tqdm.pandas() progress_apply issue

[yanxke]

Object of type "Series[Unknown]" is not callable on the progress_apply() part.

Environment data

• Language Server version: v2023.7.30
• OS and version: Windows 11
• Python version (& distribution if applicable, e.g. Anaconda): 3.10.6

Code Snippet

import pandas as pd
from tqdm import tqdm

def foo(df: pd.DataFrame) -> pd.DataFrame:
    def bar(row: pd.Series):
            return pd.NA

    tqdm.pandas()
    df["distance"] = df.progress_apply(bar, axis=1)

    return df

[judej]

Thanks for the report, unfortunately, this is not enough data for us to debug. I've tried it and don't see the Series[Unknown]" is not callable on version v2023.8.50

If this continues to be a problem, would you please reopen the issue and add the information in the report as described in the troubleshooting guide?

[yanxke]

It still appears in the latest version.

2023-09-11 09:09:43.655 [info] Experiment 'pythonPromptNewFormatterExt' is active
2023-09-11 09:09:43.655 [info] Experiment 'pythonPromptNewToolsExt' is active
2023-09-11 09:09:43.655 [info] Experiment 'pythonTestAdaptercf' is active
2023-09-11 09:09:43.655 [info] Test server listening.
2023-09-11 09:09:43.655 [info] Python interpreter path: C:\Python311\python.exe
2023-09-11 09:09:53.332 [info] > conda info --json
2023-09-11 09:09:55.463 [info] > ~\.conda\Scripts\conda.exe info --json
2023-09-11 09:09:55.472 [info] > ~\miniconda3\envs\labelme\Scripts\conda.exe info --json
2023-09-11 09:09:55.478 [info] > ~\miniconda3\Scripts\conda.exe info --json
2023-09-11 09:09:55.483 [info] > ~\miniconda3\envs\crawlpg\Scripts\conda.exe info --json
2023-09-11 09:09:56.708 [info] Starting Pylance language server.

logs.txt

[yanxke]

Logs attached. @judej could you reopen the issue?

[debonte]

Pylance ships with the pandas stubs from pandas-stubs. These stubs do not show a progress_apply member on DataFrame, so Pylance is falling back to the DataFrame.__getattr__ stub which returns a Series.

Static type checkers like Pylance don't have any way of knowing that tqdm has dynamically added the progress_apply member to DataFrame. If you feel that this should be improved, I would suggest filing an issue in the tqdm issue tracker.