From 6877342824457851f35cd6d386fa58a9d3fba6f5 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Sat, 2 Mar 2019 00:14:54 +0100
Subject: [PATCH 01/26] ridge_regression: fix selecting solver in 'auto' mode

---
 sklearn/linear_model/ridge.py            | 32 +++++++++++++++---------
 sklearn/linear_model/tests/test_ridge.py | 22 ++++++++++++++++
 2 files changed, 42 insertions(+), 12 deletions(-)

diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index e1fc9b42438e4..279099befb4b7 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -368,11 +368,26 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
                       return_n_iter=False, return_intercept=False,
                       X_scale=None, X_offset=None):
 
+    has_sw = sample_weight is not None
+
+    def _select_auto_mode():
+        if return_intercept:
+            # only sag and saga support fitting intercept directly
+            return "sag"
+        if has_sw:
+            # this should be changed since all solvers support sample_weights
+            return "cholesky"
+        if sparse.issparse(X):
+            return "sparse_cg"
+        return "cholesky"
+
+    if solver == 'auto':
+        solver = _select_auto_mode()
+
     if return_intercept and sparse.issparse(X) and solver != 'sag':
-        if solver != 'auto':
-            warnings.warn("In Ridge, only 'sag' solver can currently fit the "
-                          "intercept when X is sparse. Solver has been "
-                          "automatically changed into 'sag'.")
+        warnings.warn("In Ridge, only 'sag' solver can currently fit the "
+                      "intercept. Solver has been "
+                      "automatically changed into 'sag'.")
         solver = 'sag'
 
     _dtype = [np.float64, np.float32]
@@ -404,14 +419,7 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
         raise ValueError("Number of samples in X and y does not correspond:"
                          " %d != %d" % (n_samples, n_samples_))
 
-    has_sw = sample_weight is not None
 
-    if solver == 'auto':
-        # cholesky if it's a dense array and cg in any other case
-        if not sparse.issparse(X) or has_sw:
-            solver = 'cholesky'
-        else:
-            solver = 'sparse_cg'
 
     if has_sw:
         if np.atleast_1d(sample_weight).ndim > 1:
@@ -555,7 +563,7 @@ def fit(self, X, y, sample_weight=None):
             # add the offset which was subtracted by _preprocess_data
             self.intercept_ += y_offset
         else:
-            if sparse.issparse(X):
+            if sparse.issparse(X) and self.solver == 'sparse_cg':
                 # required to fit intercept with sparse_cg solver
                 params = {'X_offset': X_offset, 'X_scale': X_scale}
             else:
diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 4a0bd1d86e2ef..c824aec1d8b6e 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -837,6 +837,28 @@ def test_ridge_fit_intercept_sparse():
         assert_array_almost_equal(dense.coef_, sparse.coef_)
 
 
+@pytest.mark.parametrize('return_intercept', [False, True])
+@pytest.mark.parametrize('sample_weight', [None, np.ones(1000)])
+@pytest.mark.parametrize('arr_type', [np.array, sp.csr_matrix])
+def test_ridge_check_auto_modes(return_intercept, sample_weight, arr_type):
+    X = np.random.rand(1000, 3)
+    true_coefs = [1, 2, 0.1]
+    y = np.dot(X, true_coefs)
+    X_testing = arr_type(X)
+
+    target = ridge_regression(X_testing, y, 1,
+                              solver='auto',
+                              sample_weight=sample_weight,
+                              return_intercept=return_intercept
+                              )
+    try:
+        coef, intercept = target
+        assert_array_almost_equal(coef, true_coefs, decimal=1)
+        assert_array_almost_equal(intercept, 0, decimal=1)
+    except ValueError:
+        assert_array_almost_equal(target, true_coefs, decimal=1)
+
+
 def test_errors_and_values_helper():
     ridgecv = _RidgeGCV()
     rng = check_random_state(42)

From 39a387b47c5d46982f9ff250a2a2386feeb11ef1 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Sat, 2 Mar 2019 00:30:50 +0100
Subject: [PATCH 02/26] test for return_intercept=True and solver specified

---
 sklearn/linear_model/ridge.py            |  9 +++++++--
 sklearn/linear_model/tests/test_ridge.py | 21 +++++++++++++++++++++
 2 files changed, 28 insertions(+), 2 deletions(-)

diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index 279099befb4b7..70b2d903b590b 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -384,12 +384,19 @@ def _select_auto_mode():
     if solver == 'auto':
         solver = _select_auto_mode()
 
+    if solver not in ('sparse_cg', 'cholesky', 'svd', 'lsqr', 'sag', 'saga'):
+        raise ValueError('Solver %s not understood' % solver)
+
     if return_intercept and sparse.issparse(X) and solver != 'sag':
         warnings.warn("In Ridge, only 'sag' solver can currently fit the "
                       "intercept. Solver has been "
                       "automatically changed into 'sag'.")
         solver = 'sag'
 
+    if return_intercept and solver not in ['sag', 'saga']:
+        raise ValueError("return_intercept=True is only supported with sag "
+                         "and saga solvers")
+
     _dtype = [np.float64, np.float32]
 
     # SAG needs X and y columns to be C-contiguous and np.float64
@@ -440,8 +447,6 @@ def _select_auto_mode():
     if alpha.size == 1 and n_targets > 1:
         alpha = np.repeat(alpha, n_targets)
 
-    if solver not in ('sparse_cg', 'cholesky', 'svd', 'lsqr', 'sag', 'saga'):
-        raise ValueError('Solver %s not understood' % solver)
 
     n_iter = None
     if solver == 'sparse_cg':
diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index c824aec1d8b6e..856945b3d0603 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -859,6 +859,27 @@ def test_ridge_check_auto_modes(return_intercept, sample_weight, arr_type):
         assert_array_almost_equal(target, true_coefs, decimal=1)
 
 
+def test_ridge_regression_fail_with_return_intercept():
+    # return_itercept is only supported by sag/saga
+
+    X, y = make_regression(n_samples=1000, n_features=2, n_informative=2,
+                           bias=10., random_state=42)
+
+    for solver in ['sparse_cg', 'cholesky', 'svd', 'lsqr']:
+        with pytest.raises(ValueError, match="return_intercept=True is only"):
+            target = ridge_regression(X, y, 1,
+                                      solver=solver,
+                                      return_intercept=True
+                                      )
+
+    for solver in ['saga', 'sag']:
+        target = ridge_regression(X, y, 1,
+                                  solver=solver,
+                                  return_intercept=True
+                                 )
+        assert len(target) == 2
+        assert target[0].shape == (2,)
+
 def test_errors_and_values_helper():
     ridgecv = _RidgeGCV()
     rng = check_random_state(42)

From 62afd26951c11eee014f2e6a444a1b779a711bc4 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Sat, 2 Mar 2019 00:33:35 +0100
Subject: [PATCH 03/26] fix linting issue

---
 sklearn/linear_model/tests/test_ridge.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 856945b3d0603..52f5ed2637383 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -876,7 +876,7 @@ def test_ridge_regression_fail_with_return_intercept():
         target = ridge_regression(X, y, 1,
                                   solver=solver,
                                   return_intercept=True
-                                 )
+                                  )
         assert len(target) == 2
         assert target[0].shape == (2,)
 

From d3e085f38aa7da521a18a169b33a6ae96e434457 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Sat, 2 Mar 2019 00:49:42 +0100
Subject: [PATCH 04/26] change solver and issue warning instead of raising an
 exception

---
 sklearn/linear_model/ridge.py            |  6 +-----
 sklearn/linear_model/tests/test_ridge.py | 16 ++++++++++++----
 2 files changed, 13 insertions(+), 9 deletions(-)

diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index 70b2d903b590b..a4f38d5e32e35 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -387,16 +387,12 @@ def _select_auto_mode():
     if solver not in ('sparse_cg', 'cholesky', 'svd', 'lsqr', 'sag', 'saga'):
         raise ValueError('Solver %s not understood' % solver)
 
-    if return_intercept and sparse.issparse(X) and solver != 'sag':
+    if return_intercept and solver != 'sag':
         warnings.warn("In Ridge, only 'sag' solver can currently fit the "
                       "intercept. Solver has been "
                       "automatically changed into 'sag'.")
         solver = 'sag'
 
-    if return_intercept and solver not in ['sag', 'saga']:
-        raise ValueError("return_intercept=True is only supported with sag "
-                         "and saga solvers")
-
     _dtype = [np.float64, np.float32]
 
     # SAG needs X and y columns to be C-contiguous and np.float64
diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 52f5ed2637383..0a4b992b072b2 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -865,21 +865,29 @@ def test_ridge_regression_fail_with_return_intercept():
     X, y = make_regression(n_samples=1000, n_features=2, n_informative=2,
                            bias=10., random_state=42)
 
-    for solver in ['sparse_cg', 'cholesky', 'svd', 'lsqr']:
-        with pytest.raises(ValueError, match="return_intercept=True is only"):
+    for solver in ['sparse_cg', 'cholesky', 'svd', 'lsqr', 'saga']:
+        with pytest.warns(UserWarning) as record:
             target = ridge_regression(X, y, 1,
                                       solver=solver,
                                       return_intercept=True
                                       )
+            assert len(target) == 2
+            assert target[0].shape == (2,)
 
-    for solver in ['saga', 'sag']:
+        for r in record:
+            r.message.args[0].startswith("return_intercept=True is only")
+
+    with pytest.warns(None) as record:
         target = ridge_regression(X, y, 1,
-                                  solver=solver,
+                                  solver="sag",
                                   return_intercept=True
                                   )
         assert len(target) == 2
         assert target[0].shape == (2,)
 
+    # no warning should be raised
+    assert len(record) == 0
+
 def test_errors_and_values_helper():
     ridgecv = _RidgeGCV()
     rng = check_random_state(42)

From 596ae7f79236fa311adcd2c82584a2f33c2e13c7 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Sat, 2 Mar 2019 00:52:19 +0100
Subject: [PATCH 05/26] fix comment

---
 sklearn/linear_model/ridge.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index a4f38d5e32e35..24af1792f63a5 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -372,7 +372,7 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
 
     def _select_auto_mode():
         if return_intercept:
-            # only sag and saga support fitting intercept directly
+            # only sag supports fitting intercept directly
             return "sag"
         if has_sw:
             # this should be changed since all solvers support sample_weights

From b0bfbe1ea5e7d73b8fcfb2b977f5dd91d0005b99 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Sat, 2 Mar 2019 00:59:57 +0100
Subject: [PATCH 06/26] run the return_intercept test on all solvers

---
 sklearn/linear_model/tests/test_ridge.py | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 0a4b992b072b2..6032dd206a39c 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -840,14 +840,22 @@ def test_ridge_fit_intercept_sparse():
 @pytest.mark.parametrize('return_intercept', [False, True])
 @pytest.mark.parametrize('sample_weight', [None, np.ones(1000)])
 @pytest.mark.parametrize('arr_type', [np.array, sp.csr_matrix])
-def test_ridge_check_auto_modes(return_intercept, sample_weight, arr_type):
+@pytest.mark.parametrize('solver', ['auto', 'sparse_cg', 'cholesky', 'lsqr',
+                                    'sag', 'saga'])
+def test_ridge_regression_check_arguments_validity(return_intercept,
+                                                   sample_weight, arr_type,
+                                                   solver):
+    """check if all combinations of arguments give valid estimations"""
+
+    # test excludes 'svd' solver because it raises exception for sparse inputs
+
     X = np.random.rand(1000, 3)
     true_coefs = [1, 2, 0.1]
     y = np.dot(X, true_coefs)
     X_testing = arr_type(X)
 
     target = ridge_regression(X_testing, y, 1,
-                              solver='auto',
+                              solver=solver,
                               sample_weight=sample_weight,
                               return_intercept=return_intercept
                               )

From fdfbc50f32eb97d3c3f2d724cc84d3b9a2c25fbf Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Sat, 2 Mar 2019 01:02:15 +0100
Subject: [PATCH 07/26] change test name

---
 sklearn/linear_model/tests/test_ridge.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 6032dd206a39c..81f8ee2bb559c 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -867,7 +867,7 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
         assert_array_almost_equal(target, true_coefs, decimal=1)
 
 
-def test_ridge_regression_fail_with_return_intercept():
+def test_ridge_regression_warns_with_return_intercept():
     # return_itercept is only supported by sag/saga
 
     X, y = make_regression(n_samples=1000, n_features=2, n_informative=2,

From 73bbd9106419476c6cdecc054630334cc82391fe Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Sun, 3 Mar 2019 00:21:47 +0100
Subject: [PATCH 08/26] inline select auto mode

---
 sklearn/linear_model/ridge.py | 18 ++++++++----------
 1 file changed, 8 insertions(+), 10 deletions(-)

diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index 24af1792f63a5..dda7b0aceff2a 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -370,19 +370,17 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
 
     has_sw = sample_weight is not None
 
-    def _select_auto_mode():
+    if solver == 'auto':
         if return_intercept:
             # only sag supports fitting intercept directly
-            return "sag"
-        if has_sw:
+            solver = "sag"
+        elif has_sw:
             # this should be changed since all solvers support sample_weights
-            return "cholesky"
-        if sparse.issparse(X):
-            return "sparse_cg"
-        return "cholesky"
-
-    if solver == 'auto':
-        solver = _select_auto_mode()
+            solver = "cholesky"
+        elif sparse.issparse(X):
+            solver = "sparse_cg"
+        else:
+            solver = "cholesky"
 
     if solver not in ('sparse_cg', 'cholesky', 'svd', 'lsqr', 'sag', 'saga'):
         raise ValueError('Solver %s not understood' % solver)

From 70f7560b3ad02e61ca5e77a1bbd03546ab67e40c Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Sun, 3 Mar 2019 00:36:22 +0100
Subject: [PATCH 09/26] update whats new doc

---
 doc/whats_new/v0.21.rst | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/doc/whats_new/v0.21.rst b/doc/whats_new/v0.21.rst
index b2df99d9a131b..d46d412603e69 100644
--- a/doc/whats_new/v0.21.rst
+++ b/doc/whats_new/v0.21.rst
@@ -177,9 +177,9 @@ Support for Python 3.4 and below has been officially dropped.
 - |Enhancement| Minimized the validation of X in
   :class:`ensemble.AdaBoostClassifier` and :class:`ensemble.AdaBoostRegressor`
   :issue:`13174` by :user:`Christos Aridas <chkoar>`.
-  
+
 - |Enhancement| :class:`ensemble.IsolationForest` now exposes ``warm_start``
-  parameter, allowing iterative addition of trees to an isolation 
+  parameter, allowing iterative addition of trees to an isolation
   forest. :issue:`13496` by :user:`Peter Marko <petibear>`.
 
 - |Efficiency| Make :class:`ensemble.IsolationForest` more memory efficient
@@ -340,6 +340,11 @@ Support for Python 3.4 and below has been officially dropped.
   deterministic when trained in a multi-class setting on several threads.
   :issue:`13422` by :user:`Clément Doumouro <ClemDoum>`.
 
+- |Fix| Fixed bug in :func:`linear_model.ridge.ridge_regression` that
+  caused unhandled exception for arguments ``return_intercept=True`` and
+  ``solver=auto`` (default) or any other solver different from ``sag``.
+  :issue:`13363` by :user:`Bartosz Telenczuk <btel>`
+
 :mod:`sklearn.manifold`
 ............................
 

From b494e76f89ce84cc3e6e8c3666d552deecc75d1d Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Wed, 6 Mar 2019 22:55:08 +0100
Subject: [PATCH 10/26] remove try... except...

---
 sklearn/linear_model/tests/test_ridge.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 81f8ee2bb559c..7864f9ae5061c 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -859,11 +859,11 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
                               sample_weight=sample_weight,
                               return_intercept=return_intercept
                               )
-    try:
+    if return_intercept:
         coef, intercept = target
         assert_array_almost_equal(coef, true_coefs, decimal=1)
         assert_array_almost_equal(intercept, 0, decimal=1)
-    except ValueError:
+    else:
         assert_array_almost_equal(target, true_coefs, decimal=1)
 
 

From 3d44669808e116d34bf33b9712bc0f10c04ae063 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Wed, 6 Mar 2019 22:59:30 +0100
Subject: [PATCH 11/26] update error message

---
 sklearn/linear_model/ridge.py            | 3 ++-
 sklearn/linear_model/tests/test_ridge.py | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index dda7b0aceff2a..53265ae2d42a7 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -383,7 +383,8 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
             solver = "cholesky"
 
     if solver not in ('sparse_cg', 'cholesky', 'svd', 'lsqr', 'sag', 'saga'):
-        raise ValueError('Solver %s not understood' % solver)
+        raise ValueError("Known solver are 'sparse_cg', 'cholesky', 'svd'"
+                         " 'lsqr', 'sag' or 'saga'. Got %s." % solver)
 
     if return_intercept and solver != 'sag':
         warnings.warn("In Ridge, only 'sag' solver can currently fit the "
diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 7864f9ae5061c..d0af6a246c2a3 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -778,7 +778,8 @@ def test_raises_value_error_if_solver_not_supported():
     wrong_solver = "This is not a solver (MagritteSolveCV QuantumBitcoin)"
 
     exception = ValueError
-    message = "Solver %s not understood" % wrong_solver
+    message = ("Known solver are 'sparse_cg', 'cholesky', 'svd'"
+               " 'lsqr', 'sag' or 'saga'. Got %s." % wrong_solver)
 
     def func():
         X = np.eye(3)

From 9848429ae67d6f173a25aa64d3f72aea05f3644f Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Wed, 13 Mar 2019 00:56:22 +0100
Subject: [PATCH 12/26] fix typo

---
 sklearn/linear_model/ridge.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index 53265ae2d42a7..64c26f1b708ee 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -383,7 +383,7 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
             solver = "cholesky"
 
     if solver not in ('sparse_cg', 'cholesky', 'svd', 'lsqr', 'sag', 'saga'):
-        raise ValueError("Known solver are 'sparse_cg', 'cholesky', 'svd'"
+        raise ValueError("Known solvers are 'sparse_cg', 'cholesky', 'svd'"
                          " 'lsqr', 'sag' or 'saga'. Got %s." % solver)
 
     if return_intercept and solver != 'sag':

From 81573aaa39f1c6c2689209088e40ccc838c0ba74 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Fri, 15 Mar 2019 13:00:26 +0100
Subject: [PATCH 13/26] fix test (didn't update error message)

---
 sklearn/linear_model/tests/test_ridge.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index d0af6a246c2a3..3a9394d6c7711 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -778,7 +778,7 @@ def test_raises_value_error_if_solver_not_supported():
     wrong_solver = "This is not a solver (MagritteSolveCV QuantumBitcoin)"
 
     exception = ValueError
-    message = ("Known solver are 'sparse_cg', 'cholesky', 'svd'"
+    message = ("Known solvers are 'sparse_cg', 'cholesky', 'svd'"
                " 'lsqr', 'sag' or 'saga'. Got %s." % wrong_solver)
 
     def func():

From a2473ff21c287840b17b12f76f5522e475204849 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Thu, 21 Mar 2019 00:48:10 +0100
Subject: [PATCH 14/26] match warning

---
 sklearn/linear_model/tests/test_ridge.py | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 3a9394d6c7711..33823ef5b3fff 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -875,7 +875,7 @@ def test_ridge_regression_warns_with_return_intercept():
                            bias=10., random_state=42)
 
     for solver in ['sparse_cg', 'cholesky', 'svd', 'lsqr', 'saga']:
-        with pytest.warns(UserWarning) as record:
+        with pytest.warns(UserWarning, match="In Ridge, only 'sag' solver"):
             target = ridge_regression(X, y, 1,
                                       solver=solver,
                                       return_intercept=True
@@ -883,9 +883,6 @@ def test_ridge_regression_warns_with_return_intercept():
             assert len(target) == 2
             assert target[0].shape == (2,)
 
-        for r in record:
-            r.message.args[0].startswith("return_intercept=True is only")
-
     with pytest.warns(None) as record:
         target = ridge_regression(X, y, 1,
                                   solver="sag",

From 10c116cfde3ef258c34ad6d7b1518fbd9599188f Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Thu, 21 Mar 2019 01:18:30 +0100
Subject: [PATCH 15/26] change the auto selection sequence

---
 doc/whats_new/v0.21.rst       | 5 +++++
 sklearn/linear_model/ridge.py | 7 ++-----
 2 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/doc/whats_new/v0.21.rst b/doc/whats_new/v0.21.rst
index d46d412603e69..e3d9129229096 100644
--- a/doc/whats_new/v0.21.rst
+++ b/doc/whats_new/v0.21.rst
@@ -345,6 +345,11 @@ Support for Python 3.4 and below has been officially dropped.
   ``solver=auto`` (default) or any other solver different from ``sag``.
   :issue:`13363` by :user:`Bartosz Telenczuk <btel>`
 
+- |API| :func:`linear_model.ridge.ridge_regression` will choose ``sparse_cg``
+  solver for sparse inputs when ``solver=auto`` and ``sample_weight``
+  is provided (previously `cholesky` solver was selected). :issue:`13363`
+  by :user:`Bartosz Telenczuk <btel>`
+
 :mod:`sklearn.manifold`
 ............................
 
diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index 64c26f1b708ee..2a37c1a9e6964 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -374,13 +374,10 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
         if return_intercept:
             # only sag supports fitting intercept directly
             solver = "sag"
-        elif has_sw:
-            # this should be changed since all solvers support sample_weights
+        elif not sparse.issparse(X):
             solver = "cholesky"
-        elif sparse.issparse(X):
-            solver = "sparse_cg"
         else:
-            solver = "cholesky"
+            solver = "sparse_cg"
 
     if solver not in ('sparse_cg', 'cholesky', 'svd', 'lsqr', 'sag', 'saga'):
         raise ValueError("Known solvers are 'sparse_cg', 'cholesky', 'svd'"

From 05ba6272635d45c0fb77c22e0201e447a3c1e512 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Thu, 21 Mar 2019 01:18:45 +0100
Subject: [PATCH 16/26] use assert_allclose

---
 sklearn/linear_model/tests/test_ridge.py | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 33823ef5b3fff..d3b5d542bdff7 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -7,6 +7,7 @@
 
 from sklearn.utils.testing import assert_almost_equal
 from sklearn.utils.testing import assert_array_almost_equal
+from sklearn.utils.testing import assert_allclose
 from sklearn.utils.testing import assert_equal
 from sklearn.utils.testing import assert_array_equal
 from sklearn.utils.testing import assert_greater
@@ -862,10 +863,10 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
                               )
     if return_intercept:
         coef, intercept = target
-        assert_array_almost_equal(coef, true_coefs, decimal=1)
-        assert_array_almost_equal(intercept, 0, decimal=1)
+        assert_allclose(coef, true_coefs, atol=0.1)
+        assert_allclose(intercept, 0, atol=0.1)
     else:
-        assert_array_almost_equal(target, true_coefs, decimal=1)
+        assert_allclose(target, true_coefs, atol=0.1)
 
 
 def test_ridge_regression_warns_with_return_intercept():

From 90a7e564e9c23d567c2cbfbb39388de3e3d9be87 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Thu, 21 Mar 2019 14:13:56 +0100
Subject: [PATCH 17/26] changed atol to 0.03

---
 sklearn/linear_model/tests/test_ridge.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index d3b5d542bdff7..7e51a8e50b781 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -863,10 +863,10 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
                               )
     if return_intercept:
         coef, intercept = target
-        assert_allclose(coef, true_coefs, atol=0.1)
-        assert_allclose(intercept, 0, atol=0.1)
+        assert_allclose(coef, true_coefs, atol=0.03)
+        assert_allclose(intercept, 0, atol=0.03)
     else:
-        assert_allclose(target, true_coefs, atol=0.1)
+        assert_allclose(target, true_coefs, atol=0.03)
 
 
 def test_ridge_regression_warns_with_return_intercept():

From 97a532663b52f89ab91998cb509fe6671234fa60 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Tue, 9 Apr 2019 14:30:33 +0200
Subject: [PATCH 18/26] added rtol=0

---
 sklearn/linear_model/tests/test_ridge.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 7e51a8e50b781..d0aa9d02748e7 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -863,10 +863,10 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
                               )
     if return_intercept:
         coef, intercept = target
-        assert_allclose(coef, true_coefs, atol=0.03)
-        assert_allclose(intercept, 0, atol=0.03)
+        assert_allclose(coef, true_coefs, rtol=0, atol=0.03)
+        assert_allclose(intercept, 0, rtol=0, atol=0.03)
     else:
-        assert_allclose(target, true_coefs, atol=0.03)
+        assert_allclose(target, true_coefs, rtol=0, atol=0.03)
 
 
 def test_ridge_regression_warns_with_return_intercept():

From 06190130db5c4713f5c26f22dc6dc322739768b6 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Wed, 17 Apr 2019 01:18:19 +0200
Subject: [PATCH 19/26] use fixed random_state

---
 sklearn/linear_model/tests/test_ridge.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index d0aa9d02748e7..fc4bdc417dc5e 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -851,7 +851,8 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
 
     # test excludes 'svd' solver because it raises exception for sparse inputs
 
-    X = np.random.rand(1000, 3)
+    rng = check_random_state(42)
+    X = rng.rand(1000, 3)
     true_coefs = [1, 2, 0.1]
     y = np.dot(X, true_coefs)
     X_testing = arr_type(X)

From 64d21df952117d0be4981671e275e9733bc3a444 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Wed, 17 Apr 2019 01:30:20 +0200
Subject: [PATCH 20/26] update whats new

---
 doc/whats_new/v0.21.rst | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/doc/whats_new/v0.21.rst b/doc/whats_new/v0.21.rst
index e3d9129229096..b00ac7f39bc07 100644
--- a/doc/whats_new/v0.21.rst
+++ b/doc/whats_new/v0.21.rst
@@ -340,7 +340,9 @@ Support for Python 3.4 and below has been officially dropped.
   deterministic when trained in a multi-class setting on several threads.
   :issue:`13422` by :user:`Clément Doumouro <ClemDoum>`.
 
-- |Fix| Fixed bug in :func:`linear_model.ridge.ridge_regression` that
+- |Fix| Fixed bug in :func:`linear_model.ridge.ridge_regression`,
+  :class:`linear_model.ridge.Ridge` and
+  :class:`linear_model.ridge.ridge.RidgeClassifier` that
   caused unhandled exception for arguments ``return_intercept=True`` and
   ``solver=auto`` (default) or any other solver different from ``sag``.
   :issue:`13363` by :user:`Bartosz Telenczuk <btel>`

From 1d7ca02eea93c7963936ae6f14e53f7bccfd7562 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Wed, 17 Apr 2019 01:30:34 +0200
Subject: [PATCH 21/26] currently -> directly

---
 sklearn/linear_model/ridge.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index 2a37c1a9e6964..4fe89e1d5ea1e 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -384,7 +384,7 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
                          " 'lsqr', 'sag' or 'saga'. Got %s." % solver)
 
     if return_intercept and solver != 'sag':
-        warnings.warn("In Ridge, only 'sag' solver can currently fit the "
+        warnings.warn("In Ridge, only 'sag' solver can directly fit the "
                       "intercept. Solver has been "
                       "automatically changed into 'sag'.")
         solver = 'sag'

From 604f7d10f66247dd291f1d2b71654f349b9edd2d Mon Sep 17 00:00:00 2001
From: Alexandre Gramfort <alexandre.gramfort@m4x.org>
Date: Wed, 17 Apr 2019 13:51:26 +0200
Subject: [PATCH 22/26] use more strict test

---
 sklearn/linear_model/tests/test_ridge.py | 17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index fc4bdc417dc5e..c8c3b08abb2ba 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -855,19 +855,26 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
     X = rng.rand(1000, 3)
     true_coefs = [1, 2, 0.1]
     y = np.dot(X, true_coefs)
+    intercept = 0.
+    if return_intercept:
+        intercept = 10000.
+    y += intercept
     X_testing = arr_type(X)
 
-    target = ridge_regression(X_testing, y, 1,
+    alpha, atol, tol = 1e-3, 1e-4, 1e-6
+    target = ridge_regression(X_testing, y, alpha=alpha,
                               solver=solver,
                               sample_weight=sample_weight,
-                              return_intercept=return_intercept
+                              return_intercept=return_intercept,
+                              tol=tol,
                               )
+
     if return_intercept:
         coef, intercept = target
-        assert_allclose(coef, true_coefs, rtol=0, atol=0.03)
-        assert_allclose(intercept, 0, rtol=0, atol=0.03)
+        assert_allclose(coef, true_coefs, rtol=0, atol=atol)
+        assert_allclose(intercept, intercept, rtol=0, atol=atol)
     else:
-        assert_allclose(target, true_coefs, rtol=0, atol=0.03)
+        assert_allclose(target, true_coefs, rtol=0, atol=atol)
 
 
 def test_ridge_regression_warns_with_return_intercept():

From 79d04d6b61527cd3749b3550dbea1390df22b4c1 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Wed, 17 Apr 2019 22:35:17 +0200
Subject: [PATCH 23/26] fix test for true_intercept

---
 sklearn/linear_model/tests/test_ridge.py | 24 ++++++++++++------------
 1 file changed, 12 insertions(+), 12 deletions(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index c8c3b08abb2ba..e9c033e0b74b6 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -855,26 +855,26 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
     X = rng.rand(1000, 3)
     true_coefs = [1, 2, 0.1]
     y = np.dot(X, true_coefs)
-    intercept = 0.
+    true_intercept = 0.
     if return_intercept:
-        intercept = 10000.
-    y += intercept
+        true_intercept = 10000.
+    y += true_intercept
     X_testing = arr_type(X)
 
     alpha, atol, tol = 1e-3, 1e-4, 1e-6
-    target = ridge_regression(X_testing, y, alpha=alpha,
-                              solver=solver,
-                              sample_weight=sample_weight,
-                              return_intercept=return_intercept,
-                              tol=tol,
-                              )
+    out = ridge_regression(X_testing, y, alpha=alpha,
+                           solver=solver,
+                           sample_weight=sample_weight,
+                           return_intercept=return_intercept,
+                           tol=tol,
+                           )
 
     if return_intercept:
-        coef, intercept = target
+        coef, intercept = out
         assert_allclose(coef, true_coefs, rtol=0, atol=atol)
-        assert_allclose(intercept, intercept, rtol=0, atol=atol)
+        assert_allclose(intercept, true_intercept, rtol=0, atol=atol)
     else:
-        assert_allclose(target, true_coefs, rtol=0, atol=atol)
+        assert_allclose(out, true_coefs, rtol=0, atol=atol)
 
 
 def test_ridge_regression_warns_with_return_intercept():

From 9e1aefbb393ff32ac269ead62a4af480584a18f3 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Wed, 17 Apr 2019 22:45:25 +0200
Subject: [PATCH 24/26] changed warning to exception

---
 sklearn/linear_model/ridge.py            |  7 ++--
 sklearn/linear_model/tests/test_ridge.py | 45 +++++++++++++-----------
 2 files changed, 27 insertions(+), 25 deletions(-)

diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index 4fe89e1d5ea1e..625f534779d9d 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -384,10 +384,9 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
                          " 'lsqr', 'sag' or 'saga'. Got %s." % solver)
 
     if return_intercept and solver != 'sag':
-        warnings.warn("In Ridge, only 'sag' solver can directly fit the "
-                      "intercept. Solver has been "
-                      "automatically changed into 'sag'.")
-        solver = 'sag'
+        raise ValueError("In Ridge, only 'sag' solver can directly fit the "
+                         "intercept. Solver has been "
+                         "automatically changed into 'sag'.")
 
     _dtype = [np.float64, np.float32]
 
diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index e9c033e0b74b6..3a2497c1d0cda 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -834,9 +834,7 @@ def test_ridge_fit_intercept_sparse():
     # test the solver switch and the corresponding warning
     for solver in ['saga', 'lsqr']:
         sparse = Ridge(alpha=1., tol=1.e-15, solver=solver, fit_intercept=True)
-        assert_warns(UserWarning, sparse.fit, X_csr, y)
-        assert_almost_equal(dense.intercept_, sparse.intercept_)
-        assert_array_almost_equal(dense.coef_, sparse.coef_)
+        assert_raises_regex(ValueError, "In Ridge,", sparse.fit, X_csr, y)
 
 
 @pytest.mark.parametrize('return_intercept', [False, True])
@@ -862,6 +860,18 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
     X_testing = arr_type(X)
 
     alpha, atol, tol = 1e-3, 1e-4, 1e-6
+
+    if solver not in ['sag', 'auto'] and return_intercept:
+        assert_raises_regex(ValueError,
+                            "In Ridge, only 'sag' solver",
+                            ridge_regression, X_testing, y,
+                            alpha=alpha,
+                            solver=solver,
+                            sample_weight=sample_weight,
+                            return_intercept=return_intercept,
+                            tol=tol)
+        return
+
     out = ridge_regression(X_testing, y, alpha=alpha,
                            solver=solver,
                            sample_weight=sample_weight,
@@ -884,24 +894,17 @@ def test_ridge_regression_warns_with_return_intercept():
                            bias=10., random_state=42)
 
     for solver in ['sparse_cg', 'cholesky', 'svd', 'lsqr', 'saga']:
-        with pytest.warns(UserWarning, match="In Ridge, only 'sag' solver"):
-            target = ridge_regression(X, y, 1,
-                                      solver=solver,
-                                      return_intercept=True
-                                      )
-            assert len(target) == 2
-            assert target[0].shape == (2,)
-
-    with pytest.warns(None) as record:
-        target = ridge_regression(X, y, 1,
-                                  solver="sag",
-                                  return_intercept=True
-                                  )
-        assert len(target) == 2
-        assert target[0].shape == (2,)
-
-    # no warning should be raised
-    assert len(record) == 0
+
+
+        assert_raises_regex(ValueError,
+                            "In Ridge, only 'sag' solver",
+                            ridge_regression, X, y, 1, solver=solver,
+                            return_intercept=True)
+
+    target = ridge_regression(X, y, 1,
+                              solver="sag",
+                              return_intercept=True
+                              )
 
 def test_errors_and_values_helper():
     ridgecv = _RidgeGCV()

From d4cd2804700bd4831be4ef69ace0ad25e2822859 Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Wed, 17 Apr 2019 22:49:24 +0200
Subject: [PATCH 25/26] remove redundant test (covered by
 test_ridge_regression_check_arguments_validity)

---
 sklearn/linear_model/tests/test_ridge.py | 19 -------------------
 1 file changed, 19 deletions(-)

diff --git a/sklearn/linear_model/tests/test_ridge.py b/sklearn/linear_model/tests/test_ridge.py
index 3a2497c1d0cda..7bfb617d4beff 100644
--- a/sklearn/linear_model/tests/test_ridge.py
+++ b/sklearn/linear_model/tests/test_ridge.py
@@ -887,25 +887,6 @@ def test_ridge_regression_check_arguments_validity(return_intercept,
         assert_allclose(out, true_coefs, rtol=0, atol=atol)
 
 
-def test_ridge_regression_warns_with_return_intercept():
-    # return_itercept is only supported by sag/saga
-
-    X, y = make_regression(n_samples=1000, n_features=2, n_informative=2,
-                           bias=10., random_state=42)
-
-    for solver in ['sparse_cg', 'cholesky', 'svd', 'lsqr', 'saga']:
-
-
-        assert_raises_regex(ValueError,
-                            "In Ridge, only 'sag' solver",
-                            ridge_regression, X, y, 1, solver=solver,
-                            return_intercept=True)
-
-    target = ridge_regression(X, y, 1,
-                              solver="sag",
-                              return_intercept=True
-                              )
-
 def test_errors_and_values_helper():
     ridgecv = _RidgeGCV()
     rng = check_random_state(42)

From 4ed33b5ec2254c569fedd0b84d70c03394a1339a Mon Sep 17 00:00:00 2001
From: Bartosz Telenczuk <bartosz.telenczuk@gmail.com>
Date: Thu, 18 Apr 2019 00:46:16 +0200
Subject: [PATCH 26/26] updated exception message and added entry in whats new

---
 doc/whats_new/v0.21.rst       | 4 ++++
 sklearn/linear_model/ridge.py | 4 ++--
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/doc/whats_new/v0.21.rst b/doc/whats_new/v0.21.rst
index b00ac7f39bc07..259b943b4273d 100644
--- a/doc/whats_new/v0.21.rst
+++ b/doc/whats_new/v0.21.rst
@@ -347,6 +347,10 @@ Support for Python 3.4 and below has been officially dropped.
   ``solver=auto`` (default) or any other solver different from ``sag``.
   :issue:`13363` by :user:`Bartosz Telenczuk <btel>`
 
+- |Fix| :func:`linear_model.ridge.ridge_regression` will now raise an exception
+  if ``return_intercept=True`` and solver is different from ``sag``. Previously,
+  only warning was issued. :issue:`13363` by :user:`Bartosz Telenczuk <btel>`
+
 - |API| :func:`linear_model.ridge.ridge_regression` will choose ``sparse_cg``
   solver for sparse inputs when ``solver=auto`` and ``sample_weight``
   is provided (previously `cholesky` solver was selected). :issue:`13363`
diff --git a/sklearn/linear_model/ridge.py b/sklearn/linear_model/ridge.py
index 625f534779d9d..c9ef770135ef3 100644
--- a/sklearn/linear_model/ridge.py
+++ b/sklearn/linear_model/ridge.py
@@ -385,8 +385,8 @@ def _ridge_regression(X, y, alpha, sample_weight=None, solver='auto',
 
     if return_intercept and solver != 'sag':
         raise ValueError("In Ridge, only 'sag' solver can directly fit the "
-                         "intercept. Solver has been "
-                         "automatically changed into 'sag'.")
+                         "intercept. Please change solver to 'sag' or set "
+                         "return_intercept=False.")
 
     _dtype = [np.float64, np.float32]