Asthenia0412 opened a new pull request, #53727:
URL: https://github.com/apache/arrow/pull/53727
## Background
A stale guard in `from_dataframe.py` rejects pandas float16 columns with NaN
values, claiming `pc.is_nan` has no float16 kernel.
## Problem
Every `from_dataframe()` call with a float16 DataFrame raises
`NotImplementedError`.
## Solution
Removed the outdated guard block. PR #46866 already added float16 support to
`pc.is_nan` (GH-45083), so float16 now flows through the same code path as
other float widths.
## Testing
Verified that `from_dataframe(pd.DataFrame({'h': np.array([1.5, np.nan],
dtype='<f2')}))` now succeeds and produces a table with correct null handling.
Also updated the existing test `test_pandas_to_pyarrow_float16_with_missing`
to verify correct behavior instead of expecting `NotImplementedError`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
