on win32 Type &quot;help&quot;, &quot;copyright&quot;, &quot;credits&quot; or &quot;license&quot; for more information. &gt;&gt;&gt; import sqlalchemy &gt;&gt;&gt; print(sqlalchemy.__version__) 1.4.32 &gt;&gt;&gt; </code></pre> <p>Anyone else run into this? In the past, <em>session.flush()</em> with auto-commit = True will propagate changes to database. Now:</p> <pre><code>session.flush() &lt;-- Should I remove Flush altogether? session.commit() </code></pre> <p>Thanks</p>
<python><sqlalchemy>
2024-04-14 02:20:47
1
1,404
user3761555
78,322,464
10,061,193
Including `.jinja` template files inside the Python distribution using `pyproject.toml`
<p>I'm working on a package. It has a console command that enables users to generate a template from the <code>.jinja</code> files within the distribution.</p> <p>My issue is that I can't include them (template files) in my distribution. Here is a tree-look of my package and where my templates are located. (<code>template/</code>)</p> <pre><code>mypack/ ├── mypack/ │ ├── package1/ │ ├── package2/ │   └── template │   └── {{tmpl_slug}} │   └── file1.py.jinja │ └── file2.json.jinja │ └── config.yml │ └── ... └── pyproject.yml </code></pre> <p>I want my distribution to have the exact <code>template/</code> directory with all its sub-files and sub-directories.</p> <p><strong>I can do what I need using a <code>MANIFEST.in</code> file that has the following single line in it.</strong></p> <pre><code>graft mypack/template </code></pre> <p>How can I do it without the MANIFEST file, just via <code>pyproject.toml</code>?!</p>
<python><templates><jinja2><packaging><python-packaging>
2024-04-14 01:40:26
0
394
Sadra
78,322,364
1,802,225
Ethereum: how to call Smart Contract functions directly with JSON-RPC method?
<p>I need to execute and call http JSON-RPC ethereum node to get results without web3py or web3js. How to do that? I have read there is <code>eth_call</code> method, but I didn't find any examples with ABI how to use. Below is example with <code>eth_getTransactionByHash</code>.</p> <pre class="lang-py prettyprint-override"><code>import json import requests from web3 import Web3 w3 = Web3(Web3.HTTPProvider('https://eth-pokt.nodies.app')) if not w3.is_connected(): exit(&quot;w3 not connected&quot;) # basic ERC-20: name(), symbol(), decimals() balanceOf() abi = json.loads('[{&quot;constant&quot;:true,&quot;inputs&quot;:[],&quot;name&quot;:&quot;name&quot;,&quot;outputs&quot;:[{&quot;name&quot;:&quot;&quot;,&quot;type&quot;:&quot;string&quot;}],&quot;payable&quot;:false,&quot;type&quot;:&quot;function&quot;},{&quot;constant&quot;:true,&quot;inputs&quot;:[],&quot;name&quot;:&quot;decimals&quot;,&quot;outputs&quot;:[{&quot;name&quot;:&quot;&quot;,&quot;type&quot;:&quot;uint8&quot;}],&quot;payable&quot;:false,&quot;type&quot;:&quot;function&quot;},{&quot;constant&quot;:true,&quot;inputs&quot;:[{&quot;name&quot;:&quot;_owner&quot;,&quot;type&quot;:&quot;address&quot;}],&quot;name&quot;:&quot;balanceOf&quot;,&quot;outputs&quot;:[{&quot;name&quot;:&quot;balance&quot;,&quot;type&quot;:&quot;uint256&quot;}],&quot;payable&quot;:false,&quot;type&quot;:&quot;function&quot;},{&quot;constant&quot;:true,&quot;inputs&quot;:[],&quot;name&quot;:&quot;symbol&quot;,&quot;outputs&quot;:[{&quot;name&quot;:&quot;&quot;,&quot;type&quot;:&quot;string&quot;}],&quot;payable&quot;:false,&quot;type&quot;:&quot;function&quot;}]') contract = w3.eth.contract(address='0x40E64405F18e4FB01c6fc39f4F0c78df5eF9D0E0', abi=abi) balance = contract.functions.balanceOf('0x7Ba83AC9243Ff520A28FC218758608F900E2d958').call(block_identifier=19649654) print(balance) # 976055241381593241650 # example of http request to get transaction tx = requests.post('https://eth-pokt.nodies.app', json={&quot;method&quot;:&quot;eth_getTransactionByHash&quot;,&quot;params&quot;:[&quot;0xd8feb6cd2e76f738e2b42da99f9746ade1dc8e4034ae79acec01fda7b293e6c1&quot;],&quot;id&quot;:1,&quot;jsonrpc&quot;:&quot;2.0&quot;}, headers={&quot;Content-Type&quot;: &quot;application/json&quot;}).json() </code></pre>
<python><ethereum><smartcontracts><web3py>
2024-04-14 00:30:41
0
1,770
sirjay
78,322,358
219,153
Does eval() have performance advantage in Mojo vs. Python?
<p>I would like to use <code>eval()</code> for validation of short function synthesis. The search space may be fairly large, e.g. more than a million different functions. Functions will be synthesized as a string containing their source code and executed with <code>eval()</code>.</p> <p>Will Mojo have a performance advantage vs. Python for this task, or it just calls Python's <code>eval()</code> through the compatibility layer? Is there a different method than string and <code>eval()</code>, which would work for my scenario?</p> <hr /> <p>Just to add some context, so concerns about security of calling <code>eval</code> are put to rest. What I'm doing is called Program Synthesis, see: <a href="https://en.wikipedia.org/wiki/Program_synthesis" rel="nofollow noreferrer">https://en.wikipedia.org/wiki/Program_synthesis</a>. I'm not calling <code>eval</code> on an unknown source code.</p>
<python><eval><mojolang>
2024-04-14 00:29:05
1
8,585
Paul Jurczak
78,322,249
17,517,315
Correct inheritance of methods from dataclass in Python
<p>I'm quite new to dataclasses and OOP in Python and I have been trying to figure out how to inheritance works. Before anything, I am really sorry if my question has been asked in one way or another, I really tried to look for similar questions but there is nothing out there that answered me. So apologies beforehand if there is anything unclear.</p> <p>Basically I want to achieve the following:</p> <pre><code> metadata.py @dataclass(frozen=True) class IngestionMetadata: file: str sheet_name: str col_converters: dict col_names: list = field(init=False) def __post_init__(self): object.__setattr__(self, 'col_names', list(self.col_converters.keys())) @classmethod def get_columns(cls): return list(cls.col_converters.keys()) @dataclass(frozen=True) class ProductGroupLookup(IngestionMetadata): file: str = 'product_group_table.xlsx' sheet_name: str = 'INGESTION' col_converters: dict = field(default_factory= lambda: { 'Product 1': pl.String, 'Product 2': pl.String, 'Product 3': pl.String, 'Product Category Code 4': pl.String }) </code></pre> <p>My goal is to be able to import the module from another script like:</p> <pre><code>from metadata import IngestionMetadata as im product_cols = im.ProductGroupLookup.get_columns() # Should output ['Product 1', 'Product 2'...] </code></pre> <p>I tried already separating the classes into subclasses or just putting them all inside the base class but without success. I would appreciate some guidance.</p>
<python><oop><python-dataclasses>
2024-04-13 23:04:00
0
391
matt.aurelio
78,322,169
1,498,178
How to merge Django model query results?
<p>I have to modify a legacy Django web application where there are 2 tables to store the same information using the same database structure, and the only difference is the names of the tables (and the corresponding Django models):</p> <ul> <li><code>ProgramAAssignments</code> model for <code>program_a_assignments</code> table</li> <li><code>ProgramBAssignments</code> model for <code>program_b_assignments</code> table</li> </ul> <p>They have 2 distinct views that query these models like so:</p> <pre><code>ProgramAAssignments.objects.all().order_by('-assignment_date') ProgramBAssignments.objects.all().order_by('-assignment_date') </code></pre> <p>I decided to leave the database untouched and just combine the views, but don't know how to merge these queries.</p>
<python><django-models>
2024-04-13 22:10:12
1
8,705
toraritte
78,322,167
2,030,532
Why tab completion in PyCharm python console is much slower than a terminal?
<p>I am running my python script in python console of PyCharm and the tab completion of a custom object attributes is pretty slow (unusable). However when I run the same code is a terminal using IPython the tab completion is fast. I tried increasing the memory limits of PyCharm with no avail. Is this a limitation of PyCharm that is not efficient in looking up attributes? Or some setting that needs to be corrected.</p>
<python><pycharm>
2024-04-13 22:09:29
0
3,874
motam79
78,321,952
8,713,442
Python Code failing : dedupe library error
<p>I am trying to learn about dedupe library . I am trying to match name which are more than 80% match.</p> <p>Sharing code and error . Please help</p> <pre><code>import dedupe from Levenshtein import distance def test(): # Sample data (replace with your actual library data) data = [ {'name': 'Alice Smith', 'address': '123 Main St', 'phone': '555-1212'}, {'name': 'Alice SmIth', 'address': '123 Main Street', 'phone': '555-1213'}, {'name': 'Bob Johnson', 'address': '456 Elm St', 'phone': '555-3434'}, {'name': 'Charlie Brown', 'address': '789 Maple Ave', 'phone': '555-5656'}, ] # Define fields for comparison (adjust based on your data) # Define data fields and comparison functions fields = [ {'field': 'name', 'comparators': ['name_similarity']}, ] # Define similarity functions - customize based on your matching criteria def name_similarity(s1, s2): # Implement your name comparison logic here (e.g., Levenshtein distance, etc.) distance1 = distance(s1, s2) similarity = 1 - (distance1 / max(len(s1), len(s2))) # Normalize distance to 0-1 similarity return similarity # Set thresholds for field-wise and overall similarity (adjust as needed) deduper = dedupe.Dedupe(fields) deduper.threshold( threshold=0.8) # Process the data for deduplication deduped_data = deduper.dedupe(data) # Print the deduplicated results print(&quot;Deduplicated Data:&quot;) for cluster in deduped_data: print(cluster) if __name__ == '__main__': test() </code></pre> <p>.....</p> <pre><code>C:\PythonProject\pythonProject\venv\Graph_POC\Scripts\python.exe C:\PythonProject\pythonProject\matching.py Traceback (most recent call last): File &quot;C:\PythonProject\pythonProject\venv\Graph_POC\Lib\site-packages\dedupe\datamodel.py&quot;, line 152, in typify_variables variable_type = definition[&quot;type&quot;] ~~~~~~~~~~^^^^^^^^ KeyError: 'type' During handling of the above exception, another exception occurred: Traceback (most recent call last): File &quot;C:\PythonProject\pythonProject\matching.py&quot;, line 45, in &lt;module&gt; test() File &quot;C:\PythonProject\pythonProject\matching.py&quot;, line 32, in test deduper = dedupe.Dedupe(fields) ^^^^^^^^^^^^^^^^^^^^^ File &quot;C:\PythonProject\pythonProject\venv\Graph_POC\Lib\site-packages\dedupe\api.py&quot;, line 1155, in __init__ self.data_model = datamodel.DataModel(variable_definition) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;C:\PythonProject\pythonProject\venv\Graph_POC\Lib\site-packages\dedupe\datamodel.py&quot;, line 42, in __init__ self.primary_variables, all_variables = typify_variables(variable_definitions) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File &quot;C:\PythonProject\pythonProject\venv\Graph_POC\Lib\site-packages\dedupe\datamodel.py&quot;, line 161, in typify_variables raise KeyError( KeyError: &quot;Missing variable type: variable specifications are dictionaries that must include a type definition, ex. {'field' : 'Phone', type: 'String'}&quot; Process finished with exit code 1 </code></pre>
<python><python-3.x><python-dedupe>
2024-04-13 20:36:13
1
464
pbh
78,321,701
8,713,442
pip install dedupe not working for python 3.11
<p>While doing pip install dedupe getting following error .</p> <p>Python installed is 3.11 . Please help I am trying collect cluster id using dedupe . Running it from pycharm .</p> <blockquote> <pre><code> Building wheels for collected packages: affinegap, doublemetaphone Building wheel for affinegap (pyproject.toml) ... error error: subprocess-exited-with-error × Building wheel for affinegap (pyproject.toml) did not run successfully. │ exit code: 1 ╰─&gt; [10 lines of output] running bdist_wheel running build running build_py creating build creating build\lib.win-amd64-cpython-311 creating build\lib.win-amd64-cpython-311\affinegap copying affinegap\__init__.py -&gt; build\lib.win-amd64-cpython-311\affinegap running build_ext building 'affinegap.affinegap' extension error: Microsoft Visual C++ 14.0 or greater is required. Get it with &quot;Microsoft C++ Build Tools&quot;: </code></pre> <p><a href="https://visualstudio.microsoft.com/visual-cpp-build-tools/" rel="nofollow noreferrer">https://visualstudio.microsoft.com/visual-cpp-build-tools/</a> [end of output]</p> <pre><code> note: This error originates from a subprocess, and is likely not a problem with pip. ERROR: Failed building wheel for affinegap Building wheel for doublemetaphone (pyproject.toml) ... error error: subprocess-exited-with-error × Building wheel for doublemetaphone (pyproject.toml) did not run successfully. │ exit code: 1 ╰─&gt; [10 lines of output] running bdist_wheel running build running build_py creating build creating build\lib.win-amd64-cpython-311 creating build\lib.win-amd64-cpython-311\doublemetaphone copying doublemetaphone\__init__.py -&gt; build\lib.win-amd64-cpython-311\doublemetaphone running build_ext building 'doublemetaphone.doublemetaphone' extension error: Microsoft Visual C++ 14.0 or greater is required. Get it with &quot;Microsoft C++ Build Tools&quot;: </code></pre> <p><a href="https://visualstudio.microsoft.com/visual-cpp-build-tools/" rel="nofollow noreferrer">https://visualstudio.microsoft.com/visual-cpp-build-tools/</a> [end of output]</p> <pre><code> note: This error originates from a subprocess, and is likely not a problem with pip. ERROR: Failed building wheel for doublemetaphone Failed to build affinegap doublemetaphone ERROR: Could not build wheels for affinegap, doublemetaphone, which is required to install pyproject.toml-based projects [notice] A new release of pip available: 22.3.1 -&gt; 24.0 [notice] To update, run: python.exe -m pip install --upgrade pip </code></pre> </blockquote>
<python><python-dedupe>
2024-04-13 18:56:47
1
464
pbh
78,321,696
15,474,507
GetMessagesReactionsRequest - distinguish reactions between 2 users
<p>Is there a system to distinguish the reaction of one user from another? <br>I'm testing with two my personal telegram accounts, but the terminal always gives me the same username, but in reality they are 2 users that makes reaction, not the same one.</p> <p><a href="https://i.sstatic.net/5EqYt.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/5EqYt.png" alt="enter image description here" /></a></p> <p>I test with this</p> <pre><code>from telethon.sync import TelegramClient from telethon.tl.functions.messages import GetMessagesReactionsRequest from telethon.tl.types import DocumentAttributeFilename api_id = '2++++' api_hash = 'f++++' channel_id = -100+++ client = TelegramClient('session_name', api_id, api_hash) async def main(): async with client: channel_entity = await client.get_input_entity(channel_id) messages = await client.get_messages(channel_entity, limit=100) me = await client.get_me() for message in messages: reactions = await client(GetMessagesReactionsRequest(channel_entity, [message.id])) if reactions.updates: # check if the reactions list is not empty for update in reactions.updates: if hasattr(update, 'reactions'): for reaction_count in update.reactions.results: content = message.message if message.message else &quot;No caption&quot; if hasattr(message, 'media') and message.media: if hasattr(message.media, 'document') and message.media.document: for attr in message.media.document.attributes: if isinstance(attr, DocumentAttributeFilename): content = attr.file_name username = me.username if me.username else &quot;ignoto&quot; print(f&quot;Post ID: {message.id} - Content: {content[:50]} - Reaction: {reaction_count.reaction.emoticon}, Count: {reaction_count.count}, User: @{username}&quot;) client.loop.run_until_complete(main()) </code></pre>
<python><telegram><telethon>
2024-04-13 18:55:11
0
307
Alex Doc
78,321,650
13,392,257
Keycloak: requests.exceptions.MissingSchema: Invalid URL 'None': No scheme supplied
<p>I am trying to run KeyCloak server and FastAPI application locally. I am following this tutorial <a href="https://fastapi-keycloak.code-specialist.com/quick_start/" rel="nofollow noreferrer">https://fastapi-keycloak.code-specialist.com/quick_start/</a></p> <p>My actions:</p> <p>1 Created docker-compose.yaml</p> <pre><code>version: '3.9' services: postgres: image: postgres:14.11 container_name: postgres_db volumes: - postgres_data:/var/lib/postgresql/data environment: POSTGRES_DB: keycloak POSTGRES_USER: keycloak POSTGRES_PASSWORD: password keycloak_server: image: quay.io/keycloak/keycloak:20.0.2 volumes: - ./realm-export.json:/opt/keycloak/data/import/realm.json container_name: keycloak_server environment: KC_DB: postgres KC_DB_URL: jdbc:postgresql://postgres:5432/keycloak KC_DB_USERNAME: keycloak KC_DB_PASSWORD: password KC_HOSTNAME: localhost KC_HOSTNAME_STRICT: false KC_HOSTNAME_STRICT_HTTPS: false KC_LOG_LEVEL: info KC_METRICS_ENABLED: true KC_HEALTH_ENABLED: true KEYCLOAK_ADMIN: admin KEYCLOAK_ADMIN_PASSWORD: admin command: - start-dev - --import-realm depends_on: - postgres ports: - 8280:8080 volumes: postgres_data: </code></pre> <p>Faced with error in the tutorial and fixed according to this <a href="https://howtodoinjava.com/devops/keycloak-script-upload-is-disabled/" rel="nofollow noreferrer">https://howtodoinjava.com/devops/keycloak-script-upload-is-disabled/</a> (removed authorizationSettings)</p> <p>I see the following log</p> <pre><code>keycloak_server | 2024-04-13 18:30:02,620 INFO [org.keycloak.quarkus.runtime.hostname.DefaultHostnameProvider] (main) Hostname settings: Base URL: &lt;unset&gt;, Hostname: localhost, Strict HTTPS: false, Path: &lt;request&gt;, Strict BackChannel: false, Admin URL: &lt;unset&gt;, Admin: &lt;request&gt;, Port: -1, Proxied: false keycloak_server | 2024-04-13 18:30:03,742 WARN [io.quarkus.agroal.runtime.DataSources] (main) Datasource &lt;default&gt; enables XA but transaction recovery is not enabled. Please enable transaction recovery by setting quarkus.transaction-manager.enable-recovery=true, otherwise data may be lost if the application is terminated abruptly keycloak_server | 2024-04-13 18:30:04,447 WARN [org.infinispan.PERSISTENCE] (keycloak-cache-init) ISPN000554: jboss-marshalling is deprecated and planned for removal keycloak_server | 2024-04-13 18:30:04,566 WARN [org.infinispan.CONFIG] (keycloak-cache-init) ISPN000569: Unable to persist Infinispan internal caches as no global state enabled keycloak_server | 2024-04-13 18:30:04,615 INFO [org.infinispan.CONTAINER] (keycloak-cache-init) ISPN000556: Starting user marshaller 'org.infinispan.jboss.marshalling.core.JBossUserMarshaller' keycloak_server | 2024-04-13 18:30:04,892 INFO [org.infinispan.CONTAINER] (keycloak-cache-init) ISPN000128: Infinispan version: Infinispan 'Triskaidekaphobia' 13.0.10.Final keycloak_server | 2024-04-13 18:30:05,275 INFO [org.keycloak.connections.infinispan.DefaultInfinispanConnectionProviderFactory] (main) Node name: node_591749, Site name: null keycloak_server | 2024-04-13 18:30:05,281 INFO [org.keycloak.broker.provider.AbstractIdentityProviderMapper] (main) Registering class org.keycloak.broker.provider.mappersync.ConfigSyncEventListener keycloak_server | 2024-04-13 18:30:06,011 INFO [org.keycloak.services] (main) KC-SERVICES0003: Not importing realm Test from file /opt/keycloak/bin/../data/import/realm.json. It already exists. keycloak_server | 2024-04-13 18:30:06,024 INFO [org.keycloak.services] (main) KC-SERVICES0003: Not importing realm Test from file /opt/keycloak/bin/../data/import/realm.json. It already exists. keycloak_server | 2024-04-13 18:30:06,163 INFO [io.quarkus] (main) Keycloak 20.0.2 on JVM (powered by Quarkus 2.13.3.Final) started in 4.937s. Listening on: http://0.0.0.0:8080 keycloak_server | 2024-04-13 18:30:06,163 INFO [io.quarkus] (main) Profile dev activated. keycloak_server | 2024-04-13 18:30:06,163 INFO [io.quarkus] (main) Installed features: [agroal, cdi, hibernate-orm, jdbc-h2, jdbc-mariadb, jdbc-mssql, jdbc-mysql, jdbc-oracle, jdbc-postgresql, keycloak, logging-gelf, narayana-jta, reactive-routes, resteasy, resteasy-jackson, smallrye-context-propagation, smallrye-health, smallrye-metrics, vault, vertx] keycloak_server | 2024-04-13 18:30:06,209 ERROR [org.keycloak.services] (main) KC-SERVICES0010: Failed to add user 'admin' to realm 'master': user with username exists keycloak_server | 2024-04-13 18:30:06,210 WARN [org.keycloak.quarkus.runtime.KeycloakMain] (main) Running the server in development mode. DO NOT use this configuration in production. </code></pre> <p>I am not sure that error <strong>Failed to add user 'admin' to realm 'master': user with username exists</strong> is critical (I think it happens because of the second run of the docker-compose file)</p> <p>Then I am running FastAPI application <code>python app.py</code> (file and configs copied from tutorial)</p> <pre><code>import uvicorn from fastapi import FastAPI, Depends from fastapi.responses import RedirectResponse from fastapi_keycloak import FastAPIKeycloak, OIDCUser app = FastAPI() idp = FastAPIKeycloak( server_url=&quot;http://localhost:8280/auth&quot;, client_id=&quot;test-client&quot;, client_secret=&quot;GzgACcJzhzQ4j8kWhmhazt7WSdxDVUyE&quot;, admin_client_secret=&quot;BIcczGsZ6I8W5zf0rZg5qSexlloQLPKB&quot;, realm=&quot;Test&quot;, callback_uri=&quot;http://localhost:8081/callback&quot; ) idp.add_swagger_config(app) ... if __name__ == '__main__': uvicorn.run('app:app', host=&quot;127.0.0.1&quot;, port=8081) </code></pre> <p>An see this error:</p> <pre><code>idp = FastAPIKeycloak( File &quot;/Users/aamoskalenko/0_root_folder/02_dev/33_keykloak_fastapi/venv/lib/python3.8/site-packages/fastapi_keycloak/api.py&quot;, line 165, in __init__ self._get_admin_token() # Requests an admin access token on startup File &quot;/Users/aamoskalenko/0_root_folder/02_dev/33_keykloak_fastapi/venv/lib/python3.8/site-packages/fastapi_keycloak/api.py&quot;, line 349, in _get_admin_token response = requests.post(url=self.token_uri, headers=headers, data=data, timeout=self.timeout) File &quot;/Users/aamoskalenko/0_root_folder/02_dev/33_keykloak_fastapi/venv/lib/python3.8/site-packages/requests/api.py&quot;, line 115, in post return request(&quot;post&quot;, url, data=data, json=json, **kwargs) File &quot;/Users/aamoskalenko/0_root_folder/02_dev/33_keykloak_fastapi/venv/lib/python3.8/site-packages/requests/api.py&quot;, line 59, in request return session.request(method=method, url=url, **kwargs) File &quot;/Users/aamoskalenko/0_root_folder/02_dev/33_keykloak_fastapi/venv/lib/python3.8/site-packages/requests/sessions.py&quot;, line 575, in request prep = self.prepare_request(req) File &quot;/Users/aamoskalenko/0_root_folder/02_dev/33_keykloak_fastapi/venv/lib/python3.8/site-packages/requests/sessions.py&quot;, line 486, in prepare_request p.prepare( File &quot;/Users/aamoskalenko/0_root_folder/02_dev/33_keykloak_fastapi/venv/lib/python3.8/site-packages/requests/models.py&quot;, line 368, in prepare self.prepare_url(url, params) File &quot;/Users/aamoskalenko/0_root_folder/02_dev/33_keykloak_fastapi/venv/lib/python3.8/site-packages/requests/models.py&quot;, line 439, in prepare_url raise MissingSchema( requests.exceptions.MissingSchema: Invalid URL 'None': No scheme supplied. Perhaps you meant https://None? </code></pre>
<python><keycloak><fastapi>
2024-04-13 18:37:28
0
1,708
mascai
78,321,309
13,174,189
How to group rows in dataframe based on timestamps ranges and mean group size maximization?
<p>I have a data frame:</p> <pre><code>prod_id timestamp1 timestamp2 1 2023-12-02 2023-12-01 2 2023-12-05 2023-12-01 3 2023-12-06 2023-12-01 4 2023-12-07 2023-12-01 5 2023-12-08 2023-12-01 6 2023-12-08 2023-12-02 7 2023-10-10 2023-09-02 8 2023-12-11 2023-12-22 9 2023-12-12 2023-12-24 </code></pre> <p>I want to group prod_id (create new parameter group_id) by timestamps ranges. the distribution of dates at timestamp1 must not exceed 3 days (so the difference between max and min within group_id must be no higher than 3 days). Similarly, the distribution of dates at timestamp 2 must not exceed 3 days (so the difference between max and min within group_id must be no higher than 3 days). I need to maximize the average number of prod_id per group_id. so here the desired result must be:</p> <pre><code>prod_id timestamp1 timestamp2 group_id 1 2023-12-02 2023-12-01 1 2 2023-12-05 2023-12-01 2 3 2023-12-06 2023-12-01 2 4 2023-12-07 2023-12-01 2 5 2023-12-08 2023-12-01 2 6 2023-12-08 2023-12-02 2 7 2023-10-10 2023-09-02 3 8 2023-12-11 2023-12-22 4 9 2023-12-12 2023-12-24 4 </code></pre> <p>as you see here prod_id is in group_id 2 because it will create larger group_id, than if it was in group_id 1, even though it could be in group_id 1 by conditions. so this grouping maximizes the mean number of prod_id per group_id.</p> <p>How to do it? I wrote this, but it doesn't have algorithmic maximization of the mean number of prod_id per group_id:</p> <pre><code>df['timestamp1'] = pd.to_datetime(df['timestamp1']) df['timestamp2'] = pd.to_datetime(df['timestamp2']) df_sorted = df.sort_values(by=['timestamp1', 'timestamp2']) group_id = 1 df_sorted['group_id'] = 0 while not df_sorted.empty: group_start = df_sorted.iloc[0] valid_rows_t1 = df_sorted[(df_sorted['timestamp1'] &lt;= group_start['timestamp1'] + pd.Timedelta(days=30))] valid_rows_t2 = valid_rows_t1[(valid_rows_t1['timestamp2'] &gt;= valid_rows_t1['timestamp2'].min()) &amp; (valid_rows_t1['timestamp2'] &lt;= valid_rows_t1['timestamp2'].min() + pd.Timedelta(days=30))] df_sorted.loc[valid_rows_t2.index, 'group_id'] = group_id df_sorted = df_sorted.drop(valid_rows_t2.index) group_id += 1 </code></pre> <p>Here is a code to create tha dataframe:</p> <pre><code># Create the dataframe data = { 'prod_id': [1, 2, 3, 4, 5, 6, 7, 8, 9], 'timestamp1': ['2023-12-02', '2023-12-05', '2023-12-06', '2023-12-07', '2023-12-08', '2023-12-08', '2023-10-10', '2023-12-11', '2023-12-12'], 'timestamp2': ['2023-12-01', '2023-12-01', '2023-12-01', '2023-12-01', '2023-12-01', '2023-12-02', '2023-09-02', '2023-12-22', '2023-12-24'] } df = pd.DataFrame(data) </code></pre>
<python><python-3.x><algorithm><function><genetic-algorithm>
2024-04-13 16:39:34
1
1,199
french_fries
78,321,225
12,560,241
Find if a set of lists of 2 values are the result of the possible combination of a n-value list
<p>I have used itertools to find all possible 2-element combination (without repetition) from a set of 10 elements to which I have applied some filtering to reduce the 2-element combination based on given condition</p> <p>For simplicity suppose the code below: the 10 elements are the number 1 to 10. The code I have used is following:</p> <pre><code>import numpy as np M = np.array(np.transpose(np.loadtxt('C:\\py\\10.csv'))) n=2 from itertools import combinations combos = list(filter(lambda e: ((e[0]+e[1]) &gt; 14 ) , combinations(M, n))) print (combos) </code></pre> <p>The output will then be</p> <pre><code>[(5.0, 10.0), (6.0, 9.0), (6.0, 10.0), (7.0, 8.0), (7.0, 9.0), (7.0, 10.0), (8.0, 9.0), (8.0, 10.0), (9.0, 10.0)] </code></pre> <p>What I am now trying to achieve, programmatically, is to compress the above results into a n-element list that would have generated it.</p> <p>For example I can say <code>(6.0, 9.0), (6.0, 10.0), (9.0, 10.0)</code> are the two element combination of a list of 3 elements <code>(6,9,10)</code></p> <p>I can also say <code>(7.0, 8.0), (7.0, 9.0), (7.0, 10.0), (8.0, 9.0), (8.0, 10.0), (9.0, 10.0)</code> are the two element combination of a list of 4 elements (7,8,9,10)</p> <p>I can also say that the (5.0, 10.0) pair it is not the result of the combination of a sublist of the 10 numbers</p> <p>If anyone came across this problem I'd like to know how this could be achieved</p>
<python><combinations><python-itertools><combinatorics><more-itertools>
2024-04-13 16:12:56
1
317
Marco_sbt
78,321,161
8,641,778
How to csat WinID handle to capsule in pyside6
<p>In pyqt, one can access wid as <code>sip.voidptr</code>, using <code>widget.winId()</code>, and then turn it to capsule object using <code>wid.ascapsule()</code></p> <p>In pyside6, when using <code>widget.winId()</code>, I can only get an int number.</p> <p>How can I get winId() as ptr like pyqt when using pyside6?</p> <p>I need ptr as capsule because the api I use give the following error:</p> <pre><code>return WNT_Window(ctypes.c_void_p(wid)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ TypeError: __init__(): incompatible constructor arguments. The following argument types are supported: 1. OCP.WNT.WNT_Window(theTitle: str, theClass: OCP.WNT.WNT_WClass, theStyle: int, thePxLeft: int, thePxTop: int, thePxWidth: int, thePxHeight: int, theBackColor: OCP.Quantity.Quantity_NameOfColor = &lt;Quantity_NameOfColor.Quantity_NOC_MATRAGRAY: 2&gt;, theParent: capsule = None, theMenu: capsule = None, theClientStruct: capsule = None) 2. OCP.WNT.WNT_Window(theHandle: capsule, theBackColor: OCP.Quantity.Quantity_NameOfColor = &lt;Quantity_NameOfColor.Quantity_NOC_MATRAGRAY: 2&gt;) </code></pre>
<python><qt><pyqt><pyside>
2024-04-13 15:53:29
1
323
C-Entropy
78,321,095
4,398,966
Spyder stdout not being flushed
<p>Updated with complete example. I'm running this in Spyder but it's been noted in comments that this works as expected from a windows command prompt.</p> <p>I have the following:</p> <pre><code>import sys import os class Portfolio: @classmethod def menu(cls): print(&quot;Main Menu&quot;) p = Portfolio() p.clear_screen() p.transaction() def clear_screen(self): os.system('cls' if os.name == 'nt' else 'clear') print(flush=True) sys.stdout.flush() def transaction(self): print(&quot;Transaction Menu&quot;) if __name__ == &quot;__main__&quot; : Portfolio.menu() </code></pre> <p>When I run this the screen clears AFTER not BEFORE transaction() is run. How can I correct that?</p>
<python><python-3.x><buffer><screen>
2024-04-13 15:33:45
0
15,782
DCR
78,321,036
2,649,312
Python how to pull data from dictionary
<p>Using the <code>weerlive.nl</code> API to pull weather data from the Dutch meteriological institute (KNMI), I get a dictionary with one key <code>'liveweer'</code> with a long set of tuples:</p> <pre><code>dict_values([[{'plaats': 'Tilburg', 'timestamp': '1713019987', 'time': '13-04-2024 16:53', 'temp': '23.3', 'gtemp': '22.3', 'samenv': 'Licht bewolkt', 'lv': '48', 'windr': 'ZW', 'windrgr': '225', 'windms': '6', 'winds': '4', 'windk': '11.7', 'windkmh': '21.6', 'luchtd': '1020.5', 'ldmmhg': '765', 'dauwp': '11.7', 'zicht': '40', 'verw': 'Vandaag zonnige perioden, droog en vrij warm. Zondag frisser', 'sup': '06:45', 'sunder': '20:35', 'image': 'lichtbewolkt', 'd0weer': 'bewolkt', 'd0tmax': '22', 'd0tmin': '12', 'd0windk': '3', 'd0windknp': '10', 'd0windms': '5', 'd0windkmh': '19', 'd0windr': 'ZW', 'd0windrgr': '225', 'd0neerslag': '0', 'd0zon': '0', 'd1weer': 'halfbewolkt', 'd1tmax': '15', 'd1tmin': '10', 'd1windk': '3', 'd1windknp': '8', 'd1windms': '4', 'd1windkmh': '15', 'd1windr': 'W', 'd1windrgr': '270', 'd1neerslag': '20', 'd1zon': '50', 'd2weer': 'regen', 'd2tmax': '11', 'd2tmin': '7', 'd2windk': '3', 'd2windknp': '10', 'd2windms': '5', 'd2windkmh': '19', 'd2windr': 'ZW', 'd2windrgr': '225', 'd2neerslag': '90', 'd2zon': '20', 'alarm': '0', 'alarmtxt': ''}]]) </code></pre> <p>For my RaspberryPi weather station I only need the values of <code>'temp'</code> and <code>'lv'</code> (luchtvochtigheid = humidity).</p> <p>How do I pull these values from the dictionary?</p>
<python><dictionary>
2024-04-13 15:15:19
2
811
jdelange
78,321,025
3,437,012
Why is this trivial numba function with a List input so slow
<pre><code>import numba from typing import List @numba.njit def test(a: List[int]) -&gt; int: return 1 test([i for i in range(2_000_000)]) </code></pre> <p>takes 2s and scales linearly with the size of the list. Wrapping the input argument with <code>numba.typed.List</code> takes even longer. (all the time is spent on the <code>numba.typed.List</code> call.</p> <p>The timings don't get better if the function is called multiple times (while only being defined once), i.e., this is not a matter of compilation time.</p> <p>Is there a way to tell numba to just use the list as is?</p> <p>In my actual application, the raw data comes from an external library that cannot return numpy arrays or numba lists directly, only Python lists.</p> <p>I'm using numba 0.59.1 and Python 3.12 on a 4 core Ubuntu22 laptop with 16GB RAM.</p>
<python><numpy><numba>
2024-04-13 15:12:13
1
2,370
Bananach
78,320,873
19,576,917
How to achieve concurrency with python telegram bot module
<p>I am using python telegram bot module to create a bot that takes an image in Braille as an input, translates it then returns the English translation. The method responsible for the translation is defined as follows,</p> <pre><code>async def start_translation(update: Update, context: CallbackContext): # translation logic return translated_text </code></pre> <p>The problem here is that this method takes too long to finish. So, when a user sends an image to the bot while another user has already sent an image and it is being translated, the second user has to wait until the first user is done. How can I employ parallel processing in my code so that multiple users can translate at the same time? I have tried to use <code>threading</code> module but it seems like it is not supported with <code>async</code> functions. Below is how the translation process starts:</p> <pre><code>def create_handle_photo(oVoice, oCorrect, oInference, current_dir): async def handle_photo(update: Update, context: CallbackContext): image_path = await get_image(update, context, current_dir) # download the image oProcessImage = ProcessImage(image_path) text = await start_translation(update, context, oProcessImage, oInference, oCorrect) if text != '': await generate_voice(update, context, current_dir, oVoice, text) await clean_up(current_dir, image_path) #delete downloaded files return handle_photo def main(): app = ApplicationBuilder().token(TOKEN).build() handle_photo_handler = create_handle_photo(oVoice, oCorrect, oInference, current_dir) app.add_handler(MessageHandler(filters.PHOTO, handle_photo_handler)) # handle photo app.run_polling() if __name__ == '__main__': main() </code></pre> <p>The input and output of the bot:</p> <p><a href="https://i.sstatic.net/j3nRY.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/j3nRY.png" alt="Bot receives image starts translation" /></a></p>
<python><parallel-processing><python-telegram-bot>
2024-04-13 14:22:00
0
488
Chandler Bong
78,320,840
16,569,183
Piping pip freeze and pip uninstall
<p>I can clean my current python environment in two steps</p> <pre class="lang-bash prettyprint-override"><code>pip freeze &gt; requirements.txt pip uninstall -r requirements.txt -y </code></pre> <p>I was wondering if it's possible to pipe these two commands to avoid creating the temporary file (and why or why not). The naive approach (below) does not seem to work</p> <pre class="lang-bash prettyprint-override"><code>pip freeze | pip uninstall -y -r </code></pre>
<python><pip>
2024-04-13 14:11:27
1
313
alfonsoSR
78,320,789
2,494,795
Register a Pandas Dataframe to Azure Machine Learning: NOT_SUPPORTED_API_USE_ATTEMPT
<p>I have been using the following code to register Pandas dataframes to Azure ML:</p> <pre><code>workspace = Workspace(&lt;subscription_id&gt;, &lt;resource_group&gt;, &lt;workspace_name&gt;) from azureml.core import Dataset datastore = workspace.get_default_datastore() ds = Dataset.Tabular.register_pandas_dataframe(df, datastore, &quot;output_dataset_name&quot;, &quot;description&quot;) </code></pre> <p>However, starting today, I am receiving a deprecation error message:</p> <pre><code> Message: [NOT_SUPPORTED_API_USE_ATTEMPT] The [read_pandas_dataframe] API has been deprecated and is no longer supported Payload: {&quot;rslex_version&quot;: &quot;2.22.2&quot;, &quot;api_name&quot;: &quot;read_pandas_dataframe&quot;, &quot;version&quot;: &quot;5.1.6&quot;} ActivityCompleted: Activity=register_pandas_dataframe, HowEnded=Failure, Duration=14.09 [ms], Info = {'activity_name': 'register_pandas_dataframe', 'activity_type': 'PublicApi', 'app_name': 'TabularDataset', 'source': 'azureml.dataset', 'version': '1.51.0.post1', 'dataprepVersion': '5.1.6', 'completionStatus': 'Failure', 'durationMs': 1.7}, Exception=NotImplementedError; read_pandas_dataframe is no longer supported. Write dataframe out as Parquet or csv and use read_parquet or read_csv instead.. </code></pre> <p>Can anybody recommend the best way to register Pandas dataframes as AML datasets after this deprecation? Thanks!</p>
<python><pandas><azure><azure-machine-learning-service>
2024-04-13 13:58:03
1
1,636
Irina
78,320,581
1,089,412
Pydantic - change data in nested model before validation
<p>I have this small example of nested object with pydantic.</p> <pre class="lang-py prettyprint-override"><code>from typing import Dict from pydantic import BaseModel, Field, ValidationError class UserType(BaseModel): name: str = Field(min_length=1) type: str = Field(min_length=1) class AppConfig(BaseModel): key1: int = Field(gt=0) objects: Dict[str, UserType] try: data = { &quot;key1&quot;: 1, &quot;objects&quot;: { &quot;type1&quot;: { &quot;name&quot;: &quot;Name 2&quot;, }, &quot;type2&quot;: { &quot;name&quot;: &quot;Name 1&quot; } } } c = AppConfig(**data) print(c.model_dump_json()) except ValidationError as e: print(e) </code></pre> <p>This obviously fail because <code>type</code> is not set in <code>UserType</code> model. My goal is to somehow set each <code>UserType.type</code> using associated key from <code>objects</code> dictionary before the actual validation. Something like:</p> <ul> <li>I pass data to model</li> <li>Model validates main keys</li> <li>Before validating nested objects it copies dict key as <code>type = key</code> inside data that is next is passed to UserType model</li> </ul> <p>Is this possible somehow with <code>pydantic</code>?</p> <p>I know I could do all this before passing data to main model but I want to know if this is possible in <code>pydantic</code> model.</p> <p>I also tried using <code>model_post_init()</code> method like this:</p> <pre class="lang-py prettyprint-override"><code>from typing import Dict from pydantic import BaseModel, Field, ValidationError class UserType(BaseModel): name: str = Field(min_length=1) type: str = Field(min_length=1) class AppConfig(BaseModel): key1: int = Field(gt=0) objects: Dict[str, UserType] def model_post_init(self, __context) -&gt; None: values = self.dict() for obj_type, obj in values[&quot;objects&quot;].items(): print(obj, obj_type) obj[&quot;type&quot;] = obj_type try: data = { &quot;key1&quot;: 1, &quot;objects&quot;: { &quot;type1&quot;: { &quot;name&quot;: &quot;Name 2&quot;, #&quot;type&quot;: &quot;t&quot; }, &quot;type2&quot;: { &quot;name&quot;: &quot;Name 1&quot;, #&quot;type&quot;: &quot;t&quot; } } } c = AppConfig(**data) print(c.model_dump_json()) except ValidationError as e: print(e) </code></pre> <p>But this method is executed after validation and validation failed earlier. Also tried to set some dummy type values in data payload and then override it in <code>model_post_init()</code> but this did not work at all and model had only original dummy values for type.</p>
<python><validation><pydantic>
2024-04-13 12:30:53
1
3,306
piotrekkr
78,320,551
5,473,482
Find intersection point between line and ellipse in Python
<p>I am trying to find the point which intersects with the line and the ellipse. I can find the point manually as seen in the code but how can I find it automatically. Also when the red point is inside the ellipse the green point should also be projected on to the ellipse when extending the line. How can I solve this? Code examples are appreciated! Thanks!</p> <p><a href="https://i.sstatic.net/zOZLf1J5.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/zOZLf1J5.png" alt="enter image description here" /></a></p> <pre><code>import cv2 import numpy as np # Parameters ellipse_center = np.array([238, 239]) ellipse_axes = np.array([150, 100]) # Semi-major and semi-minor axes ellipse_angle = np.radians(70) # Angle of rotation (in radians) line_point1 = ellipse_center # Point on the line line_point2 = np.array([341, 125]) # Another point on the line (moving point) # Initialize OpenCV window window_name = &quot;Projecting Point on Rotated Ellipse&quot; cv2.namedWindow(window_name) canvas = np.zeros((400, 400, 3), dtype=np.uint8) while True: # Clear canvas canvas.fill(0) # Draw the rotated ellipse cv2.ellipse(canvas, tuple(ellipse_center), tuple(ellipse_axes), np.degrees(ellipse_angle), 0, 360, (255, 255, 255), 2) # Draw the moving point cv2.circle(canvas, (int(line_point2[0]), int(line_point2[1])), 5, (0, 0, 255), -1) # Draw the center point of the ellipse cv2.circle(canvas, tuple(ellipse_center), 5, (0, 255, 255), -1) # Draw the line between moving point and the center of the ellipse cv2.line(canvas, (int(line_point2[0]), int(line_point2[1])), tuple(ellipse_center), (0, 0, 255), 1) intersection_points = (310, 159) # &lt;- How to find this? cv2.circle(canvas, intersection_points, 5, (0, 255, 0), -1) # Show canvas cv2.imshow(window_name, canvas) # Exit if 'q' is pressed if cv2.waitKey(1) &amp; 0xFF == ord('q'): break # Close OpenCV window cv2.destroyAllWindows() </code></pre> <p>When using the ellipse equation and the line equation I get this:</p> <pre><code>import cv2 import numpy as np from scipy.optimize import fsolve # Parameters ellipse_center = np.array([238, 239]) ellipse_axes = np.array([150, 100]) # Semi-major and semi-minor axes ellipse_angle = np.radians(70) # Angle of rotation (in radians) line_point1 = ellipse_center # Point on the line line_point2 = np.array([341, 125]) # Another point on the line (moving point) # Ellipse equation: (x - h)^2 / a^2 + (y - k)^2 / b^2 = 1, where (h, k) is the center of the ellipse def ellipse_eq(params): x, y = params h, k = ellipse_center a, b = ellipse_axes return ((x - h) ** 2 / a ** 2) + ((y - k) ** 2 / b ** 2) - 1 # Line equation: y = mx + c def line_eq(x): m = (line_point2[1] - line_point1[1]) / (line_point2[0] - line_point1[0]) c = line_point1[1] - m * line_point1[0] return m * x + c # Intersection of the ellipse and the line def intersection_point(): x_intercept = fsolve(lambda x: ellipse_eq([x, line_eq(x)]), line_point2[0]) y_intercept = line_eq(x_intercept) return x_intercept[0], y_intercept[0] # Initialize OpenCV window window_name = &quot;Projecting Point on Rotated Ellipse&quot; cv2.namedWindow(window_name) canvas = np.zeros((400, 400, 3), dtype=np.uint8) while True: # Clear canvas canvas.fill(0) # Draw the rotated ellipse cv2.ellipse(canvas, tuple(ellipse_center), tuple(ellipse_axes), np.degrees(ellipse_angle), 0, 360, (255, 255, 255), 2) # Draw the moving point cv2.circle(canvas, (int(line_point2[0]), int(line_point2[1])), 5, (0, 0, 255), -1) # Draw the center point of the ellipse cv2.circle(canvas, tuple(ellipse_center), 5, (0, 255, 255), -1) # Draw the line between moving point and the center of the ellipse cv2.line(canvas, (int(line_point2[0]), int(line_point2[1])), tuple(ellipse_center), (0, 0, 255), 1) # Find and draw the intersection point intersection_points = intersection_point() cv2.circle(canvas, (int(intersection_points[0]), int(intersection_points[1])), 5, (0, 255, 0), -1) # Show canvas cv2.imshow(window_name, canvas) # Exit if 'q' is pressed if cv2.waitKey(1) &amp; 0xFF == ord('q'): break # Close OpenCV window cv2.destroyAllWindows() </code></pre> <p><a href="https://i.sstatic.net/8M6M2H0T.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/8M6M2H0T.png" alt="enter image description here" /></a></p> <p>Which does not solve my problem. What am I doing wrong here?</p> <p>EDIT: Using MBo's approach with the code:</p> <pre><code>import cv2 import numpy as np # Parameters ellipse_center = np.array([238, 239]) ellipse_axes = np.array([150, 100]) # Semi-major and semi-minor axes ellipse_angle = np.radians(70) # Angle of rotation (in radians) line_point1 = ellipse_center # Point on the line (ellipse center) line_point2 = np.array([341, 125]) # Another point on the line (moving point) # Initialize OpenCV window window_name = &quot;Projecting Point on Rotated Ellipse&quot; cv2.namedWindow(window_name) canvas = np.zeros((400, 400, 3), dtype=np.uint8) while True: # Clear canvas canvas.fill(0) # Draw the rotated ellipse cv2.ellipse(canvas, tuple(ellipse_center), tuple(ellipse_axes), np.degrees(ellipse_angle), 0, 360, (255, 255, 255), 2) # Draw the moving point cv2.circle(canvas, (int(line_point2[0]), int(line_point2[1])), 5, (0, 0, 255), -1) # Draw the center point of the ellipse cv2.circle(canvas, tuple(ellipse_center), 5, (0, 255, 255), -1) # Draw the line between moving point and the center of the ellipse cv2.line(canvas, (int(line_point2[0]), int(line_point2[1])), tuple(ellipse_center), (0, 0, 255), 1) # Calculate the intersection point t = np.arctan2(line_point2[1] - ellipse_center[1], line_point2[0] - ellipse_center[0]) - np.arctan2(np.sin(ellipse_angle) * ellipse_axes[0], np.cos(ellipse_angle) * ellipse_axes[1]) x_intersection = ellipse_center[0] + ellipse_axes[0] * np.cos(t) * np.cos(ellipse_angle) - ellipse_axes[1] * np.sin(t) * np.sin(ellipse_angle) y_intersection = ellipse_center[1] + ellipse_axes[0] * np.cos(t) * np.sin(ellipse_angle) + ellipse_axes[1] * np.sin(t) * np.cos(ellipse_angle) intersection_point = (int(x_intersection), int(y_intersection)) # Draw the intersection point cv2.circle(canvas, intersection_point, 5, (0, 255, 0), -1) # Show canvas cv2.imshow(window_name, canvas) # Exit if 'q' is pressed if cv2.waitKey(1) &amp; 0xFF == ord('q'): break # Close OpenCV window cv2.destroyAllWindows() </code></pre> <p>Gets me:</p> <p><a href="https://i.sstatic.net/s1mAK.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/s1mAK.png" alt="enter image description here" /></a></p>
<python><math><geometry><ellipse>
2024-04-13 12:17:38
2
1,047
Blind0ne
78,320,404
3,047,069
Facing problems saving JSON data to postgres using Python FastAPI
<p>I am using python and Fastapi to build a project. I am fairly new to both the technologies, and now getting stuck at a task of uploading a JSON file to an endpoint and then saving its contents to postgresSQL.</p> <p>Here are some details:</p> <p><strong>vehicles.json: It can have list of 500-600 entries in the format shown below</strong></p> <pre><code>{ &quot;vehicleList&quot;: [ { &quot;id&quot;: 1, &quot;coordinate&quot;: { &quot;latitude&quot;: 54.5532316, &quot;longitude&quot;: 67.0087783 }, &quot;condition&quot;: &quot;GOOD&quot; }, { &quot;id&quot;: 2, &quot;coordinate&quot;: { &quot;latitude&quot;: 37.442316, &quot;longitude&quot;: 38.0087783 }, &quot;condition&quot;: &quot;BREAKDOWM&quot; }, ] } </code></pre> <p><strong>connect.py:</strong></p> <pre><code>vehicle_data = Table( &quot;vehicle_data&quot;, metadata_obj, Column(&quot;coordinates&quot;, ARRAY(Float), nullable=False), Column(&quot;condition&quot;, String(6), nullable=False), Column(&quot;id&quot;, Integer, primary_key=True), ) </code></pre> <p><strong>router.py:</strong></p> <pre><code>@router.post(&quot;/upload&quot;) async def upload_json(file: UploadFile = File(...)): try: # Read the uploaded file as bytes contents = await file.read() # Decode the bytes to string assuming it's JSON decoded_content = contents.decode(&quot;utf-8&quot;) # Parse the JSON content json_data = json.loads(decoded_content) fields = [ 'coordinates', #List of floats 'condition', #str 'id' #int ] for item in json_data: my_data = [tuple(item[field] for field in fields) for item in json_data] insert_query = &quot;INSERT INTO vehicle_data (coordinates, condition, id) VALUES %s&quot; execute_values(insert_query, tuple(my_data)) return JSONResponse(status_code=200, content={&quot;message&quot;: &quot;JSON file uploaded successfully&quot;, &quot;data&quot;: json_data}) except Exception as e: return JSONResponse(content={&quot;error&quot;: str(e)}, status_code=500) </code></pre> <p><strong>Expected output: I want to only store all the entries of JSON into postgress rows</strong></p> <p><a href="https://i.sstatic.net/r3PvF.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/r3PvF.png" alt="enter image description here" /></a></p> <p><strong>What I tried:</strong></p> <p>I tried reading about file upload on fastapi as well as went through multiple stackoverflow questions, and even found a question that was close to the solution I am trying which is:</p> <pre><code>for item in json_data: my_data = [tuple(item[field] for field in fields) for item in json_data] insert_query = &quot;INSERT INTO vehicle_data (coordinates, condition, id) VALUES %s&quot; execute_values(insert_query, tuple(my_data)) </code></pre> <p>But I suspect that the mistake is somewhere in the above code which I am not able to solve, I am able to correctly parse data in the form of python dict object till <code>json_data = json.loads(decoded_content)</code> line.</p> <p>It'll be very helpful if you just help me to point in right direction or state what needs to be replaced instead of the lines written above.</p> <p>Here is the error I am getting: <a href="https://i.sstatic.net/8bWVi.png" rel="nofollow noreferrer"><img src="https://i.sstatic.net/8bWVi.png" alt="enter image description here" /></a></p> <p>Have a great day!</p>
<python><json><postgresql><fastapi><psycopg2>
2024-04-13 11:21:27
0
821
Radheya
78,320,399
5,378,816
Type hint for a function returning the same type not working
<p>What is wrong in this example? Isn't the <code>str -&gt; str</code> case included in <code>T -&gt; T</code> ?</p> <pre><code>from typing import TypeVar T = TypeVar(&quot;T&quot;) def unch(arg:T) -&gt; T: if isinstance(arg, str): return arg return arg </code></pre> <p><code>mypy</code> checking result:</p> <pre><code># error: Incompatible return value type (got &quot;str&quot;, expected &quot;T&quot;) </code></pre> <hr /> <p>Update 1: deleted</p> <hr /> <p>Update 2: don't know why, but this works:</p> <pre><code>from typing import TypeVar, Any T = TypeVar(&quot;T&quot;, str, Any) def unch(arg:T) -&gt; T: if isinstance(arg, str): return arg return arg </code></pre>
<python><mypy><python-typing>
2024-04-13 11:19:05
0
17,998
VPfB
78,320,329
6,303,377
Can't run inference SDK on AWS Lambda due to error with multiprocessing
<p>I am running this code on AWS Lambda</p> <pre><code>import os from inference_sdk import InferenceHTTPClient def handler(event, context): client = InferenceHTTPClient(api_url=&quot;https://detect.roboflow.com&quot;, api_key=os.environ[&quot;ROBOFLOW_API_KEY&quot;]) img_path = &quot;./pizza.jpg&quot; return client.infer(img_path, model_id=&quot;pizza-identifier/3&quot;) </code></pre> <p>As part of a docker container that looks like this:</p> <pre><code>FROM public.ecr.aws/lambda/python:3.11 RUN yum install -y mesa-libGL COPY requirements.txt ${LAMBDA_TASK_ROOT} RUN pip install -r requirements.txt COPY pizza.jpg ${LAMBDA_TASK_ROOT} COPY lambda_function.py ${LAMBDA_TASK_ROOT} CMD [ &quot;lambda_function.handler&quot; ] </code></pre> <p>My requirements.txt contains nothing but <code>inference==0.9.17</code></p> <p>When the code runs I get the following error. I have been trying to fix this and tried workarounds but to no avail. I understand that the error is somehow related to multiprocessing. I found <a href="https://stackoverflow.com/questions/59638035/using-python-multiprocessing-queue-inside-aws-lambda-function">this post</a> from which I understand that multiprocessing isn't possible on AWS Lambda, however, my script does not control or trigger any multiprocessing. I am incredibly frustrated since I've been working on this for 9 hours now and would appreciate any hints!</p> <p>This is the full error:</p> <pre><code>{ &quot;errorMessage&quot;: &quot;[Errno 38] Function not implemented&quot;, &quot;errorType&quot;: &quot;OSError&quot;, &quot;requestId&quot;: &quot;703be804-fd86-4b44-88f9-ac54c87717be&quot;, &quot;stackTrace&quot;: [ &quot; File \&quot;/var/task/lambda_function.py\&quot;, line 10, in handler\n return client.infer(img_path, model_id=\&quot;pizza-identifier/3\&quot;)\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/site-packages/inference_sdk/http/client.py\&quot;, line 82, in decorate\n return function(*args, **kwargs)\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/site-packages/inference_sdk/http/client.py\&quot;, line 237, in infer\n return self.infer_from_api_v0(\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/site-packages/inference_sdk/http/client.py\&quot;, line 299, in infer_from_api_v0\n responses = execute_requests_packages(\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/site-packages/inference_sdk/http/utils/executors.py\&quot;, line 42, in execute_requests_packages\n responses = make_parallel_requests(\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/site-packages/inference_sdk/http/utils/executors.py\&quot;, line 58, in make_parallel_requests\n with ThreadPool(processes=workers) as pool:\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/multiprocessing/pool.py\&quot;, line 930, in __init__\n Pool.__init__(self, processes, initializer, initargs)\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/multiprocessing/pool.py\&quot;, line 196, in __init__\n self._change_notifier = self._ctx.SimpleQueue()\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/multiprocessing/context.py\&quot;, line 113, in SimpleQueue\n return SimpleQueue(ctx=self.get_context())\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/multiprocessing/queues.py\&quot;, line 341, in __init__\n self._rlock = ctx.Lock()\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/multiprocessing/context.py\&quot;, line 68, in Lock\n return Lock(ctx=self.get_context())\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/multiprocessing/synchronize.py\&quot;, line 169, in __init__\n SemLock.__init__(self, SEMAPHORE, 1, 1, ctx=ctx)\n&quot;, &quot; File \&quot;/var/lang/lib/python3.11/multiprocessing/synchronize.py\&quot;, line 57, in __init__\n sl = self._semlock = _multiprocessing.SemLock(\n&quot; ] } </code></pre>
<python><python-3.x><amazon-web-services><aws-lambda>
2024-04-13 10:50:15
0
1,789
Dominique Paul
78,320,199
1,920,003
How to improve Nginx/Uvicorn upload speed in FastAPI application?
<p>I have an Ubuntu server, running a Nginx reverse proxy for a FastAPI application using Uvicorn. The server is an AWS EC2 g4n.xlarge in Virginia. On the frontend I'm using HTMX My personal home upload speed is close to 1GBPs, fiber optics, I'm connected via the ethernet cable.</p> <p>The application uploads the file first and then processes it. To upload 48mb, 35 mins mp3 file, the application takes 10min or so, it's not acceptable. In fact, given my fast internet speed, anything above 30 seconds or a minute is not acceptable. I already tried chunk upload, didn't make a difference</p> <pre><code>async with aiofiles.open(video_file_path, &quot;wb&quot;) as out_file: while True: content = await file.read(1024 * 1024) # Read chunks of 1 MB if not content: break await out_file.write(content) </code></pre> <p>I believe the issue has to do with NGINX, because on my PC for testing, I use Uvicorn directly without NGINX and the upload is instant. My upload speed is fast and EC2 upload speed is fast, so the only one left to blame is Nginx, I think</p> <p>nginx config</p> <pre><code>server { server_name example.com; # Increase client max body size client_max_body_size 6000M; # Allow 6GB uploads # Adjust timeouts client_body_timeout 7200s; client_header_timeout 7200s; location / { proxy_pass http://127.0.0.1:8001; # Proxy pass to the FastAPI port proxy_set_header Host $host; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; # Necessary for WebSocket support proxy_http_version 1.1; proxy_set_header Upgrade $http_upgrade; proxy_set_header Connection &quot;upgrade&quot;; # Proxy timeouts proxy_read_timeout 7200s; proxy_connect_timeout 7200s; proxy_send_timeout 7200s; } </code></pre>
<python><nginx><fastapi><uvicorn>
2024-04-13 09:55:03
1
5,375
Lynob
78,320,086
710,955
How get the value of a 'charset' meta element with Xpath?
<p>With Selenium webdriver, I'm trying to parse a <code>charset</code> meta element from a page.</p> <pre class="lang-html prettyprint-override"><code>&lt;meta charset=&quot;UTF-8&quot;&gt; </code></pre> <p>This is what I have so far</p> <pre class="lang-py prettyprint-override"><code>from selenium.webdriver.common.by import By xpath='//meta[@charset]' charset_meta_element = driver.find_element(By.XPATH, xpath) </code></pre> <p>I get a <code>WebElement</code> object. How from this element I can get the value (eg: 'UTF-8')?</p>
<python><selenium-webdriver><xpath><html-meta>
2024-04-13 09:03:57
2
5,809
LeMoussel