scikit-learn · adrinjalali · Jan 13, 2020 · Dec 5, 2019 · Dec 5, 2019 · Dec 5, 2019
diff --git a/sklearn/linear_model/_least_angle.py b/sklearn/linear_model/_least_angle.py
@@ -47,12 +47,6 @@ def lars_path(X, y, Xy=None, Gram=None, max_iter=500, alpha_min=0,
         Input data. Note that if X is None then the Gram matrix must be
         specified, i.e., cannot be None or False.
 
-        .. deprecated:: 0.21
-
-           The use of ``X`` is ``None`` in combination with ``Gram`` is not
-           ``None`` will be removed in v0.23. Use :func:`lars_path_gram`
-           instead.
-
     y : None or array-like of shape (n_samples,)
         Input targets.
 
@@ -67,11 +61,6 @@ def lars_path(X, y, Xy=None, Gram=None, max_iter=500, alpha_min=0,
         matrix is precomputed from the given X, if there are more samples
         than features.
 
-        .. deprecated:: 0.21
-
-           The use of ``X`` is ``None`` in combination with ``Gram`` is not
-           None will be removed in v0.23. Use :func:`lars_path_gram` instead.
-
     max_iter : int, default=500
         Maximum number of iterations to perform, set to infinity for no limit.
 
@@ -155,9 +144,10 @@ def lars_path(X, y, Xy=None, Gram=None, max_iter=500, alpha_min=0,
 
     """
     if X is None and Gram is not None:
-        warnings.warn('Use lars_path_gram to avoid passing X and y. '
-                      'The current option will be removed in v0.23.',
-                      FutureWarning)
+        raise ValueError(
+            'X cannot be None if Gram is not None'
+            'Use lars_path_gram to avoid passing X and y.'
+        )
     return _lars_path_solver(
         X=X, y=y, Xy=Xy, Gram=Gram, n_samples=None, max_iter=max_iter,
         alpha_min=alpha_min, method=method, copy_X=copy_X,

diff --git a/sklearn/linear_model/tests/test_least_angle.py b/sklearn/linear_model/tests/test_least_angle.py
@@ -15,7 +15,8 @@
 from sklearn.utils._testing import TempMemmap
 from sklearn.exceptions import ConvergenceWarning
 from sklearn import linear_model, datasets
-from sklearn.linear_model._least_angle import _lars_path_residues, LassoLarsIC
+from sklearn.linear_model._least_angle import _lars_path_residues
+from sklearn.linear_model import LassoLarsIC, lars_path
 
 # TODO: use another dataset that has multiple drops
 diabetes = datasets.load_diabetes()
@@ -730,3 +731,9 @@ def test_lasso_lars_fit_copyX_behaviour(copy_X):
     y = X[:, 2]
     lasso_lars.fit(X, y, copy_X=copy_X)
     assert copy_X == np.array_equal(X, X_copy)
+
+
+def test_X_none_gram_not_none():
+    with pytest.raises(ValueError,
+                       match="X cannot be None if Gram is not None"):
+        lars_path(X=None, y=[1], Gram='not None')
diff --git a/sklearn/multioutput.py b/sklearn/multioutput.py
@@ -265,44 +265,6 @@ def partial_fit(self, X, y, sample_weight=None):
         super().partial_fit(
             X, y, sample_weight=sample_weight)
 
-    # XXX Remove this method in 0.23
-    def score(self, X, y, sample_weight=None):
-        """Returns the coefficient of determination R^2 of the prediction.
-
-        The coefficient R^2 is defined as (1 - u/v), where u is the residual
-        sum of squares ((y_true - y_pred) ** 2).sum() and v is the regression
-        sum of squares ((y_true - y_true.mean()) ** 2).sum().
-        Best possible score is 1.0 and it can be negative (because the
-        model can be arbitrarily worse). A constant model that always
-        predicts the expected value of y, disregarding the input features,
-        would get a R^2 score of 0.0.
-
-        Notes
-        -----
-        R^2 is calculated by weighting all the targets equally using
-        `multioutput='uniform_average'`.
-
-        Parameters
-        ----------
-        X : array-like, shape (n_samples, n_features)
-            Test samples.
-
-        y : array-like, shape (n_samples) or (n_samples, n_outputs)
-            True values for X.
-
-        sample_weight : array-like, shape [n_samples], optional
-            Sample weights.
-
-        Returns
-        -------
-        score : float
-            R^2 of self.predict(X) wrt. y.
-        """
-        # XXX remove in 0.19 when r2_score default for multioutput changes
-        from .metrics import r2_score
-        return r2_score(y, self.predict(X), sample_weight=sample_weight,
-                        multioutput='uniform_average')
-
 
 class MultiOutputClassifier(ClassifierMixin, _MultiOutputEstimator):
     """Multi target classification

diff --git a/sklearn/svm/_classes.py b/sklearn/svm/_classes.py
@@ -214,18 +214,6 @@ def fit(self, X, y, sample_weight=None):
         self : object
             An instance of the estimator.
         """
-        # FIXME Remove l1/l2 support in 0.23 ----------------------------------
-        msg = ("loss='%s' has been deprecated in favor of "
-               "loss='%s' as of 0.16. Backward compatibility"
-               " for the loss='%s' will be removed in %s")
-
-        if self.loss in ('l1', 'l2'):
-            old_loss = self.loss
-            self.loss = {'l1': 'hinge', 'l2': 'squared_hinge'}.get(self.loss)
-            warnings.warn(msg % (old_loss, self.loss, old_loss, '0.23'),
-                          FutureWarning)
-        # ---------------------------------------------------------------------
-
         if self.C < 0:
             raise ValueError("Penalty term must be positive; got (C=%r)"
                              % self.C)
@@ -403,20 +391,6 @@ def fit(self, X, y, sample_weight=None):
         -------
         self : object
         """
-        # FIXME Remove l1/l2 support in 0.23 ----------------------------------
-        msg = ("loss='%s' has been deprecated in favor of "
-               "loss='%s' as of 0.16. Backward compatibility"
-               " for the loss='%s' will be removed in %s")
-
-        if self.loss in ('l1', 'l2'):
-            old_loss = self.loss
-            self.loss = {'l1': 'epsilon_insensitive',
-                         'l2': 'squared_epsilon_insensitive'
-                         }.get(self.loss)
-            warnings.warn(msg % (old_loss, self.loss, old_loss, '0.23'),
-                          FutureWarning)
-        # ---------------------------------------------------------------------
-
         if self.C < 0:
             raise ValueError("Penalty term must be positive; got (C=%r)"
                              % self.C)

diff --git a/sklearn/svm/tests/test_svm.py b/sklearn/svm/tests/test_svm.py
@@ -736,39 +736,6 @@ def test_linearsvc_parameters():
         svm.LinearSVC(loss="l3").fit(X, y)
 
 
-# FIXME remove in 0.23
-def test_linearsvx_loss_penalty_deprecations():
-    X, y = [[0.0], [1.0]], [0, 1]
-
-    msg = ("loss='%s' has been deprecated in favor of "
-           "loss='%s' as of 0.16. Backward compatibility"
-           " for the %s will be removed in %s")
-
-    # LinearSVC
-    # loss l1 --> hinge
-    assert_warns_message(FutureWarning,
-                         msg % ("l1", "hinge", "loss='l1'", "0.23"),
-                         svm.LinearSVC(loss="l1").fit, X, y)
-
-    # loss l2 --> squared_hinge
-    assert_warns_message(FutureWarning,
-                         msg % ("l2", "squared_hinge", "loss='l2'", "0.23"),
-                         svm.LinearSVC(loss="l2").fit, X, y)
-
-    # LinearSVR
-    # loss l1 --> epsilon_insensitive
-    assert_warns_message(FutureWarning,
-                         msg % ("l1", "epsilon_insensitive", "loss='l1'",
-                                "0.23"),
-                         svm.LinearSVR(loss="l1").fit, X, y)
-
-    # loss l2 --> squared_epsilon_insensitive
-    assert_warns_message(FutureWarning,
-                         msg % ("l2", "squared_epsilon_insensitive",
-                                "loss='l2'", "0.23"),
-                         svm.LinearSVR(loss="l2").fit, X, y)
-
-
 def test_linear_svx_uppercase_loss_penality_raises_error():
     # Check if Upper case notation raises error at _fit_liblinear
     # which is called by fit