[Warning] UserWarning: Using categorical_feature in Dataset.

[ucasiggcas]

hi,dear all
I tried according to the link docs, but has the warning, could you pls help me ?
train_data = lgb.Dataset(data, label=label, feature_name=['c1', 'c2', 'c3'], categorical_feature=['c3'])

Warning (from warnings module):
  File "D:\python\lib\site-packages\lightgbm\basic.py", line 1291
    warnings.warn('Using categorical_feature in Dataset.')
UserWarning: Using categorical_feature in Dataset.

[StrikerRUS]

Hi @ucasiggcas !
Everything is fine! This warning is to make you confident that categorical features are not lost. Refer to
#3625 (comment).

[domkm]

This is extremely odd. I've never come across a library that logs a warning when things are working correctly. I'd recommend at least updating the message to make it clear to the user that nothing is wrong.

[jameslamb]

This is an open source project, you're welcome to propose a pull request changing its behavior.

[domkm]

Apologies if my tone was impolite. It was unintentional but, upon rereading, I believe it may have been.

Would a PR to remove this warning be merged?

[jameslamb]

Thanks for saying that!

Would a PR to remove this warning be merged?

Definitely, I agree with your assessment. But I looked tonight and found that actually...this warning has already been removed from lightgbm.

This issue is a duplicate of #3379, and the warning was removed in #4768 (November 2021).

Unfortunately, most changes merged to master after October 27, 2021 still haven't made it into a release. So if you pip install lightgbm, you may still experience this warning. The fix will be in v4.0.0, the next release of LightGBM. You can follow #5153 to be automatically notified when that release comes out. We apologize for the long delay, and are trying to get that release ready as fast as possiblle.

[domkm]

That's great! Thank you very much.

[github-actions]

This issue has been automatically locked since there has not been any recent activity since it was closed.
To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues
including a reference to this.