Skrub

==========================

Major changes
-------------

* New encoder: :class:`DatetimeEncoder` can transform a datetime column into several numerical columns
(year, month, day, hour, minute, second, ...). It is now the default transformer used
in the :class:`TableVectorizer` for datetime columns. :pr:`239` by :user:`Leo Grinsztajn <LeoGrin>`

* The :class:`TableVectorizer` has seen some major improvements and bug fixes:

- Fixes the automatic casting logic in ``transform``.
- To avoid dimensionality explosion when a feature has two unique values, the default encoder (:class:`~sklearn.preprocessing.OneHotEncoder`) now drops one of the two vectors (see parameter `drop="if_binary"`).
- ``fit_transform`` and ``transform`` can now return unencoded features, like the :class:`~sklearn.compose.ColumnTransformer`'s behavior. Previously, a ``RuntimeError`` was raised.

:pr:`300` by :user:`Lilian Boulard <LilianBoulard>`

* **Backward-incompatible change in the TableVectorizer**:
To apply ``remainder`` to features (with the ``*_transformer`` parameters),
the value ``'remainder'`` must be passed, instead of ``None`` in previous versions.
``None`` now indicates that we want to use the default transformer. :pr:`303` by :user:`Lilian Boulard <LilianBoulard>`

* Support for Python 3.6 and 3.7 has been dropped. Python >= 3.8 is now required. :pr:`289` by :user:`Lilian Boulard <LilianBoulard>`

* Bumped minimum dependencies:

- scikit-learn>=0.23
- scipy>=1.4.0
- numpy>=1.17.3
- pandas>=1.2.0 :pr:`299` and :pr:`300` by :user:`Lilian Boulard <LilianBoulard>`

* Dropped support for Jaro, Jaro-Winkler and Levenshtein distances.

- The :class:`SimilarityEncoder` now exclusively uses ``ngram`` for similarities,
and the `similarity` parameter is deprecated. It will be removed in 0.5. :pr:`282` by :user:`Lilian Boulard <LilianBoulard>`

Notes
-----

* The ``transformers_`` attribute of the :class:`TableVectorizer` now contains column
names instead of column indices for the "remainder" columns. :pr:`266` by :user:`Leo Grinsztajn <LeoGrin>`

=========================

Bug fixes
---------

* Fixed a bug in the :class:`TableVectorizer` causing a :class:`FutureWarning`
when using the :func:`get_feature_names_out` method. :pr:`262` by :user:`Lilian Boulard <LilianBoulard>`

==========================

Major changes
-------------

* Improvements to the :class:`TableVectorizer`

- Type detection works better: handles dates, numerics columns encoded as strings, or numeric columns containing strings for missing values.

:pr:`238` by :user:`Leo Grinsztajn <LeoGrin>`

* :func:`get_feature_names` becomes :func:`get_feature_names_out`, following changes in the scikit-learn API.
:func:`get_feature_names` is deprecated in scikit-learn > 1.0. :pr:`241` by :user:`Gael Varoquaux <GaelVaroquaux>`

* Improvements to the :class:`MinHashEncoder`
- It is now possible to fit multiple columns simultaneously with the :class:`MinHashEncoder`.
Very useful when using for instance the :func:`~sklearn.compose.make_column_transformer` function,
on multiple columns.

:pr:`243` by :user:`Jovan Stojanovic <jovan-stojanovic>`


Bug-fixes
---------

* Fixed a bug that resulted in the :class:`GapEncoder` ignoring the analyzer argument. :pr:`242` by :user:`Jovan Stojanovic <jovan-stojanovic>`

* :class:`GapEncoder`'s `get_feature_names_out` now accepts all iterators, not just lists. :pr:`255` by :user:`Lilian Boulard <LilianBoulard>`

* Fixed :class:`DeprecationWarning` raised by the usage of `distutils.version.LooseVersion`. :pr:`261` by :user:`Lilian Boulard <LilianBoulard>`

Notes
-----

* Remove trailing imports in the :class:`MinHashEncoder`.

* Fix typos and update links for website.

* Documentation of the :class:`TableVectorizer` and the :class:`SimilarityEncoder` improved.

=========================

Also see pre-release 0.2.0a1 below for additional changes.

Major changes
-------------

* Bump minimum dependencies:

- scikit-learn (>=0.21.0) :pr:`202` by :user:`Lilian Boulard <LilianBoulard>`
- pandas (>=1.1.5) **! NEW REQUIREMENT !** :pr:`155` by :user:`Lilian Boulard <LilianBoulard>`

* **datasets.fetching** - backward-incompatible changes to the example
datasets fetchers:

- The backend has changed: we now exclusively fetch the datasets from OpenML.
End users should not see any difference regarding this.
- The frontend, however, changed a little: the fetching functions stay the same
but their return values were modified in favor of a more Pythonic interface.
Refer to the docstrings of functions `dirty_cat.datasets.fetch_*`
for more information.
- The example notebooks were updated to reflect these changes. :pr:`155` by :user:`Lilian Boulard <LilianBoulard>`

* **Backward incompatible change to** :class:`MinHashEncoder`: The :class:`MinHashEncoder` now
only supports two dimensional inputs of shape (N_samples, 1).
:pr:`185` by :user:`Lilian Boulard <LilianBoulard>` and :user:`Alexis Cvetkov <alexis-cvetkov>`.

* Update `handle_missing` parameters:

- :class:`GapEncoder`: the default value "zero_impute" becomes "empty_impute" (see doc).
- :class:`MinHashEncoder`: the default value "" becomes "zero_impute" (see doc).

:pr:`210` by :user:`Alexis Cvetkov <alexis-cvetkov>`.

* Add a method "get_feature_names_out" for the :class:`GapEncoder` and the :class:`TableVectorizer`,
since `get_feature_names` will be depreciated in scikit-learn 1.2. :pr:`216` by :user:`Alexis Cvetkov <alexis-cvetkov>`

Notes
-----

* Removed hard-coded CSV file `dirty_cat/data/FiveThirtyEight_Midwest_Survey.csv`.


* Improvements to the :class:`TableVectorizer`

- Missing values are not systematically imputed anymore
- Type casting and per-column imputation are now learnt during fitting
- Several bugfixes

:pr:`201` by :user:`Lilian Boulard <LilianBoulard>`

============================

Version 0.2.0a1 is a pre-release.
To try it, you have to install it manually using::

pip install --pre dirty_cat==0.2.0a1

or from the GitHub repository::

pip install git+https://github.com/dirty-cat/dirty_cat.git

Major changes
-------------

* Bump minimum dependencies:

- Python (>= 3.6)
- NumPy (>= 1.16)
- SciPy (>= 1.2)
- scikit-learn (>= 0.20.0)

* :class:`TableVectorizer`: Added automatic transform through the
:class:`TableVectorizer` class. It transforms
columns automatically based on their type. It provides a replacement
for scikit-learn's :class:`~sklearn.compose.ColumnTransformer` simpler to use on heterogeneous
pandas DataFrame. :pr:`167` by :user:`Lilian Boulard <LilianBoulard>`

* **Backward incompatible change to** :class:`GapEncoder`: The :class:`GapEncoder` now only
supports two-dimensional inputs of shape (n_samples, n_features).
Internally, features are encoded by independent :class:`GapEncoder` models,
and are then concatenated into a single matrix.
:pr:`185` by :user:`Lilian Boulard <LilianBoulard>` and :user:`Alexis Cvetkov <alexis-cvetkov>`.


Bug-fixes
---------

* Fix `get_feature_names` for scikit-learn > 0.21. :pr:`216` by :user:`Alexis Cvetkov <alexis-cvetkov>`

========================

Major changes
-------------

Bug-fixes
---------

* RuntimeWarnings due to overflow in :class:`GapEncoder`. :pr:`161` by :user:`Alexis Cvetkov <alexis-cvetkov>`