scikit-learn · glemaitre · Oct 24, 2021 · Oct 23, 2021 · Oct 23, 2021 · Oct 24, 2021
diff --git a/doc/whats_new/v1.1.rst b/doc/whats_new/v1.1.rst
@@ -89,6 +89,13 @@ Changelog
   message when running in a jupyter notebook that is not trusted. :pr:`21316`
   by `Thomas Fan`_.
 
+:mod:`sklearn.neighbors`
+........................
+
+- |Fix| :class:`neighbors.KernelDensity` now validates input parameters in `fit`
+  instead of `__init__`. :pr:`21430` by :user:`Desislava Vasileva <DessyVV>` and
+  :user:`Lucy Jimenez <LucyJimenez>`.
+
 Code and Documentation Contributors
 -----------------------------------
 

diff --git a/sklearn/neighbors/_kde.py b/sklearn/neighbors/_kde.py
@@ -135,16 +135,6 @@ def __init__(
         self.leaf_size = leaf_size
         self.metric_params = metric_params
 
-        # run the choose algorithm code so that exceptions will happen here
-        # we're using clone() in the GenerativeBayes classifier,
-        # so we can't do this kind of logic in __init__
-        self._choose_algorithm(self.algorithm, self.metric)
-
-        if bandwidth <= 0:
-            raise ValueError("bandwidth must be positive")
-        if kernel not in VALID_KERNELS:
-            raise ValueError("invalid kernel: '{0}'".format(kernel))
-
     def _choose_algorithm(self, algorithm, metric):
         # given the algorithm string + metric string, choose the optimal
         # algorithm to compute the result.
@@ -188,7 +178,14 @@ def fit(self, X, y=None, sample_weight=None):
         self : object
             Returns the instance itself.
         """
+
         algorithm = self._choose_algorithm(self.algorithm, self.metric)
+
+        if self.bandwidth <= 0:
+            raise ValueError("bandwidth must be positive")
+        if self.kernel not in VALID_KERNELS:
+            raise ValueError("invalid kernel: '{0}'".format(self.kernel))
+
         X = self._validate_data(X, order="C", dtype=DTYPE)
 
         if sample_weight is not None:

diff --git a/sklearn/neighbors/tests/test_kde.py b/sklearn/neighbors/tests/test_kde.py
@@ -107,11 +107,12 @@ def test_kde_algorithm_metric_choice(algorithm, metric):
     X = rng.randn(10, 2)  # 2 features required for haversine dist.
     Y = rng.randn(10, 2)
 
+    kde = KernelDensity(algorithm=algorithm, metric=metric)
+
     if algorithm == "kd_tree" and metric not in KDTree.valid_metrics:
         with pytest.raises(ValueError):
-            KernelDensity(algorithm=algorithm, metric=metric)
+            kde.fit(X)
     else:
-        kde = KernelDensity(algorithm=algorithm, metric=metric)
         kde.fit(X)
         y_dens = kde.score_samples(Y)
         assert y_dens.shape == Y.shape[:1]
@@ -126,16 +127,17 @@ def test_kde_score(n_samples=100, n_features=3):
 
 
 def test_kde_badargs():
+    X = np.random.random((200, 10))
     with pytest.raises(ValueError):
-        KernelDensity(algorithm="blah")
+        KernelDensity(algorithm="blah").fit(X)
     with pytest.raises(ValueError):
-        KernelDensity(bandwidth=0)
+        KernelDensity(bandwidth=0).fit(X)
     with pytest.raises(ValueError):
-        KernelDensity(kernel="blah")
+        KernelDensity(kernel="blah").fit(X)
     with pytest.raises(ValueError):
-        KernelDensity(metric="blah")
+        KernelDensity(metric="blah").fit(X)
     with pytest.raises(ValueError):
-        KernelDensity(algorithm="kd_tree", metric="blah")
+        KernelDensity(algorithm="kd_tree", metric="blah").fit(X)
     kde = KernelDensity()
     with pytest.raises(ValueError):
         kde.fit(np.random.random((200, 10)), sample_weight=np.random.random((200, 10)))

diff --git a/sklearn/tests/test_common.py b/sklearn/tests/test_common.py
@@ -413,7 +413,6 @@ def test_transformers_get_feature_names_out(transformer):
     "FeatureUnion",
     "GridSearchCV",
     "HalvingGridSearchCV",
-    "KernelDensity",
     "KernelPCA",
     "LabelBinarizer",
     "NuSVC",