Glum

------------------

**New feature:**

- Added Negative Binomial distribution by setting the ``'family'`` parameter of
:class:`~glum.GeneralizedLinearRegressor` and :class:`~glum.GeneralizedLinearRegressorCV`
to ``'negative.binomial'``.

------------------

**Bug fixes:**

- Fixed an issue with :meth:`~glum.ExponentialDispersionModel._score_matrix` which failed when called with a tabmat matrix input.

**Other changes**:

- Removes unused scikit-learn cython imports.

------------------

**Other changes**:

- :class:`~glum._link.LogitLink` has been made public.
- Apple Silicon wheels are now uploaded to PyPI.

------------------

**Bug fixes:**

- A data frame with dense and sparse columns was transformed to a dense matrix instead of a split matrix by :meth:`~glum.GeneralizedLinearRegressor._set_up_and_check_fit_args`.
Fixed by calling ``tabmat.from_pandas`` on any data frame.

**New features:**

- The following classes and functions have been made public:
:class:`~glum._distribution.BinomialDistribution`,
:class:`~glum._distribution.ExponentialDispersionModel`,
:class:`~glum._distribution.GammaDistribution`,
:class:`~glum._distribution.GeneralizedHyperbolicSecant`,
:class:`~glum._distribution.InverseGaussianDistribution`,
:class:`~glum._distribution.NormalDistribution`,
:class:`~glum._distribution.PoissonDistribution`,
:class:`~glum._link.IdentityLink`,
:class:`~glum._link.Link`,
:class:`~glum._link.LogLink`,
:class:`~glum._link.TweedieLink`,
:func:`~glum._glm.get_family` and
:func:`~glum._glm.get_link`.
- The distribution and link classes now feature a more lenient equality check instead of the default identity check,
so that, e.g., ``TweedieDistribution(1) == TweedieDistribution(1)`` now returns ``True``.

------------------

**Other changes:**

- Fixing pypi upload issue. Version 2.2.0 will not be available through the standard distribution channels.

------------------

**New features:**

- Add an argument to GeneralizedLinearRegressorBase to drop the first category in a Categorical column using [implementation in tabmat](https://github.com/Quantco/tabmat/pull/168)
- One may now request the Tweedie loss by setting the ``'family'`` parameter of
:class:`~glum.GeneralizedLinearRegressor` and :class:`~glum.GeneralizedLinearRegressorCV`
to ``'tweedie'``.

**Bug fixes:**

- Setting bounds for constant columns was not working (bounds were internally modified to 0).
A similar issue was preventing inequalities from working with constant columns. This is now fixed.

**Other changes:**

- No more builds for 32-bit systems with python >= 3.8. This is due to scipy not supporting it anymore.