Hi,

I'm pleased to announce the availability pandas 0.21.0.

This is a major release from 0.20.3 and includes a number of API changes, deprecations, new features, enhancements, and performance improvements along with a large number of bug fixes. We recommend that all users upgrade to this version.

Highlights include:

- Integration with Apache Parquet https://parquet.apache.org, including a new top-level read_parquet http://pandas.pydata.org/pandas-docs/version/0.21/generated/pandas.read_parquet.html#pandas.read_parquet function and a DataFrame.to_parquet http://pandas.pydata.org/pandas-docs/version/0.21/generated/pandas.DataFrame.to_parquet.html#pandas.DataFrame.to_parquet method, see here http://pandas.pydata.org/pandas-docs/version/0.21/io.html#io-parquet.

- New user-facing dtype pandas.api.types.CategoricalDtype http://pandas.pydata.org/pandas-docs/version/0.21/generated/pandas.api.types.CategoricalDtype.html#pandas.api.types.CategoricalDtype for specifying categoricals independent of the data, see here http://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#whatsnew-0210-enhancements-categorical-dtype .

- The behavior of sum and prod on all-NaN Series/DataFrames is now consistent and no longer depends on whether bottleneck https://kwgoodman.github.io/bottleneck-doc/ is installed, see here http://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#whatsnew-0210-api-breaking-bottleneck .

- Compatibility fixes for pypy, see here http://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#whatsnew-0210-pypy .

- Additions to the drop, reindex, and rename API to make them more consistent, see here http://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#whatsnew-0210-enhancements-drop-api

- Addition of the new methods DataFrame.infer_objects (see here http://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#whatsnew-0210-enhancements-infer-objects) and GroupBy.pipe (see here http://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#whatsnew-0210-enhancements-groupby-pipe ).

- Indexing with a list of labels, where one or more of the labels is missing, is deprecated and will raise a KeyError in a future version, see here http://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#whatsnew-0210-api-breaking-loc .

Check the whatsnew https://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html for detailed changes, including backwards incompatible changes http://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#backwards-incompatible-api-changes and deprecations http://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#whatsnew-0190-deprecations. Please report any issues you find on the pandas issue tracker https://github.com/pandas-dev/pandas.

Binary packages will be available in the defaults and conda-forge channels shortly.

conda install pandas

Wheels and a source distribution are available on PyPI.

pip install --upgrade pip setuptools pip install --upgrade --upgrade-strategy=only-if-needed pandas

Tom