MAINT change the default solver in QuantileRegressor (scikit-learn#27943)

glemaitre · web-flow · commit e430f6e127f3 · 2023-12-12T17:01:54.000+01:00
diff --git a/sklearn/decomposition/tests/test_sparse_pca.py b/sklearn/decomposition/tests/test_sparse_pca.py
@@ -268,7 +268,7 @@ def test_spca_feature_names_out(SPCA):
     assert_array_equal([f"{estimator_name}{i}" for i in range(4)], names)
 
 
-# TODO (1.6): remove in 1.6
+# TODO(1.6): remove in 1.6
 def test_spca_max_iter_None_deprecation():
     """Check that we raise a warning for the deprecation of `max_iter=None`."""
     rng = np.random.RandomState(0)
diff --git a/sklearn/linear_model/_quantile.py b/sklearn/linear_model/_quantile.py
@@ -11,7 +11,7 @@
 from ..base import BaseEstimator, RegressorMixin, _fit_context
 from ..exceptions import ConvergenceWarning
 from ..utils import _safe_indexing
-from ..utils._param_validation import Hidden, Interval, StrOptions
+from ..utils._param_validation import Interval, StrOptions
 from ..utils.fixes import parse_version, sp_version
 from ..utils.validation import _check_sample_weight
 from ._base import LinearModel
@@ -44,7 +44,7 @@ class QuantileRegressor(LinearModel, RegressorMixin, BaseEstimator):
         Whether or not to fit the intercept.
 
     solver : {'highs-ds', 'highs-ipm', 'highs', 'interior-point', \
-            'revised simplex'}, default='interior-point'
+            'revised simplex'}, default='highs'
         Method used by :func:`scipy.optimize.linprog` to solve the linear
         programming formulation.
 
@@ -55,7 +55,7 @@ class QuantileRegressor(LinearModel, RegressorMixin, BaseEstimator):
         From `scipy>=1.11.0`, "interior-point" is not available anymore.
 
         .. versionchanged:: 1.4
-           The default of `solver` will change to `"highs"` in version 1.4.
+           The default of `solver` changed to `"highs"` in version 1.4.
 
     solver_options : dict, default=None
         Additional parameters passed to :func:`scipy.optimize.linprog` as
@@ -121,7 +121,6 @@ class QuantileRegressor(LinearModel, RegressorMixin, BaseEstimator):
                     "revised simplex",
                 }
             ),
-            Hidden(StrOptions({"warn"})),
         ],
         "solver_options": [dict, None],
     }
@@ -132,7 +131,7 @@ def __init__(
         quantile=0.5,
         alpha=1.0,
         fit_intercept=True,
-        solver="warn",
+        solver="highs",
         solver_options=None,
     ):
         self.quantile = quantile
@@ -182,17 +181,7 @@ def fit(self, X, y, sample_weight=None):
         # So we rescale the penalty term, which is equivalent.
         alpha = np.sum(sample_weight) * self.alpha
 
-        if self.solver == "warn":
-            warnings.warn(
-                (
-                    "The default solver will change from 'interior-point' to 'highs' in"
-                    " version 1.4. Set `solver='highs'` or to the desired solver to"
-                    " silence this warning."
-                ),
-                FutureWarning,
-            )
-            solver = "interior-point"
-        elif self.solver in (
+        if self.solver in (
             "highs-ds",
             "highs-ipm",
             "highs",
diff --git a/sklearn/linear_model/tests/test_quantile.py b/sklearn/linear_model/tests/test_quantile.py
@@ -292,19 +292,6 @@ def test_sparse_input(sparse_container, solver, fit_intercept, default_solver):
         assert 0.45 <= np.mean(y < quant_sparse.predict(X_sparse)) <= 0.57
 
 
-# TODO (1.4): remove this test in 1.4
-@pytest.mark.skipif(
-    parse_version(sp_version.base_version) >= parse_version("1.11"),
-    reason="interior-point solver is not available in SciPy 1.11",
-)
-def test_warning_new_default(X_y_data):
-    """Check that we warn about the new default solver."""
-    X, y = X_y_data
-    model = QuantileRegressor()
-    with pytest.warns(FutureWarning, match="The default solver will change"):
-        model.fit(X, y)
-
-
 def test_error_interior_point_future(X_y_data, monkeypatch):
     """Check that we will raise a proper error when requesting
     `solver='interior-point'` in SciPy >= 1.11.
