[MRG+1] Fixes #10393 Fixed error when fitting RidgeCV with integers

doc/whats_new/v0.20.rst

@@ -277,6 +277,9 @@ Classifiers and regressors
   where the coefficient had wrong shape when ``fit_intercept=False``.
   :issue:`10687` by :user:`Martin Hahn <martin-hahn>`.
 
+- Fixed a bug in :class:`linear_model.RidgeCV` where using integer ``alphas``
+  raised an error. :issue:`10393` by :user:`Mabel Villalba-Jiménez <mabelvj>`.
+
 Decomposition, manifold learning and clustering
 
 - Fix for uninformative error in :class:`decomposition.IncrementalPCA`:
@@ -471,4 +474,4 @@ Changes to estimator checks
 
 - Add test :func:`estimator_checks.check_methods_subset_invariance` to check
   that estimators methods are invariant if applied to a data subset.
-  :issue:`10420` by :user:`Jonathan Ohayon <Johayon>`
+  :issue:`10420` by :user:`Jonathan Ohayon <Johayon>`

sklearn/linear_model/ridge.py

@@ -778,6 +778,7 @@ class RidgeClassifier(LinearClassifierMixin, _BaseRidge):
     a one-versus-all approach. Concretely, this is implemented by taking
     advantage of the multi-variate response support in Ridge.
     """
+
     def __init__(self, alpha=1.0, fit_intercept=True, normalize=False,
                  copy_X=True, max_iter=None, tol=1e-3, class_weight=None,
                  solver="auto", random_state=None):
@@ -1041,11 +1042,16 @@ def fit(self, X, y, sample_weight=None):
         scorer = check_scoring(self, scoring=self.scoring, allow_none=True)
         error = scorer is None
 
+        if np.any(self.alphas < 0):
+            raise ValueError("alphas cannot be negative. "
+                             "Got {} containing some "
+                             "negative value instead.".format(self.alphas))
+
         for i, alpha in enumerate(self.alphas):
             if error:
-                out, c = _errors(alpha, y, v, Q, QT_y)
+                out, c = _errors(float(alpha), y, v, Q, QT_y)
             else:
-                out, c = _values(alpha, y, v, Q, QT_y)
+                out, c = _values(float(alpha), y, v, Q, QT_y)
             cv_values[:, i] = out.ravel()
             C.append(c)
 
@@ -1085,7 +1091,7 @@ def __init__(self, alphas=(0.1, 1.0, 10.0),
                  fit_intercept=True, normalize=False, scoring=None,
                  cv=None, gcv_mode=None,
                  store_cv_values=False):
-        self.alphas = alphas
+        self.alphas = np.asarray(alphas)
         self.fit_intercept = fit_intercept
         self.normalize = normalize
         self.scoring = scoring
@@ -1328,6 +1334,7 @@ class RidgeClassifierCV(LinearClassifierMixin, _BaseRidgeCV):
     a one-versus-all approach. Concretely, this is implemented by taking
     advantage of the multi-variate response support in Ridge.
     """
+
     def __init__(self, alphas=(0.1, 1.0, 10.0), fit_intercept=True,
                  normalize=False, scoring=None, cv=None, class_weight=None):
         super(RidgeClassifierCV, self).__init__(

sklearn/linear_model/tests/test_ridge.py

@@ -11,6 +11,7 @@
 from sklearn.utils.testing import assert_greater
 from sklearn.utils.testing import assert_raises
 from sklearn.utils.testing import assert_raise_message
+from sklearn.utils.testing import assert_raises_regex
 from sklearn.utils.testing import ignore_warnings
 from sklearn.utils.testing import assert_warns
 
@@ -51,6 +52,7 @@
 X_iris = sp.csr_matrix(iris.data)
 y_iris = iris.target
 
+
 DENSE_FILTER = lambda X: X
 SPARSE_FILTER = lambda X: sp.csr_matrix(X)
 
@@ -704,6 +706,34 @@ def test_sparse_design_with_sample_weights():
                                       decimal=6)
 
 
+def test_ridgecv_int_alphas():
+    X = np.array([[-1.0, -1.0], [-1.0, 0], [-.8, -1.0],
+                  [1.0, 1.0], [1.0, 0.0]])
+    y = [1, 1, 1, -1, -1]
+
+    # Integers
+    ridge = RidgeCV(alphas=(1, 10, 100))
+    ridge.fit(X, y)
+
+
+def test_ridgecv_negative_alphas():
+    X = np.array([[-1.0, -1.0], [-1.0, 0], [-.8, -1.0],
+                  [1.0, 1.0], [1.0, 0.0]])
+    y = [1, 1, 1, -1, -1]
+
+    # Negative integers
+    ridge = RidgeCV(alphas=(-1, -10, -100))
+    assert_raises_regex(ValueError,
+                        "alphas cannot be negative.",
+                        ridge.fit, X, y)
+
+    # Negative floats
+    ridge = RidgeCV(alphas=(-0.1, -1.0, -10.0))
+    assert_raises_regex(ValueError,
+                        "alphas cannot be negative.",
+                        ridge.fit, X, y)
+
+
 def test_raises_value_error_if_solver_not_supported():
     # Tests whether a ValueError is raised if a non-identified solver
     # is passed to ridge_regression