MNT Replace DeprecationWarning with FutureWarning (#15080)

* bruteforce renaming

* WIP

* WIP

* some more

* removed weird line

* update -Werror

* testiforest

* again

* Fixed some tests

* fixed some tests

* removed -Werror

* fixed test_docstring_param issue

* fixed some tests

* some more

* renamed to SklearnDeprecationWarning

* pep8

* updated docs

* pep8

* merge

* changed to FutureWarning

* fixes

* Update doc/developers/tips.rst

Co-Authored-By: Adrin Jalali <adrin.jalali@gmail.com>

* avoid duplicates

* fixed warning for deprecations

* Still make CI break if DeprecationWarning isn't caught

* updated one warning

* fixed test

* fixed some renamings

* updated new dep warnings

* fixed bad import

* ignore warnings

* update again

* ignore futurewarning when walking packages

* Fixed test

* pep8

* Added whatsnew

doc/developers/contributing.rst

@@ -826,15 +826,17 @@ E.g., renaming an attribute ``labels_`` to ``classes_`` can be done as::
     def labels_(self):
         return self.classes_
 
-If a parameter has to be deprecated, use ``DeprecationWarning`` appropriately.
+If a parameter has to be deprecated, a ``FutureWarning`` warning
+must be raised too.
 In the following example, k is deprecated and renamed to n_clusters::
 
     import warnings
 
     def example_function(n_clusters=8, k='deprecated'):
         if k != 'deprecated':
             warnings.warn("'k' was renamed to n_clusters in version 0.13 and "
-                          "will be removed in 0.15.", DeprecationWarning)
+                          "will be removed in 0.15.",
+                          FutureWarning)
             n_clusters = k
 
 When the change is in a class, we validate and raise warning in ``fit``::
@@ -849,7 +851,8 @@ When the change is in a class, we validate and raise warning in ``fit``::
       def fit(self, X, y):
           if self.k != 'deprecated':
               warnings.warn("'k' was renamed to n_clusters in version 0.13 and "
-                            "will be removed in 0.15.", DeprecationWarning)
+                            "will be removed in 0.15.",
+                            FutureWarning)
               self._n_clusters = self.k
           else:
               self._n_clusters = self.n_clusters

doc/developers/tips.rst

@@ -102,10 +102,10 @@ Other `pytest` options that may become useful include:
     statements
   - ``--tb=short`` or ``--tb=line`` to control the length of the logs
 
-Since our continuous integration tests will error if ``DeprecationWarning``
-or ``FutureWarning`` aren't properly caught, it is also recommended to run
-``pytest`` along with the ``-Werror::DeprecationWarning`` and
-``-Werror::FutureWarning`` flags.
+Since our continuous integration tests will error if
+``FutureWarning`` isn't properly caught,
+it is also recommended to run ``pytest`` along with the
+``-Werror::FutureWarning`` flag.
 
 .. _saved_replies:
 

doc/whats_new/v0.22.rst

@@ -13,6 +13,19 @@ For a short description of the main highlights of the release, please
 refer to
 :ref:`sphx_glr_auto_examples_release_highlights_plot_release_highlights_0_22_0.py`.
 
+
+Deprecations: using ``FutureWarning`` from now on
+-------------------------------------------------
+
+When deprecating a feature, previous versions of scikit-learn used to raise
+a ``DeprecationWarning``. Since the ``DeprecationWarnings`` aren't shown by
+default by Python, scikit-learn needed to resort to a custom warning filter
+that would always show the warnings.
+
+This filter is now removed, and starting from 0.22 scikit-learn will show
+``FutureWarnings`` for deprecations. :pr:`15080` by `Nicolas Hug`_.
+
+
 Changed models
 --------------
 

sklearn/__init__.py

@@ -14,7 +14,6 @@
 """
 import sys
 import re
-import warnings
 import logging
 import os
 
@@ -25,10 +24,6 @@
 logger.setLevel(logging.INFO)
 
 
-# Make sure that DeprecationWarning within this package always gets printed
-warnings.filterwarnings('always', category=DeprecationWarning,
-                        module=r'^{0}\.'.format(re.escape(__name__)))
-
 # PEP0440 compatible formatted version, see:
 # https://www.python.org/dev/peps/pep-0440/
 #

sklearn/cluster/tests/test_hierarchical.py

@@ -734,6 +734,6 @@ def test_n_components_deprecation():
 
     match = ("``n_components_`` attribute was deprecated "
              "in favor of ``n_connected_components_``")
-    with pytest.warns(DeprecationWarning, match=match):
+    with pytest.warns(FutureWarning, match=match):
         n = agc.n_components_
     assert n == agc.n_connected_components_

sklearn/compose/_column_transformer.py

@@ -422,7 +422,7 @@ def _validate_features(self, n_features, feature_names):
             warnings.warn("Given feature/column names or counts do not match "
                           "the ones for the data given during fit. This will "
                           "fail from v0.24.",
-                          DeprecationWarning)
+                          FutureWarning)
 
     def _log_message(self, name, idx, total):
         if not self.verbose:

sklearn/compose/tests/test_column_transformer.py

@@ -500,7 +500,7 @@ def test_column_transformer_invalid_columns(remainder):
     X_array_more = np.array([[0, 1, 2], [2, 4, 6], [3, 6, 9]]).T
     msg = ("Given feature/column names or counts do not match the ones for "
            "the data given during fit.")
-    with pytest.warns(DeprecationWarning, match=msg):
+    with pytest.warns(FutureWarning, match=msg):
         ct.transform(X_array_more)  # Should accept added columns, for now
     X_array_fewer = np.array([[0, 1, 2], ]).T
     err_msg = 'Number of features'
@@ -1108,7 +1108,7 @@ def test_column_transformer_reordered_column_names_remainder(explicit_colname):
     # No error for added columns if ordering is identical
     X_extended_df = X_fit_df.copy()
     X_extended_df['third'] = [3, 6, 9]
-    with pytest.warns(DeprecationWarning, match=warn_msg):
+    with pytest.warns(FutureWarning, match=warn_msg):
         tf.transform(X_extended_df)  # No error should be raised, for now
 
     # No 'columns' AttributeError when transform input is a numpy array
@@ -1133,13 +1133,13 @@ def test_feature_name_validation():
 
     msg = ("Given feature/column names or counts do not match the ones for "
            "the data given during fit.")
-    with pytest.warns(DeprecationWarning, match=msg):
+    with pytest.warns(FutureWarning, match=msg):
         tf.transform(df_extra)
 
     tf = ColumnTransformer([('bycol', Trans(), [0])])
     tf.fit(df)
 
-    with pytest.warns(DeprecationWarning, match=msg):
+    with pytest.warns(FutureWarning, match=msg):
         tf.transform(X_extra)
 
     with warnings.catch_warnings(record=True) as warns:
@@ -1149,7 +1149,7 @@ def test_feature_name_validation():
     tf = ColumnTransformer([('bycol', Trans(), ['a'])],
                            remainder=Trans())
     tf.fit(df)
-    with pytest.warns(DeprecationWarning, match=msg):
+    with pytest.warns(FutureWarning, match=msg):
         tf.transform(df_extra)
 
     tf = ColumnTransformer([('bycol', Trans(), [0, -1])])

sklearn/datasets/_base.py

@@ -932,6 +932,6 @@ def _refresh_cache(files, compress):
                        "import path. Consider removing the following files "
                        "and allowing it to be cached anew:\n%s"
                        % ("\n".join(files)))
-            warnings.warn(message=message, category=DeprecationWarning)
+            warnings.warn(message=message, category=FutureWarning)
 
     return data[0] if len(data) == 1 else data

sklearn/datasets/tests/test_base.py

@@ -292,11 +292,11 @@ def _load_warn(*args, **kwargs):
                "If this warning is raised when loading pickled models, you "
                "may need to re-serialize those models with scikit-learn "
                "0.21+.")
-        warnings.warn(msg, DeprecationWarning)
+        warnings.warn(msg, FutureWarning)
         return 0
 
     def _load_warn_unrelated(*args, **kwargs):
-        warnings.warn("unrelated warning", DeprecationWarning)
+        warnings.warn("unrelated warning", FutureWarning)
         return 0
 
     def _dump_safe(*args, **kwargs):
@@ -312,7 +312,7 @@ def _dump_raise(*args, **kwargs):
     monkeypatch.setattr(joblib, "load", _load_warn)
     monkeypatch.setattr(joblib, "dump", _dump_raise)
     msg = "This dataset will stop being loadable in scikit-learn"
-    with pytest.warns(DeprecationWarning, match=msg):
+    with pytest.warns(FutureWarning, match=msg):
         _refresh_cache('test', 0)
 
     # make sure no warning is raised if load raises the warning, but dump
@@ -327,5 +327,5 @@ def _dump_raise(*args, **kwargs):
     # by _refresh_cache
     monkeypatch.setattr(joblib, "load", _load_warn_unrelated)
     monkeypatch.setattr(joblib, "dump", _dump_safe)
-    with pytest.warns(DeprecationWarning, match="unrelated warning"):
+    with pytest.warns(FutureWarning, match="unrelated warning"):
         _refresh_cache('test', 0)

sklearn/decomposition/_sparse_pca.py

@@ -20,7 +20,7 @@ def _check_normalize_components(normalize_components, estimator_name):
             warnings.warn(
                 "'normalize_components' has been deprecated in 0.22 and "
                 "will be removed in 0.24. Remove the parameter from the "
-                " constructor.", DeprecationWarning
+                " constructor.", FutureWarning
             )
         else:
             raise NotImplementedError(

sklearn/decomposition/tests/test_sparse_pca.py

@@ -13,7 +13,6 @@
 from sklearn.decomposition import SparsePCA, MiniBatchSparsePCA, PCA
 from sklearn.utils import check_random_state
 
-
 def generate_toy_data(n_components, n_samples, image_size, random_state=None):
     n_features = image_size[0] * image_size[1]
 
@@ -196,7 +195,7 @@ def test_spca_deprecation_warning(spca):
     Y, _, _ = generate_toy_data(3, 10, (8, 8), random_state=rng)
 
     warn_msg = "'normalize_components' has been deprecated in 0.22"
-    with pytest.warns(DeprecationWarning, match=warn_msg):
+    with pytest.warns(FutureWarning, match=warn_msg):
         spca(normalize_components=True).fit(Y)
 
 

sklearn/ensemble/_base.py

@@ -230,7 +230,7 @@ def _validate_estimators(self):
             warnings.warn(
                 "Using 'None' to drop an estimator from the ensemble is "
                 "deprecated in 0.22 and support will be dropped in 0.24. "
-                "Use the string 'drop' instead.", DeprecationWarning
+                "Use the string 'drop' instead.", FutureWarning
             )
 
         has_estimator = any(est not in (None, 'drop') for est in estimators)

sklearn/ensemble/_gb.py

@@ -1339,7 +1339,7 @@ def _check_params(self):
                           "suppress this warning by not passing any value "
                           "to the 'presort' parameter. We also recommend "
                           "using HistGradientBoosting models instead.",
-                          DeprecationWarning)
+                          FutureWarning)
 
     def _init_state(self):
         """Initialize model state and allocate model state data structures. """

sklearn/ensemble/_iforest.py

@@ -233,7 +233,7 @@ def fit(self, X, y=None, sample_weight=None):
                 warn(
                     "'behaviour' is deprecated in 0.22 and will be removed "
                     "in 0.24. You should not pass or set this parameter.",
-                    DeprecationWarning
+                    FutureWarning
                 )
             else:
                 raise NotImplementedError(

sklearn/ensemble/tests/test_forest.py

@@ -1238,7 +1238,8 @@ def test_min_impurity_split():
 
     for Estimator in all_estimators:
         est = Estimator(min_impurity_split=0.1)
-        est = assert_warns_message(DeprecationWarning, "min_impurity_decrease",
+        est = assert_warns_message(FutureWarning,
+                                   "min_impurity_decrease",
                                    est.fit, X, y)
         for tree in est.estimators_:
             assert tree.min_impurity_split == 0.1

sklearn/ensemble/tests/test_gradient_boosting.py

@@ -1087,7 +1087,8 @@ def test_min_impurity_split(GBEstimator):
     X, y = datasets.make_hastie_10_2(n_samples=100, random_state=1)
 
     est = GBEstimator(min_impurity_split=0.1)
-    est = assert_warns_message(DeprecationWarning, "min_impurity_decrease",
+    est = assert_warns_message(FutureWarning,
+                               "min_impurity_decrease",
                                est.fit, X, y)
     for tree in est.estimators_.flat:
         assert tree.min_impurity_split == 0.1
@@ -1400,6 +1401,6 @@ def test_presort_deprecated(Cls, presort):
     X = np.zeros((10, 10))
     y = np.r_[[0] * 5, [1] * 5]
     gb = Cls(presort=presort)
-    with pytest.warns(DeprecationWarning,
+    with pytest.warns(FutureWarning,
                       match="The parameter 'presort' is deprecated "):
         gb.fit(X, y)

sklearn/ensemble/tests/test_iforest.py

@@ -320,7 +320,7 @@ def test_iforest_chunks_works2(
 def test_iforest_deprecation():
     iforest = IsolationForest(behaviour='new')
     warn_msg = "'behaviour' is deprecated in 0.22 and will be removed in 0.24"
-    with pytest.warns(DeprecationWarning, match=warn_msg):
+    with pytest.warns(FutureWarning, match=warn_msg):
         iforest.fit(iris.data)
 
 

sklearn/ensemble/tests/test_partial_dependence.py

@@ -27,7 +27,7 @@
 iris = datasets.load_iris()
 
 
-@ignore_warnings(category=DeprecationWarning)
+@ignore_warnings(category=FutureWarning)
 def test_partial_dependence_classifier():
     # Test partial dependence for classifier
     clf = GradientBoostingClassifier(n_estimators=10, random_state=1)
@@ -66,7 +66,7 @@ def test_partial_dependence_classifier():
     assert np.all(np.abs(pdp_w2 - pdp_w) / np.abs(pdp_w) > 0.1)
 
 
-@ignore_warnings(category=DeprecationWarning)
+@ignore_warnings(category=FutureWarning)
 def test_partial_dependence_multiclass():
     # Test partial dependence for multi-class classifier
     clf = GradientBoostingClassifier(n_estimators=10, random_state=1)
@@ -82,7 +82,7 @@ def test_partial_dependence_multiclass():
     assert axes[0].shape[0] == grid_resolution
 
 
-@ignore_warnings(category=DeprecationWarning)
+@ignore_warnings(category=FutureWarning)
 def test_partial_dependence_regressor():
     # Test partial dependence for regressor
     clf = GradientBoostingRegressor(n_estimators=10, random_state=1)
@@ -96,7 +96,7 @@ def test_partial_dependence_regressor():
     assert axes[0].shape[0] == grid_resolution
 
 
-@ignore_warnings(category=DeprecationWarning)
+@ignore_warnings(category=FutureWarning)
 def test_partial_dependence_sample_weight():
     # Test near perfect correlation between partial dependence and diagonal
     # when sample weights emphasize y = x predictions
@@ -122,7 +122,7 @@ def test_partial_dependence_sample_weight():
     assert np.corrcoef(np.ravel(pdp[0]), grid)[0, 1] > 0.99
 
 
-@ignore_warnings(category=DeprecationWarning)
+@ignore_warnings(category=FutureWarning)
 def test_partial_dependecy_input():
     # Test input validation of partial dependence.
     clf = GradientBoostingClassifier(n_estimators=10, random_state=1)
@@ -151,7 +151,7 @@ def test_partial_dependecy_input():
     assert_raises(ValueError, partial_dependence, clf, [0], grid=grid)
 
 
-@ignore_warnings(category=DeprecationWarning)
+@ignore_warnings(category=FutureWarning)
 @pytest.mark.filterwarnings('ignore: Using or importing the ABCs from')
 # matplotlib Python3.7 warning
 def test_plot_partial_dependence(pyplot):
@@ -187,7 +187,7 @@ def test_plot_partial_dependence(pyplot):
 
 @pytest.mark.filterwarnings('ignore: Using or importing the ABCs from')
 # matplotlib Python3.7 warning
-@ignore_warnings(category=DeprecationWarning)
+@ignore_warnings(category=FutureWarning)
 def test_plot_partial_dependence_input(pyplot):
     # Test partial dependence plot function input checks.
     clf = GradientBoostingClassifier(n_estimators=10, random_state=1)
@@ -224,7 +224,7 @@ def test_plot_partial_dependence_input(pyplot):
 
 @pytest.mark.filterwarnings('ignore: Using or importing the ABCs from')
 # matplotlib Python3.7 warning
-@ignore_warnings(category=DeprecationWarning)
+@ignore_warnings(category=FutureWarning)
 def test_plot_partial_dependence_multiclass(pyplot):
     # Test partial dependence plot function on multi-class input.
     clf = GradientBoostingClassifier(n_estimators=10, random_state=1)
@@ -273,5 +273,5 @@ def test_raise_deprecation_warning(pyplot, func, params):
     warn_msg = "The function ensemble.{} has been deprecated".format(
         func.__name__
     )
-    with pytest.warns(DeprecationWarning, match=warn_msg):
+    with pytest.warns(FutureWarning, match=warn_msg):
         func(clf, **params, grid_resolution=grid_resolution)

sklearn/ensemble/tests/test_voting.py

@@ -558,7 +558,7 @@ def test_deprecate_none_transformer(Voter, BaseEstimator):
     msg = ("Using 'None' to drop an estimator from the ensemble is "
            "deprecated in 0.22 and support will be dropped in 0.24. "
            "Use the string 'drop' instead.")
-    with pytest.warns(DeprecationWarning, match=msg):
+    with pytest.warns(FutureWarning, match=msg):
         est.fit(X, y)
 
 

sklearn/externals/joblib/__init__.py

@@ -12,4 +12,4 @@
        "those models with scikit-learn 0.21+.")
 
 if not hasattr(sys, "_is_pytest_session"):
-    warnings.warn(msg, category=DeprecationWarning)
+    warnings.warn(msg, category=FutureWarning)

sklearn/externals/six.py

@@ -28,7 +28,7 @@
 warnings.warn("The module is deprecated in version 0.21 and will be removed "
               "in version 0.23 since we've dropped support for Python 2.7. "
               "Please rely on the official version of six "
-              "(https://pypi.org/project/six/).", DeprecationWarning)
+              "(https://pypi.org/project/six/).", FutureWarning)
 
 __author__ = "Benjamin Peterson <benjamin@python.org>"
 __version__ = "1.4.1"

sklearn/feature_extraction/tests/test_feature_hasher.py

@@ -144,7 +144,7 @@ def test_hasher_zeros():
     assert X.data.shape == (0,)
 
 
-@ignore_warnings(category=DeprecationWarning)
+@ignore_warnings(category=FutureWarning)
 def test_hasher_alternate_sign():
     X = [list("Thequickbrownfoxjumped")]
 

sklearn/feature_extraction/tests/test_image.py

@@ -340,5 +340,5 @@ def test_width_patch():
 def test_extract_patches_deprecated():
     msg = ("The function feature_extraction.image.extract_patches has been "
            "deprecated in 0.22 and will be removed in 0.24.")
-    with pytest.warns(DeprecationWarning, match=msg):
+    with pytest.warns(FutureWarning, match=msg):
         extract_patches(downsampled_face)