What’s new in 2.1.2 (October 26, 2023)

These are the changes in pandas 2.1.2. See Release notes for a full changelog including other versions of pandas.

Deprecations

• Reverted deprecation of fill_method=None in DataFrame.pct_change(), Series.pct_change(), DataFrameGroupBy.pct_change(), and SeriesGroupBy.pct_change(); the values 'backfill', 'bfill', 'pad', and 'ffill' are still deprecated (GH 53491)

Fixed regressions

• Fixed regression in DataFrame.join() where result has missing values and dtype is arrow backed string (GH 55348)

• Fixed regression in rolling() where non-nanosecond index or on column would produce incorrect results (GH 55026, GH 55106, GH 55299)

• Fixed regression in DataFrame.resample() which was extrapolating back to origin when origin was outside its bounds (GH 55064)

• Fixed regression in DataFrame.sort_index() which was not sorting correctly when the index was a sliced MultiIndex (GH 55379)

• Fixed regression in DataFrameGroupBy.agg() and SeriesGroupBy.agg() where if the option compute.use_numba was set to True, groupby methods not supported by the numba engine would raise a TypeError (GH 55520)

• Fixed performance regression with wide DataFrames, typically involving methods where all columns were accessed individually (GH 55256, GH 55245)

• Fixed regression in merge_asof() raising TypeError for by with datetime and timedelta dtypes (GH 55453)

• Fixed regression in read_parquet() when reading a file with a string column consisting of more than 2 GB of string data and using the "string" dtype (GH 55606)

• Fixed regression in DataFrame.to_sql() not roundtripping datetime columns correctly for sqlite when using detect_types (GH 55554)

• Fixed regression in construction of certain DataFrame or Series subclasses (GH 54922)

Bug fixes

• Fixed bug in DataFrameGroupBy reductions not preserving object dtype when infer_string is set (GH 55620)

• Fixed bug in SeriesGroupBy.value_counts() returning incorrect dtype for string columns (GH 55627)

• Fixed bug in Categorical.equals() if other has arrow backed string dtype (GH 55364)

• Fixed bug in DataFrame.__setitem__() not inferring string dtype for zero-dimensional array with infer_string=True (GH 55366)

• Fixed bug in DataFrame.idxmin() and DataFrame.idxmax() raising for arrow dtypes (GH 55368)

• Fixed bug in DataFrame.interpolate() raising incorrect error message (GH 55347)

• Fixed bug in Index.insert() raising when inserting None into Index with dtype="string[pyarrow_numpy]" (GH 55365)

• Fixed bug in Series.all() and Series.any() not treating missing values correctly for dtype="string[pyarrow_numpy]" (GH 55367)

• Fixed bug in Series.floordiv() for ArrowDtype (GH 55561)

• Fixed bug in Series.mode() not sorting values for arrow backed string dtype (GH 55621)

• Fixed bug in Series.rank() for string[pyarrow_numpy] dtype (GH 55362)

• Fixed bug in Series.str.extractall() for ArrowDtype dtype being converted to object (GH 53846)

• Fixed bug where PDEP-6 warning about setting an item of an incompatible dtype was being shown when creating a new conditional column (GH 55025)

• Silence Period[B] warnings introduced by GH 53446 during normal plotting activity (GH 55138)

• Fixed bug in Series constructor not inferring string dtype when NA is the first value and infer_string is set (:issue:` 55655`)

Other

• Fixed non-working installation of optional dependency group output_formatting. Replacing underscore _ with a dash - fixes broken dependency resolution. A correct way to use now is pip install pandas[output-formatting].

Contributors

For contributors, please see /usr/share/doc/contributors_list.txt or https://github.com/pandas-dev/pandas/graphs/contributors