Fixed regression in Series.duplicated for categorical dtype with bool categories #44356
Conversation
phofl
commented
Nov 8, 2021
phofl
commented
- closes REGR: Series.duplicated with category dtype and nulls raises ValueError #44351
- tests added / passed
- Ensure all linting tests pass, see here for how to run them
- whatsnew entry
phofl
added
Categorical
There was a problem hiding this comment.
Thanks @phofl Generally LGTM.
I think could maybe also mention drop_duplicates in the release note, but probably don't need extra tests.
However, the code sample in the issue OP is based on test_drop_duplicates_categorical_bool in pandas/tests/series/methods/test_drop_duplicates.py so could make sense to co-locate the tests and also test for drop_duplicates
|
Added the test and modified the release note |
| ) | ||
| result = ser.drop_duplicates() | ||
| expected = Series( | ||
| Categorical([True, False, np.nan], categories=[True, False], ordered=True), |
There was a problem hiding this comment.
this might be a bug as this is not preserving NA (but unrelated / not a regression, so pls open a new issue)
There was a problem hiding this comment.
The test case is actually a bit misleading. This regression was just about a boolean categorical with missing values, not specifically with NA (also if you create the data with np.nan, you had the same issue)
|
@meeseeksdev backport 1.3.x |
|
thanks @phofl |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
… categories (pandas-dev#44356) (cherry picked from commit 9f54f70)
