scikit-learn · lorentzenchr · Apr 2, 2021 · Mar 31, 2021 · Mar 31, 2021 · Mar 31, 2021
diff --git a/asv_benchmarks/benchmarks/ensemble.py b/asv_benchmarks/benchmarks/ensemble.py
@@ -1,4 +1,3 @@
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import (RandomForestClassifier,
                              GradientBoostingClassifier,
                              HistGradientBoostingClassifier)

diff --git a/benchmarks/bench_hist_gradient_boosting.py b/benchmarks/bench_hist_gradient_boosting.py
@@ -4,8 +4,6 @@
 import matplotlib.pyplot as plt
 import numpy as np
 from sklearn.model_selection import train_test_split
-# To use this experimental feature, we need to explicitly ask for it:
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor
 from sklearn.ensemble import HistGradientBoostingClassifier
 from sklearn.datasets import make_classification

diff --git a/benchmarks/bench_hist_gradient_boosting_adult.py b/benchmarks/bench_hist_gradient_boosting_adult.py
@@ -4,7 +4,6 @@
 from sklearn.model_selection import train_test_split
 from sklearn.datasets import fetch_openml
 from sklearn.metrics import accuracy_score, roc_auc_score
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingClassifier
 from sklearn.ensemble._hist_gradient_boosting.utils import (
    get_equivalent_estimator)

diff --git a/benchmarks/bench_hist_gradient_boosting_categorical_only.py b/benchmarks/bench_hist_gradient_boosting_categorical_only.py
@@ -3,7 +3,6 @@

 from sklearn.preprocessing import KBinsDiscretizer
 from sklearn.datasets import make_classification
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingClassifier
 from sklearn.ensemble._hist_gradient_boosting.utils import (
    get_equivalent_estimator)

diff --git a/benchmarks/bench_hist_gradient_boosting_higgsboson.py b/benchmarks/bench_hist_gradient_boosting_higgsboson.py
@@ -9,8 +9,6 @@
 from joblib import Memory
 from sklearn.model_selection import train_test_split
 from sklearn.metrics import accuracy_score, roc_auc_score
-# To use this experimental feature, we need to explicitly ask for it:
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingClassifier
 from sklearn.ensemble._hist_gradient_boosting.utils import (
    get_equivalent_estimator)

diff --git a/benchmarks/bench_hist_gradient_boosting_threading.py b/benchmarks/bench_hist_gradient_boosting_threading.py
@@ -7,8 +7,6 @@
 from threadpoolctl import threadpool_limits
 import sklearn
 from sklearn.model_selection import train_test_split
-# To use this experimental feature, we need to explicitly ask for it:
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor
 from sklearn.ensemble import HistGradientBoostingClassifier
 from sklearn.datasets import make_classification

diff --git a/doc/conf.py b/doc/conf.py
@@ -358,7 +358,6 @@ def __call__(self, directory):

 # enable experimental module so that experimental estimators can be
 # discovered properly by sphinx
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.experimental import enable_iterative_imputer  # noqa
 from sklearn.experimental import enable_halving_search_cv  # noqa


diff --git a/doc/developers/maintainer.rst b/doc/developers/maintainer.rst
@@ -363,10 +363,17 @@ deprecation cycle.

 To create an experimental module, you can just copy and modify the content of
 `enable_hist_gradient_boosting.py
-<https://github.com/scikit-learn/scikit-learn/blob/main/sklearn/experimental/enable_hist_gradient_boosting.py>`_,
+<https://github.com/scikit-learn/scikit-learn/blob/c9c89cfc85dd8dfefd7921c16c87327d03140a06/sklearn/experimental/enable_hist_gradient_boosting.py>`__,
 or
 `enable_iterative_imputer.py
-<https://github.com/scikit-learn/scikit-learn/blob/main/sklearn/experimental/enable_iterative_imputer.py>`_.
+<https://github.com/scikit-learn/scikit-learn/blob/c9c89cfc85dd8dfefd7921c16c87327d03140a06/sklearn/experimental/enable_iterative_imputer.py>`_.
+
+.. note::
+
+  These are permalink as in 0.24, where these estimators are still
+  experimental. They might be stable at the time of reading - hence the
+  permalink. See below for instructions on the transition from experimental
+  to stable.

 Note that the public import path must be to a public subpackage (like
 ``sklearn/ensemble`` or ``sklearn/impute``), not just a ``.py`` module.
@@ -379,14 +386,15 @@ in the future when the features aren't experimental anymore.
 To avoid type checker (e.g. mypy) errors a direct import of experimental
 estimators should be done in the parent module, protected by the
 ``if typing.TYPE_CHECKING`` check. See `sklearn/ensemble/__init__.py
-<https://github.com/scikit-learn/scikit-learn/blob/main/sklearn/ensemble/__init__.py>`_,
+<https://github.com/scikit-learn/scikit-learn/blob/c9c89cfc85dd8dfefd7921c16c87327d03140a06/sklearn/ensemble/__init__.py>`_,
 or `sklearn/impute/__init__.py
-<https://github.com/scikit-learn/scikit-learn/blob/main/sklearn/impute/__init__.py>`_
+<https://github.com/scikit-learn/scikit-learn/blob/c9c89cfc85dd8dfefd7921c16c87327d03140a06/sklearn/impute/__init__.py>`_
 for an example.

 Please also write basic tests following those in
 `test_enable_hist_gradient_boosting.py
-<https://github.com/scikit-learn/scikit-learn/blob/main/sklearn/experimental/tests/test_enable_hist_gradient_boosting.py>`_.
+<https://github.com/scikit-learn/scikit-learn/blob/c9c89cfc85dd8dfefd7921c16c87327d03140a06/sklearn/experimental/tests/test_enable_hist_gradient_boosting.py>`__.
+

 Make sure every user-facing code you write explicitly mentions that the feature
 is experimental, and add a ``# noqa`` comment to avoid pep8-related warnings::
@@ -402,3 +410,14 @@ sklearn.experimental import *`` **does not work**.

 Note that some experimental classes / functions are not included in the
 :mod:`sklearn.experimental` module: ``sklearn.datasets.fetch_openml``.
+
+Once the feature become stable, remove all `enable_my_experimental_feature`
+in the scikit-learn code (even feature highlights etc.) and make the
+`enable_my_experimental_feature` a no-op that just raises a warning:
+`enable_hist_gradient_boosting.py
+<https://github.com/scikit-learn/scikit-learn/blob/main/sklearn/experimental/enable_hist_gradient_boosting.py>`__.
+The file should stay there indefinitely as we don't want to break users code:
+we just incentivize them to remove that import with the warning.
+
+Also update the tests accordingly: `test_enable_hist_gradient_boosting.py
+<https://github.com/scikit-learn/scikit-learn/blob/main/sklearn/experimental/tests/test_enable_hist_gradient_boosting.py>`__.
diff --git a/doc/modules/ensemble.rst b/doc/modules/ensemble.rst
@@ -467,7 +467,7 @@ trees.

 .. note::

-  Scikit-learn 0.21 introduces two new experimental implementations of
+  Scikit-learn 0.21 introduces two new implementations of
  gradient boosting trees, namely :class:`HistGradientBoostingClassifier`
  and :class:`HistGradientBoostingRegressor`, inspired by
  `LightGBM <https://github.com/Microsoft/LightGBM>`__ (See [LightGBM]_).
@@ -898,7 +898,7 @@ based on permutation of the features.
 Histogram-Based Gradient Boosting
 =================================

-Scikit-learn 0.21 introduced two new experimental implementations of
+Scikit-learn 0.21 introduced two new implementations of
 gradient boosting trees, namely :class:`HistGradientBoostingClassifier`
 and :class:`HistGradientBoostingRegressor`, inspired by
 `LightGBM <https://github.com/Microsoft/LightGBM>`__ (See [LightGBM]_).
@@ -920,15 +920,6 @@ estimators is slightly different, and some of the features from
 :class:`GradientBoostingClassifier` and :class:`GradientBoostingRegressor`
 are not yet supported, for instance some loss functions.

-These estimators are still **experimental**: their predictions
-and their API might change without any deprecation cycle. To use them, you
-need to explicitly import ``enable_hist_gradient_boosting``::
-
-  >>> # explicitly require this experimental feature
-  >>> from sklearn.experimental import enable_hist_gradient_boosting  # noqa
-  >>> # now you can import normally from ensemble
-  >>> from sklearn.ensemble import HistGradientBoostingClassifier
-
 .. topic:: Examples:

 * :ref:`sphx_glr_auto_examples_inspection_plot_partial_dependence.py`
@@ -941,7 +932,6 @@ Most of the parameters are unchanged from
 One exception is the ``max_iter`` parameter that replaces ``n_estimators``, and
 controls the number of iterations of the boosting process::

-  >>> from sklearn.experimental import enable_hist_gradient_boosting
  >>> from sklearn.ensemble import HistGradientBoostingClassifier
  >>> from sklearn.datasets import make_hastie_10_2

@@ -992,7 +982,6 @@ with missing values should go to the left or right child, based on the
 potential gain. When predicting, samples with missing values are assigned to
 the left or right child consequently::

-  >>> from sklearn.experimental import enable_hist_gradient_boosting  # noqa
  >>> from sklearn.ensemble import HistGradientBoostingClassifier
  >>> import numpy as np

@@ -1146,7 +1135,6 @@ You can specify a monotonic constraint on each feature using the
 constraint, while -1 and 1 indicate a negative and positive constraint,
 respectively::

-  >>> from sklearn.experimental import enable_hist_gradient_boosting  # noqa
  >>> from sklearn.ensemble import HistGradientBoostingRegressor

  ... # positive, negative, and no constraint on the 3 features

diff --git a/doc/whats_new/v0.21.rst b/doc/whats_new/v0.21.rst
@@ -420,6 +420,11 @@ Support for Python 3.4 and below has been officially dropped.
    >>> from sklearn.experimental import enable_hist_gradient_boosting  # noqa
    >>> # now you can import normally from sklearn.ensemble
    >>> from sklearn.ensemble import HistGradientBoostingClassifier
+
+  .. note::
+      Update: since version 1.0, these estimators are not experimental
+      anymore and you don't need to use `from sklearn.experimental import
+      enable_hist_gradient_boosting`.

  :pr:`12807` by :user:`Nicolas Hug<NicolasHug>`.


diff --git a/doc/whats_new/v1.0.rst b/doc/whats_new/v1.0.rst
@@ -130,6 +130,11 @@ Changelog
  target. Additional private refactoring was performed.
  :pr:`19162` by :user:`Guillaume Lemaitre <glemaitre>`.

+- |Enhancement| :class:`~sklearn.ensemble.HistGradientBoostingClassifier` and
+  :class:`~sklearn.ensemble.HistGradientBoostingRegressor` are no longer
+  experimental. They are now considered stable and are subject to the same
+  deprecation cycles as all other estimators. :pr:`19799` by `Nicolas Hug`_.
+
 :mod:`sklearn.feature_extraction`
 .................................


diff --git a/examples/ensemble/plot_gradient_boosting_categorical.py b/examples/ensemble/plot_gradient_boosting_categorical.py
@@ -45,7 +45,6 @@
 # As a baseline, we create an estimator where the categorical features are
 # dropped:

-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor
 from sklearn.pipeline import make_pipeline
 from sklearn.compose import make_column_transformer

diff --git a/examples/ensemble/plot_monotonic_constraints.py b/examples/ensemble/plot_monotonic_constraints.py
@@ -18,7 +18,6 @@
 This example was inspired by the `XGBoost documentation
 <https://xgboost.readthedocs.io/en/latest/tutorials/monotonic.html>`_.
 """
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor
 from sklearn.inspection import plot_partial_dependence
 import numpy as np

diff --git a/examples/ensemble/plot_stack_predictors.py b/examples/ensemble/plot_stack_predictors.py
@@ -160,7 +160,6 @@ def load_ames_housing():
 rf_pipeline

 # %%
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor

 gbdt_pipeline = make_pipeline(

diff --git a/examples/inspection/plot_partial_dependence.py b/examples/inspection/plot_partial_dependence.py
@@ -134,7 +134,6 @@
 # Let's now fit a :class:`~sklearn.ensemble.HistGradientBoostingRegressor` and
 # compute the partial dependence on the same features.

-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor

 print("Training HistGradientBoostingRegressor...")

diff --git a/examples/linear_model/plot_poisson_regression_non_normal_loss.py b/examples/linear_model/plot_poisson_regression_non_normal_loss.py
@@ -258,7 +258,6 @@ def score_estimator(estimator, df_test):
 # least-squares loss. Here we only fit trees with the Poisson loss to keep this
 # example concise.

-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor
 from sklearn.preprocessing import OrdinalEncoder


diff --git a/examples/release_highlights/plot_release_highlights_0_22_0.py b/examples/release_highlights/plot_release_highlights_0_22_0.py
@@ -131,7 +131,6 @@
 # support for missing values (NaNs). This means that there is no need for
 # imputing data when training or predicting.

-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingClassifier

 X = np.array([0, 1, 2, np.nan]).reshape(-1, 1)

diff --git a/examples/release_highlights/plot_release_highlights_0_23_0.py b/examples/release_highlights/plot_release_highlights_0_23_0.py
@@ -36,7 +36,6 @@
 import numpy as np
 from sklearn.model_selection import train_test_split
 from sklearn.linear_model import PoissonRegressor
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor

 n_samples, n_features = 1000, 20
@@ -124,7 +123,6 @@
 from matplotlib import pyplot as plt
 from sklearn.model_selection import train_test_split
 from sklearn.inspection import plot_partial_dependence
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor

 n_samples = 500

diff --git a/sklearn/ensemble/__init__.py b/sklearn/ensemble/__init__.py
@@ -2,8 +2,6 @@
 The :mod:`sklearn.ensemble` module includes ensemble-based methods for
 classification, regression and anomaly detection.
 """
-import typing
-
 from ._base import BaseEnsemble
 from ._forest import RandomForestClassifier
 from ._forest import RandomForestRegressor
@@ -21,13 +19,9 @@
 from ._voting import VotingRegressor
 from ._stacking import StackingClassifier
 from ._stacking import StackingRegressor
-
-if typing.TYPE_CHECKING:
-    # Avoid errors in type checkers (e.g. mypy) for experimental estimators.
-    # TODO: remove this check once the estimator is no longer experimental.
-    from ._hist_gradient_boosting.gradient_boosting import (  # noqa
-        HistGradientBoostingRegressor, HistGradientBoostingClassifier
-    )
+from ._hist_gradient_boosting.gradient_boosting import (
+    HistGradientBoostingRegressor, HistGradientBoostingClassifier
+)

 __all__ = ["BaseEnsemble",
           "RandomForestClassifier", "RandomForestRegressor",
@@ -37,4 +31,5 @@
           "GradientBoostingRegressor", "AdaBoostClassifier",
           "AdaBoostRegressor", "VotingClassifier", "VotingRegressor",
           "StackingClassifier", "StackingRegressor",
+           'HistGradientBoostingClassifier', 'HistGradientBoostingRegressor',
           ]
diff --git a/sklearn/ensemble/_hist_gradient_boosting/gradient_boosting.py b/sklearn/ensemble/_hist_gradient_boosting/gradient_boosting.py
@@ -887,17 +887,6 @@ class HistGradientBoostingRegressor(RegressorMixin, BaseHistGradientBoosting):
    This implementation is inspired by
    `LightGBM <https://github.com/Microsoft/LightGBM>`_.

-    .. note::
-
-      This estimator is still **experimental** for now: the predictions
-      and the API might change without any deprecation cycle. To use it,
-      you need to explicitly import ``enable_hist_gradient_boosting``::
-
-        >>> # explicitly require this experimental feature
-        >>> from sklearn.experimental import enable_hist_gradient_boosting  # noqa
-        >>> # now you can import normally from ensemble
-        >>> from sklearn.ensemble import HistGradientBoostingRegressor
-
    Read more in the :ref:`User Guide <histogram_based_gradient_boosting>`.

    .. versionadded:: 0.21
@@ -1040,8 +1029,6 @@ class HistGradientBoostingRegressor(RegressorMixin, BaseHistGradientBoosting):

    Examples
    --------
-    >>> # To use this experimental feature, we need to explicitly ask for it:
-    >>> from sklearn.experimental import enable_hist_gradient_boosting  # noqa
    >>> from sklearn.ensemble import HistGradientBoostingRegressor
    >>> from sklearn.datasets import load_diabetes
    >>> X, y = load_diabetes(return_X_y=True)
@@ -1156,17 +1143,6 @@ class HistGradientBoostingClassifier(ClassifierMixin,
    This implementation is inspired by
    `LightGBM <https://github.com/Microsoft/LightGBM>`_.

-    .. note::
-
-      This estimator is still **experimental** for now: the predictions
-      and the API might change without any deprecation cycle. To use it,
-      you need to explicitly import ``enable_hist_gradient_boosting``::
-
-        >>> # explicitly require this experimental feature
-        >>> from sklearn.experimental import enable_hist_gradient_boosting  # noqa
-        >>> # now you can import normally from ensemble
-        >>> from sklearn.ensemble import HistGradientBoostingClassifier
-
    Read more in the :ref:`User Guide <histogram_based_gradient_boosting>`.

    .. versionadded:: 0.21
@@ -1304,8 +1280,6 @@ class HistGradientBoostingClassifier(ClassifierMixin,

    Examples
    --------
-    >>> # To use this experimental feature, we need to explicitly ask for it:
-    >>> from sklearn.experimental import enable_hist_gradient_boosting  # noqa
    >>> from sklearn.ensemble import HistGradientBoostingClassifier
    >>> from sklearn.datasets import load_iris
    >>> X, y = load_iris(return_X_y=True)

diff --git a/sklearn/ensemble/_hist_gradient_boosting/tests/test_compare_lightgbm.py b/sklearn/ensemble/_hist_gradient_boosting/tests/test_compare_lightgbm.py
@@ -4,8 +4,6 @@
 import numpy as np
 import pytest

-# To use this experimental feature, we need to explicitly ask for it:
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor
 from sklearn.ensemble import HistGradientBoostingClassifier
 from sklearn.ensemble._hist_gradient_boosting.binning import _BinMapper

diff --git a/sklearn/ensemble/_hist_gradient_boosting/tests/test_gradient_boosting.py b/sklearn/ensemble/_hist_gradient_boosting/tests/test_gradient_boosting.py
@@ -13,8 +13,6 @@
 from sklearn.exceptions import NotFittedError
 from sklearn.compose import make_column_transformer

-# To use this experimental feature, we need to explicitly ask for it:
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor
 from sklearn.ensemble import HistGradientBoostingClassifier
 from sklearn.ensemble._hist_gradient_boosting.loss import _LOSSES

diff --git a/sklearn/ensemble/_hist_gradient_boosting/tests/test_monotonic_contraints.py b/sklearn/ensemble/_hist_gradient_boosting/tests/test_monotonic_contraints.py
@@ -10,7 +10,6 @@
    compute_node_value
 )
 from sklearn.ensemble._hist_gradient_boosting.histogram import HistogramBuilder
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor
 from sklearn.ensemble import HistGradientBoostingClassifier


diff --git a/sklearn/ensemble/_hist_gradient_boosting/tests/test_warm_start.py b/sklearn/ensemble/_hist_gradient_boosting/tests/test_warm_start.py
@@ -7,8 +7,6 @@
 from sklearn.base import clone
 from sklearn.datasets import make_classification, make_regression

-# To use this experimental feature, we need to explicitly ask for it:
-from sklearn.experimental import enable_hist_gradient_boosting  # noqa
 from sklearn.ensemble import HistGradientBoostingRegressor
 from sklearn.ensemble import HistGradientBoostingClassifier
 from sklearn.metrics import check_scoring