scikit-learn-intelex 2025.0.1__py311-none-win_amd64.whl → 2025.2.0__py311-none-win_amd64.whl

scikit_learn_intelex-2025.2.0.data/data/Lib/site-packages/sklearnex/manifold/tests/test_tsne.py

@@ -0,0 +1,250 @@
+# ===============================================================================
+# Copyright 2021 Intel Corporation
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ===============================================================================
+
+import numpy as np
+import pytest
+from numpy.testing import assert_allclose
+from sklearn.metrics.pairwise import pairwise_distances
+
+# Note: n_components must be 2 for now
+from onedal.tests.utils._dataframes_support import (
+    _as_numpy,
+    _convert_to_dataframe,
+    get_dataframes_and_queues,
+)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+def test_sklearnex_import(dataframe, queue):
+    """Test TSNE compatibility with different backends and queues, and validate sklearnex module."""
+    from sklearnex.manifold import TSNE
+
+    X = np.array([[0, 0, 0], [0, 1, 1], [1, 0, 1], [1, 1, 1]])
+    X_df = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
+    tsne = TSNE(n_components=2, perplexity=2.0, random_state=42, init="pca").fit(X_df)
+    embedding = tsne.fit_transform(X_df)
+    embedding = _as_numpy(embedding)
+    assert "daal4py" in tsne.__module__
+    assert tsne.n_components == 2
+    assert tsne.perplexity == 2.0
+    assert tsne.random_state == 42
+    assert tsne.init == "pca"
+
+
+@pytest.mark.parametrize(
+    "X_generator,n_components,perplexity,expected_shape,should_raise",
+    [
+        pytest.param(
+            lambda rng: np.array([[0, 0, 0], [0, 1, 1], [1, 0, 1], [1, 1, 1]]),
+            2,
+            2.0,
+            (4, 2),
+            False,
+            id="Basic functionality",
+        ),
+        pytest.param(
+            lambda rng: rng.random((100, 10)),
+            2,
+            30.0,
+            (100, 2),
+            False,
+            id="Random data",
+        ),
+        pytest.param(
+            lambda rng: np.array([[0, 0], [1, 1], [2, 2]]),
+            2,
+            2.0,
+            (3, 2),
+            False,
+            id="Valid minimal data",
+        ),
+        pytest.param(
+            lambda rng: np.empty((0, 10)),
+            2,
+            5.0,
+            None,
+            True,
+            id="Empty data",
+        ),
+        pytest.param(
+            lambda rng: np.array([[0, 0], [1, np.nan], [2, np.inf]]),
+            2,
+            5.0,
+            None,
+            True,
+            id="Data with NaN/Inf",
+        ),
+        pytest.param(
+            lambda rng: rng.random((50, 500)) * (rng.random((50, 500)) > 0.99),
+            2,
+            30.0,
+            (50, 2),
+            False,
+            id="Sparse-like high-dimensional data",
+        ),
+        pytest.param(
+            lambda rng: np.hstack(
+                [
+                    np.ones((50, 1)),  # First column is 1
+                    rng.random((50, 499)) * (rng.random((50, 499)) > 0.99),
+                ]
+            ),
+            2,
+            30.0,
+            (50, 2),
+            False,
+            id="Sparse-like data with constant column",
+        ),
+        pytest.param(
+            lambda rng: np.where(
+                np.arange(50 * 500).reshape(50, 500) % 10 == 0, 0, rng.random((50, 500))
+            ),
+            2,
+            30.0,
+            (50, 2),
+            False,
+            id="Sparse-like data with every tenth element zero",
+        ),
+        pytest.param(
+            lambda rng: rng.random((10, 5)),
+            2,
+            0.5,
+            (10, 2),
+            False,
+            id="Extremely low perplexity",
+        ),
+    ],
+)
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_tsne_functionality_and_edge_cases(
+    X_generator,
+    n_components,
+    perplexity,
+    expected_shape,
+    should_raise,
+    dataframe,
+    queue,
+    dtype,
+):
+    from sklearnex.manifold import TSNE
+
+    rng = np.random.default_rng(
+        seed=42
+    )  # Use generator to ensure independent dataset per test
+    X = X_generator(rng)
+    X = X.astype(dtype) if X.size > 0 else X
+    X_df = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
+
+    if should_raise:
+        with pytest.raises(ValueError):
+            TSNE(n_components=n_components, perplexity=perplexity).fit_transform(X_df)
+    else:
+        tsne = TSNE(n_components=n_components, perplexity=perplexity, random_state=42)
+        embedding = tsne.fit_transform(X_df)
+        embedding = _as_numpy(embedding)
+        assert embedding.shape == expected_shape
+        assert np.all(np.isfinite(embedding))
+        assert np.any(embedding != 0)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("init", ["pca", "random"])
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_tsne_constant_data(init, dataframe, queue, dtype):
+    from sklearnex.manifold import TSNE
+
+    X = np.ones((10, 10), dtype=dtype)
+    X_df = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
+    tsne = TSNE(n_components=2, init=init, perplexity=5, random_state=42)
+    embedding = tsne.fit_transform(X_df)
+    embedding = _as_numpy(embedding)
+    assert embedding.shape == (10, 2)
+    if init == "pca":
+        assert np.isclose(embedding[:, 0].std(), 0, atol=1e-6)  # Constant first dimension
+        assert np.allclose(embedding[:, 1], 0, atol=1e-6)  # Zero second dimension
+    elif init == "random":
+        assert np.all(np.isfinite(embedding))
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_tsne_reproducibility(dataframe, queue, dtype):
+    from sklearnex.manifold import TSNE
+
+    rng = np.random.default_rng(seed=42)
+    X = rng.random((50, 10)).astype(dtype)
+    X_df = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
+    tsne_1 = TSNE(n_components=2, random_state=42).fit_transform(X_df)
+    tsne_2 = TSNE(n_components=2, random_state=42).fit_transform(X_df)
+    # in case of dpctl.tensor.usm_ndarray covert to numpy array
+    tsne_1 = _as_numpy(tsne_1)
+    tsne_2 = _as_numpy(tsne_2)
+    assert_allclose(tsne_1, tsne_2, rtol=1e-5)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_tsne_complex_and_gpu_validation(dataframe, queue, dtype):
+    from sklearnex.manifold import TSNE
+
+    X = np.array(
+        [
+            [1, 1, 1, 1],
+            [1.1, 1.1, 1.1, 1.1],
+            [0.9, 0.9, 0.9, 0.9],
+            [2e9, 2e-9, -2e9, -2e-9],
+            [5e-5, 5e5, -5e-5, -5e5],
+            [9e-7, -9e7, 9e-7, -9e7],
+            [1, -1, 1, -1],
+            [-1e-9, 1e-9, -1e-9, 1e-9],
+            [42, 42, 42, 42],
+            [8, -8, 8e8, -8e-8],
+            [1e-3, 1e3, -1e3, -1e-3],
+            [0, 1e9, -1e-9, 1],
+            [0, 0, 1, -1],
+            [0, 0, 0, 0],
+            [-1e5, 0, 1e5, -1],
+            [1, 0, -1e8, 1e8],
+        ]
+    )
+    n_components = 2
+    perplexity = 3.0
+    expected_shape = (16, 2)
+
+    X = X.astype(dtype)
+    X_df = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
+    tsne = TSNE(n_components=n_components, perplexity=perplexity, random_state=42)
+    embedding = tsne.fit_transform(X_df)
+
+    # Validate results
+    assert embedding.shape == expected_shape
+    embedding = _as_numpy(embedding)
+    assert np.all(np.isfinite(embedding))
+    assert np.any(embedding != 0)
+
+    # Ensure close points in original space remain close in embedding
+    group_a_indices = [0, 1, 2]  # Hardcoded index of similar points
+    group_b_indices = [3, 4, 5]  # Hardcoded index of dissimilar points from a
+    embedding_distances = pairwise_distances(
+        X, metric="euclidean"
+    )  # Get an array of distance where [i, j] is distance b/t i and j
+    # Check for distance b/t two points in group A < distance of this point and any point in group B
+    for i in group_a_indices:
+        for j in group_a_indices:
+            assert (
+                embedding_distances[i, j] < embedding_distances[i, group_b_indices].min()
+            ), f"Point {i} in Group A is closer to a point in Group B than to another point in Group A."

{scikit_learn_intelex-2025.0.1.data → scikit_learn_intelex-2025.2.0.data}/data/Lib/site-packages/sklearnex/neighbors/_lof.py

@@ -17,7 +17,7 @@
 import warnings
 
 import numpy as np
-from sklearn.neighbors import LocalOutlierFactor as sklearn_LocalOutlierFactor
+from sklearn.neighbors import LocalOutlierFactor as _sklearn_LocalOutlierFactor
 from sklearn.utils.metaestimators import available_if
 from sklearn.utils.validation import check_is_fitted
 
@@ -29,18 +29,23 @@ from sklearnex.neighbors.knn_unsupervised import NearestNeighbors
 
 from ..utils._array_api import get_namespace
 
+if sklearn_check_version("1.6"):
+    from sklearn.utils.validation import validate_data
+else:
+    validate_data = _sklearn_LocalOutlierFactor._validate_data
 
-@control_n_jobs(decorated_methods=["fit", "_kneighbors"])
-class LocalOutlierFactor(KNeighborsDispatchingBase, sklearn_LocalOutlierFactor):
+
+@control_n_jobs(decorated_methods=["fit", "kneighbors", "_kneighbors"])
+class LocalOutlierFactor(KNeighborsDispatchingBase, _sklearn_LocalOutlierFactor):
     __doc__ = (
-        sklearn_LocalOutlierFactor.__doc__
+        _sklearn_LocalOutlierFactor.__doc__
         + "\n NOTE: When X=None, methods kneighbors, kneighbors_graph, and predict will"
         + "\n only output numpy arrays. In that case, the only way to offload to gpu"
         + "\n is to use a global queue (e.g. using config_context)"
     )
     if sklearn_check_version("1.2"):
         _parameter_constraints: dict = {
-            **sklearn_LocalOutlierFactor._parameter_constraints
+            **_sklearn_LocalOutlierFactor._parameter_constraints
         }
 
     # Only certain methods should be taken from knn to prevent code
@@ -115,7 +120,7 @@ class LocalOutlierFactor(KNeighborsDispatchingBase, sklearn_LocalOutlierFactor):
             "fit",
             {
                 "onedal": self.__class__._onedal_fit,
-                "sklearn": sklearn_LocalOutlierFactor.fit,
+                "sklearn": _sklearn_LocalOutlierFactor.fit,
             },
             X,
             None,
@@ -140,7 +145,7 @@ class LocalOutlierFactor(KNeighborsDispatchingBase, sklearn_LocalOutlierFactor):
     # argument is given and it is a dpctl tensor or dpnp array.
     # This would cause issues in fit_predict. Also, available_if
     # is hard to unwrap, and this is the most straighforward way.
-    @available_if(sklearn_LocalOutlierFactor._check_novelty_fit_predict)
+    @available_if(_sklearn_LocalOutlierFactor._check_novelty_fit_predict)
     @wrap_output_data
     def fit_predict(self, X, y=None):
         """Fit the model to the training set X and return the labels.
@@ -174,7 +179,7 @@ class LocalOutlierFactor(KNeighborsDispatchingBase, sklearn_LocalOutlierFactor):
             "kneighbors",
             {
                 "onedal": self.__class__._onedal_kneighbors,
-                "sklearn": sklearn_LocalOutlierFactor.kneighbors,
+                "sklearn": _sklearn_LocalOutlierFactor.kneighbors,
             },
             X,
             n_neighbors=n_neighbors,
@@ -183,7 +188,7 @@ class LocalOutlierFactor(KNeighborsDispatchingBase, sklearn_LocalOutlierFactor):
 
     kneighbors = wrap_output_data(_kneighbors)
 
-    @available_if(sklearn_LocalOutlierFactor._check_novelty_score_samples)
+    @available_if(_sklearn_LocalOutlierFactor._check_novelty_score_samples)
     @wrap_output_data
     def score_samples(self, X):
         """Opposite of the Local Outlier Factor of X.
@@ -227,5 +232,5 @@ class LocalOutlierFactor(KNeighborsDispatchingBase, sklearn_LocalOutlierFactor):
 
         return -np.mean(lrd_ratios_array, axis=1)
 
-    fit.__doc__ = sklearn_LocalOutlierFactor.fit.__doc__
-    kneighbors.__doc__ = sklearn_LocalOutlierFactor.kneighbors.__doc__
+    fit.__doc__ = _sklearn_LocalOutlierFactor.fit.__doc__
+    kneighbors.__doc__ = _sklearn_LocalOutlierFactor.kneighbors.__doc__

{scikit_learn_intelex-2025.0.1.data → scikit_learn_intelex-2025.2.0.data}/data/Lib/site-packages/sklearnex/neighbors/common.py

@@ -20,7 +20,7 @@ import numpy as np
 from scipy import sparse as sp
 from sklearn.neighbors._ball_tree import BallTree
 from sklearn.neighbors._base import VALID_METRICS, KNeighborsMixin
-from sklearn.neighbors._base import NeighborsBase as sklearn_NeighborsBase
+from sklearn.neighbors._base import NeighborsBase as _sklearn_NeighborsBase
 from sklearn.neighbors._kd_tree import KDTree
 from sklearn.utils.validation import check_is_fitted
 
@@ -64,7 +64,7 @@ class KNeighborsDispatchingBase:
             elif p == np.inf:
                 self.effective_metric_ = "chebyshev"
 
-        if not isinstance(X, (KDTree, BallTree, sklearn_NeighborsBase)):
+        if not isinstance(X, (KDTree, BallTree, _sklearn_NeighborsBase)):
             self._fit_X = _check_array(
                 X, dtype=[np.float64, np.float32], accept_sparse=True
             )
@@ -97,7 +97,7 @@ class KNeighborsDispatchingBase:
             delattr(self, "_onedal_estimator")
         # To cover test case when we pass patched
         # estimator as an input for other estimator
-        if isinstance(X, sklearn_NeighborsBase):
+        if isinstance(X, _sklearn_NeighborsBase):
             self._fit_X = X._fit_X
             self._tree = X._tree
             self._fit_method = X._fit_method
@@ -155,7 +155,7 @@ class KNeighborsDispatchingBase:
             return patching_status
 
         if not patching_status.and_condition(
-            not isinstance(data[0], (KDTree, BallTree, sklearn_NeighborsBase)),
+            not isinstance(data[0], (KDTree, BallTree, _sklearn_NeighborsBase)),
             f"Input type {type(data[0])} is not supported.",
         ):
             return patching_status

{scikit_learn_intelex-2025.0.1.data → scikit_learn_intelex-2025.2.0.data}/data/Lib/site-packages/sklearnex/neighbors/knn_classification.py

@@ -16,9 +16,9 @@
 
 from sklearn.metrics import accuracy_score
 from sklearn.neighbors._classification import (
-    KNeighborsClassifier as sklearn_KNeighborsClassifier,
+    KNeighborsClassifier as _sklearn_KNeighborsClassifier,
 )
-from sklearn.neighbors._unsupervised import NearestNeighbors as sklearn_NearestNeighbors
+from sklearn.neighbors._unsupervised import NearestNeighbors as _sklearn_NearestNeighbors
 from sklearn.utils.validation import _deprecate_positional_args, check_is_fitted
 
 from daal4py.sklearn._n_jobs_support import control_n_jobs
@@ -28,15 +28,20 @@ from onedal.neighbors import KNeighborsClassifier as onedal_KNeighborsClassifier
 from .._device_offload import dispatch, wrap_output_data
 from .common import KNeighborsDispatchingBase
 
+if sklearn_check_version("1.6"):
+    from sklearn.utils.validation import validate_data
+else:
+    validate_data = _sklearn_KNeighborsClassifier._validate_data
+
 
 @control_n_jobs(
     decorated_methods=["fit", "predict", "predict_proba", "kneighbors", "score"]
 )
-class KNeighborsClassifier(KNeighborsDispatchingBase, sklearn_KNeighborsClassifier):
-    __doc__ = sklearn_KNeighborsClassifier.__doc__
+class KNeighborsClassifier(KNeighborsDispatchingBase, _sklearn_KNeighborsClassifier):
+    __doc__ = _sklearn_KNeighborsClassifier.__doc__
     if sklearn_check_version("1.2"):
         _parameter_constraints: dict = {
-            **sklearn_KNeighborsClassifier._parameter_constraints
+            **_sklearn_KNeighborsClassifier._parameter_constraints
         }
 
     if sklearn_check_version("1.0"):
@@ -98,7 +103,7 @@ class KNeighborsClassifier(KNeighborsDispatchingBase, sklearn_KNeighborsClassifi
             "fit",
             {
                 "onedal": self.__class__._onedal_fit,
-                "sklearn": sklearn_KNeighborsClassifier.fit,
+                "sklearn": _sklearn_KNeighborsClassifier.fit,
             },
             X,
             y,
@@ -115,7 +120,7 @@ class KNeighborsClassifier(KNeighborsDispatchingBase, sklearn_KNeighborsClassifi
             "predict",
             {
                 "onedal": self.__class__._onedal_predict,
-                "sklearn": sklearn_KNeighborsClassifier.predict,
+                "sklearn": _sklearn_KNeighborsClassifier.predict,
             },
             X,
         )
@@ -130,7 +135,7 @@ class KNeighborsClassifier(KNeighborsDispatchingBase, sklearn_KNeighborsClassifi
             "predict_proba",
             {
                 "onedal": self.__class__._onedal_predict_proba,
-                "sklearn": sklearn_KNeighborsClassifier.predict_proba,
+                "sklearn": _sklearn_KNeighborsClassifier.predict_proba,
             },
             X,
         )
@@ -145,7 +150,7 @@ class KNeighborsClassifier(KNeighborsDispatchingBase, sklearn_KNeighborsClassifi
             "score",
             {
                 "onedal": self.__class__._onedal_score,
-                "sklearn": sklearn_KNeighborsClassifier.score,
+                "sklearn": _sklearn_KNeighborsClassifier.score,
             },
             X,
             y,
@@ -162,7 +167,7 @@ class KNeighborsClassifier(KNeighborsDispatchingBase, sklearn_KNeighborsClassifi
             "kneighbors",
             {
                 "onedal": self.__class__._onedal_kneighbors,
-                "sklearn": sklearn_KNeighborsClassifier.kneighbors,
+                "sklearn": _sklearn_KNeighborsClassifier.kneighbors,
             },
             X,
             n_neighbors=n_neighbors,
@@ -219,8 +224,8 @@ class KNeighborsClassifier(KNeighborsDispatchingBase, sklearn_KNeighborsClassifi
         self.outputs_2d_ = self._onedal_estimator.outputs_2d_
         self._tree = self._onedal_estimator._tree
 
-    fit.__doc__ = sklearn_KNeighborsClassifier.fit.__doc__
-    predict.__doc__ = sklearn_KNeighborsClassifier.predict.__doc__
-    predict_proba.__doc__ = sklearn_KNeighborsClassifier.predict_proba.__doc__
-    score.__doc__ = sklearn_KNeighborsClassifier.score.__doc__
-    kneighbors.__doc__ = sklearn_KNeighborsClassifier.kneighbors.__doc__
+    fit.__doc__ = _sklearn_KNeighborsClassifier.fit.__doc__
+    predict.__doc__ = _sklearn_KNeighborsClassifier.predict.__doc__
+    predict_proba.__doc__ = _sklearn_KNeighborsClassifier.predict_proba.__doc__
+    score.__doc__ = _sklearn_KNeighborsClassifier.score.__doc__
+    kneighbors.__doc__ = _sklearn_KNeighborsClassifier.kneighbors.__doc__

{scikit_learn_intelex-2025.0.1.data → scikit_learn_intelex-2025.2.0.data}/data/Lib/site-packages/sklearnex/neighbors/knn_regression.py

@@ -16,9 +16,8 @@
 
 from sklearn.metrics import r2_score
 from sklearn.neighbors._regression import (
-    KNeighborsRegressor as sklearn_KNeighborsRegressor,
+    KNeighborsRegressor as _sklearn_KNeighborsRegressor,
 )
-from sklearn.neighbors._unsupervised import NearestNeighbors as sklearn_NearestNeighbors
 from sklearn.utils.validation import _deprecate_positional_args, check_is_fitted
 
 from daal4py.sklearn._n_jobs_support import control_n_jobs
@@ -28,13 +27,18 @@ from onedal.neighbors import KNeighborsRegressor as onedal_KNeighborsRegressor
 from .._device_offload import dispatch, wrap_output_data
 from .common import KNeighborsDispatchingBase
 
+if sklearn_check_version("1.6"):
+    from sklearn.utils.validation import validate_data
+else:
+    validate_data = _sklearn_KNeighborsRegressor._validate_data
 
-@control_n_jobs(decorated_methods=["fit", "predict", "kneighbors"])
-class KNeighborsRegressor(KNeighborsDispatchingBase, sklearn_KNeighborsRegressor):
-    __doc__ = sklearn_KNeighborsRegressor.__doc__
+
+@control_n_jobs(decorated_methods=["fit", "predict", "kneighbors", "score"])
+class KNeighborsRegressor(KNeighborsDispatchingBase, _sklearn_KNeighborsRegressor):
+    __doc__ = _sklearn_KNeighborsRegressor.__doc__
     if sklearn_check_version("1.2"):
         _parameter_constraints: dict = {
-            **sklearn_KNeighborsRegressor._parameter_constraints
+            **_sklearn_KNeighborsRegressor._parameter_constraints
         }
 
     if sklearn_check_version("1.0"):
@@ -96,7 +100,7 @@ class KNeighborsRegressor(KNeighborsDispatchingBase, sklearn_KNeighborsRegressor
             "fit",
             {
                 "onedal": self.__class__._onedal_fit,
-                "sklearn": sklearn_KNeighborsRegressor.fit,
+                "sklearn": _sklearn_KNeighborsRegressor.fit,
             },
             X,
             y,
@@ -113,7 +117,7 @@ class KNeighborsRegressor(KNeighborsDispatchingBase, sklearn_KNeighborsRegressor
             "predict",
             {
                 "onedal": self.__class__._onedal_predict,
-                "sklearn": sklearn_KNeighborsRegressor.predict,
+                "sklearn": _sklearn_KNeighborsRegressor.predict,
             },
             X,
         )
@@ -128,7 +132,7 @@ class KNeighborsRegressor(KNeighborsDispatchingBase, sklearn_KNeighborsRegressor
             "score",
             {
                 "onedal": self.__class__._onedal_score,
-                "sklearn": sklearn_KNeighborsRegressor.score,
+                "sklearn": _sklearn_KNeighborsRegressor.score,
             },
             X,
             y,
@@ -145,7 +149,7 @@ class KNeighborsRegressor(KNeighborsDispatchingBase, sklearn_KNeighborsRegressor
             "kneighbors",
             {
                 "onedal": self.__class__._onedal_kneighbors,
-                "sklearn": sklearn_KNeighborsRegressor.kneighbors,
+                "sklearn": _sklearn_KNeighborsRegressor.kneighbors,
             },
             X,
             n_neighbors=n_neighbors,
@@ -197,7 +201,7 @@ class KNeighborsRegressor(KNeighborsDispatchingBase, sklearn_KNeighborsRegressor
         self._fit_method = self._onedal_estimator._fit_method
         self._tree = self._onedal_estimator._tree
 
-    fit.__doc__ = sklearn_KNeighborsRegressor.__doc__
-    predict.__doc__ = sklearn_KNeighborsRegressor.predict.__doc__
-    kneighbors.__doc__ = sklearn_KNeighborsRegressor.kneighbors.__doc__
-    score.__doc__ = sklearn_KNeighborsRegressor.score.__doc__
+    fit.__doc__ = _sklearn_KNeighborsRegressor.__doc__
+    predict.__doc__ = _sklearn_KNeighborsRegressor.predict.__doc__
+    kneighbors.__doc__ = _sklearn_KNeighborsRegressor.kneighbors.__doc__
+    score.__doc__ = _sklearn_KNeighborsRegressor.score.__doc__

{scikit_learn_intelex-2025.0.1.data → scikit_learn_intelex-2025.2.0.data}/data/Lib/site-packages/sklearnex/neighbors/knn_unsupervised.py

@@ -14,7 +14,7 @@
 # limitations under the License.
 # ===============================================================================
 
-from sklearn.neighbors._unsupervised import NearestNeighbors as sklearn_NearestNeighbors
+from sklearn.neighbors._unsupervised import NearestNeighbors as _sklearn_NearestNeighbors
 from sklearn.utils.validation import _deprecate_positional_args, check_is_fitted
 
 from daal4py.sklearn._n_jobs_support import control_n_jobs
@@ -24,12 +24,19 @@ from onedal.neighbors import NearestNeighbors as onedal_NearestNeighbors
 from .._device_offload import dispatch, wrap_output_data
 from .common import KNeighborsDispatchingBase
 
+if sklearn_check_version("1.6"):
+    from sklearn.utils.validation import validate_data
+else:
+    validate_data = _sklearn_NearestNeighbors._validate_data
 
-@control_n_jobs(decorated_methods=["fit", "kneighbors"])
-class NearestNeighbors(KNeighborsDispatchingBase, sklearn_NearestNeighbors):
-    __doc__ = sklearn_NearestNeighbors.__doc__
+
+@control_n_jobs(decorated_methods=["fit", "kneighbors", "radius_neighbors"])
+class NearestNeighbors(KNeighborsDispatchingBase, _sklearn_NearestNeighbors):
+    __doc__ = _sklearn_NearestNeighbors.__doc__
     if sklearn_check_version("1.2"):
-        _parameter_constraints: dict = {**sklearn_NearestNeighbors._parameter_constraints}
+        _parameter_constraints: dict = {
+            **_sklearn_NearestNeighbors._parameter_constraints
+        }
 
     @_deprecate_positional_args
     def __init__(
@@ -60,7 +67,7 @@ class NearestNeighbors(KNeighborsDispatchingBase, sklearn_NearestNeighbors):
             "fit",
             {
                 "onedal": self.__class__._onedal_fit,
-                "sklearn": sklearn_NearestNeighbors.fit,
+                "sklearn": _sklearn_NearestNeighbors.fit,
             },
             X,
             None,
@@ -77,7 +84,7 @@ class NearestNeighbors(KNeighborsDispatchingBase, sklearn_NearestNeighbors):
             "kneighbors",
             {
                 "onedal": self.__class__._onedal_kneighbors,
-                "sklearn": sklearn_NearestNeighbors.kneighbors,
+                "sklearn": _sklearn_NearestNeighbors.kneighbors,
             },
             X,
             n_neighbors=n_neighbors,
@@ -93,13 +100,14 @@ class NearestNeighbors(KNeighborsDispatchingBase, sklearn_NearestNeighbors):
             or getattr(self, "_tree", 0) is None
             and self._fit_method == "kd_tree"
         ):
-            sklearn_NearestNeighbors.fit(self, self._fit_X, getattr(self, "_y", None))
+            _sklearn_NearestNeighbors.fit(self, self._fit_X, getattr(self, "_y", None))
+        check_is_fitted(self)
         return dispatch(
             self,
             "radius_neighbors",
             {
                 "onedal": None,
-                "sklearn": sklearn_NearestNeighbors.radius_neighbors,
+                "sklearn": _sklearn_NearestNeighbors.radius_neighbors,
             },
             X,
             radius=radius,
@@ -115,7 +123,7 @@ class NearestNeighbors(KNeighborsDispatchingBase, sklearn_NearestNeighbors):
             "radius_neighbors_graph",
             {
                 "onedal": None,
-                "sklearn": sklearn_NearestNeighbors.radius_neighbors_graph,
+                "sklearn": _sklearn_NearestNeighbors.radius_neighbors_graph,
             },
             X,
             radius=radius,
@@ -162,9 +170,9 @@ class NearestNeighbors(KNeighborsDispatchingBase, sklearn_NearestNeighbors):
         self._fit_method = self._onedal_estimator._fit_method
         self._tree = self._onedal_estimator._tree
 
-    fit.__doc__ = sklearn_NearestNeighbors.__doc__
-    kneighbors.__doc__ = sklearn_NearestNeighbors.kneighbors.__doc__
-    radius_neighbors.__doc__ = sklearn_NearestNeighbors.radius_neighbors.__doc__
+    fit.__doc__ = _sklearn_NearestNeighbors.__doc__
+    kneighbors.__doc__ = _sklearn_NearestNeighbors.kneighbors.__doc__
+    radius_neighbors.__doc__ = _sklearn_NearestNeighbors.radius_neighbors.__doc__
     radius_neighbors_graph.__doc__ = (
-        sklearn_NearestNeighbors.radius_neighbors_graph.__doc__
+        _sklearn_NearestNeighbors.radius_neighbors_graph.__doc__
     )

{scikit_learn_intelex-2025.0.1.data → scikit_learn_intelex-2025.2.0.data}/data/Lib/site-packages/sklearnex/neighbors/tests/test_neighbors.py

@@ -52,20 +52,22 @@ def test_sklearnex_import_knn_regression(dataframe, queue):
     assert_allclose(pred, 0.5)
 
 
+@pytest.mark.parametrize("algorithm", ["auto", "brute"])
 @pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
 @pytest.mark.parametrize(
     "estimator",
     [LocalOutlierFactor, NearestNeighbors],
 )
-def test_sklearnex_kneighbors(estimator, dataframe, queue):
+def test_sklearnex_kneighbors(algorithm, estimator, dataframe, queue):
     X = [[0, 0, 2], [1, 0, 0], [0, 0, 1]]
     X = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
     test = _convert_to_dataframe([[0, 0, 1.3]], sycl_queue=queue, target_df=dataframe)
-    neigh = estimator(n_neighbors=2).fit(X)
+    neigh = estimator(n_neighbors=2, algorithm=algorithm).fit(X)
     result = neigh.kneighbors(test, 2, return_distance=False)
     result = _as_numpy(result)
     assert "sklearnex" in neigh.__module__
     assert_allclose(result, [[2, 0]])
+    result = neigh.kneighbors()
 
 
 @pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())

{scikit_learn_intelex-2025.0.1.data → scikit_learn_intelex-2025.2.0.data}/data/Lib/site-packages/sklearnex/preview/__init__.py

@@ -14,4 +14,4 @@
 # limitations under the License.
 # ==============================================================================
 
-__all__ = ["covariance", "decomposition", "linear_model"]
+__all__ = ["covariance", "decomposition"]