Skip to content

FIX: drop(index=[pd.NA]) fails for Arrow-backed index (#63304)#63933

Closed
dedlYTiTAN wants to merge 19 commits intopandas-dev:mainfrom
dedlYTiTAN:fix-gh63304-v3
Closed

FIX: drop(index=[pd.NA]) fails for Arrow-backed index (#63304)#63933
dedlYTiTAN wants to merge 19 commits intopandas-dev:mainfrom
dedlYTiTAN:fix-gh63304-v3

Conversation

@dedlYTiTAN
Copy link

Description:
This PR resolves an ArrowInvalid error when dropping pd.NA from Arrow-backed indices. It ensures the value_set in ArrowExtensionArray.isin is correctly cast to the array's native type.

Key Verifications:

  • Validated inplace=True stability.
  • Confirmed no type coercion to NumPy.
  • Verified data alignment across columns.

Your Name added 19 commits January 28, 2026 10:39
This PR resolves an ArrowInvalid error when dropping pd.NA from Arrow-backed indices. It ensures the value_set in ArrowExtensionArray.isin is correctly cast to the array's native type.
Fixes CI failures on older environments where ArrowDtype cannot be instantiated/used. Requires pyarrow>=13.0.0.
@jbrockmendel jbrockmendel added the AI Slop Suspected of being AI-generated, which is not welcome. label Jan 29, 2026
@mroeschke
Copy link
Member

You've already been warning about submitting AI generated pull requests #63898 (comment)

One more similar contribution and we'll have to ban your contributions. Closing

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

AI Slop Suspected of being AI-generated, which is not welcome.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

BUG(pandas 3.0 regression): drop(index=...) doesn't accept NA values when using arrow dtype in index

3 participants