- Bug in :meth:`Categorical.fillna` would replace all values, not just those that are
NaN(:issue:`26215`)
- Bug in :func:`to_datetime` where passing a timezone-naive :class:`DatetimeArray` or :class:`DatetimeIndex` and
utc=Truewould incorrectly return a timezone-naive result (:issue:`27733`)
- Bug in :class:`Index` where a numpy object array with a timezone aware :class:`Timestamp` and
np.nanwould not return a :class:`DatetimeIndex` (:issue:`27011`)
- Bug in :meth:`Series.interpolate` when using a timezone aware :class:`DatetimeIndex` (:issue:`27548`)
- Bug when printing negative floating point complex numbers would raise an
IndexError(:issue:`27484`)
- Improved the warnings for the deprecated methods :meth:`Series.real` and :meth:`Series.imag` (:issue:`27610`)
- Bug in :class:`IntervalIndex` where dir(obj) would raise
ValueError(:issue:`27571`)
- Bug in partial-string indexing returning a NumPy array rather than a
Serieswhen indexing with a scalar like.loc['2015'](:issue:`27516`) - Break reference cycle involving :class:`Index` and other index classes to allow garbage collection of index objects without running the GC. (:issue:`27585`, :issue:`27840`)
- Fix regression in assigning values to a single column of a DataFrame with a
MultiIndexcolumns (:issue:`27841`).
- Avoid calling
S3File.s3when reading parquet, as this was removed in s3fs version 0.3.0 (:issue:`27756`) - Better error message when a negative header is passed in :func:`pandas.read_csv` (:issue:`27779`)
- Added a pandas_plotting_backends entrypoint group for registering plot backends. See :ref:`extending.plotting-backends` for more (:issue:`26747`).
- Fixed the re-instatement of Matplotlib datetime converters after calling pandas.plotting.deregister_matplotlib_converters() (:issue:`27481`).
- Fix compatibility issue with matplotlib when passing a pandas
Indexto a plot call (:issue:`27775`).
- Bug in :meth:`pandas.core.groupby.DataFrameGroupBy.transform` where applying a timezone conversion lambda function would drop timezone information (:issue:`27496`)
- Bug in windowing over read-only arrays (:issue:`27766`)
- Fixed segfault in pandas.core.groupby.DataFrameGroupBy.quantile when an invalid quantile was passed (:issue:`27470`)
- A
KeyErroris now raised if.unstack()is called on a :class:`Series` or :class:`DataFrame` with a flat :class:`Index` passing a name which is not the correct one (:issue:`18303`) - Bug in :meth:`DataFrame.crosstab` when
marginsset toTrueandnormalizeis notFalse, an error is raised. (:issue:`27500`) - :meth:`DataFrame.join` now suppresses the
FutureWarningwhen the sort parameter is specified (:issue:`21952`) - Bug in :meth:`DataFrame.join` raising with readonly arrays (:issue:`27943`)
- Bug in reductions for :class:`Series` with Sparse dtypes (:issue:`27080`)
- Bug in :meth:`Series.replace` and :meth:`DataFrame.replace` when replacing timezone-aware timestamps using a dict-like replacer (:issue:`27720`)
Some users may unknowingly have an incomplete Python installation, which lacks the lzma module from the standard library. In this case, import pandas failed due to an ImportError (:issue: 27575). Pandas will now warn, rather than raising an ImportError if the lzma module is not present. Any subsequent attempt to use lzma methods will raise a RuntimeError. A possible fix for the lack of the lzma module is to ensure you have the necessary libraries and then re-install Python. For example, on MacOS installing Python with pyenv may lead to an incomplete Python installation due to unmet system dependencies at compilation time (like xz). Compilation will succeed, but Python might fail at run time. The issue can be solved by installing the necessary dependencies and then re-installing Python.
.. contributors:: v0.25.0..HEAD