scikit-learn-intelex 2025.1.0__py39-none-manylinux_2_28_x86_64.whl

sklearnex/basic_statistics/tests/test_incremental_basic_statistics.py

@@ -0,0 +1,404 @@
+# ===============================================================================
+# Copyright 2024 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 daal4py.sklearn._utils import daal_check_version
+from onedal.basic_statistics.tests.test_basic_statistics import (
+    expected_max,
+    expected_mean,
+    expected_sum,
+    options_and_tests,
+)
+from onedal.tests.utils._dataframes_support import (
+    _convert_to_dataframe,
+    get_dataframes_and_queues,
+)
+from sklearnex.basic_statistics import IncrementalBasicStatistics
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("weighted", [True, False])
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_partial_fit_multiple_options_on_gold_data(dataframe, queue, weighted, dtype):
+    X = np.array([[0, 0], [1, 1]])
+    X = X.astype(dtype=dtype)
+    X_split = np.array_split(X, 2)
+    if weighted:
+        weights = np.array([1, 0.5])
+        weights = weights.astype(dtype=dtype)
+        weights_split = np.array_split(weights, 2)
+
+    incbs = IncrementalBasicStatistics()
+    for i in range(2):
+        X_split_df = _convert_to_dataframe(
+            X_split[i], sycl_queue=queue, target_df=dataframe
+        )
+        if weighted:
+            weights_split_df = _convert_to_dataframe(
+                weights_split[i], sycl_queue=queue, target_df=dataframe
+            )
+            result = incbs.partial_fit(X_split_df, sample_weight=weights_split_df)
+        else:
+            result = incbs.partial_fit(X_split_df)
+
+    if weighted:
+        expected_weighted_mean = np.array([0.25, 0.25])
+        expected_weighted_min = np.array([0, 0])
+        expected_weighted_max = np.array([0.5, 0.5])
+        assert_allclose(expected_weighted_mean, result.mean)
+        assert_allclose(expected_weighted_max, result.max)
+        assert_allclose(expected_weighted_min, result.min)
+    else:
+        expected_mean = np.array([0.5, 0.5])
+        expected_min = np.array([0, 0])
+        expected_max = np.array([1, 1])
+        assert_allclose(expected_mean, result.mean)
+        assert_allclose(expected_max, result.max)
+        assert_allclose(expected_min, result.min)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("num_batches", [2, 10])
+@pytest.mark.parametrize("option", options_and_tests)
+@pytest.mark.parametrize("row_count", [100, 1000])
+@pytest.mark.parametrize("column_count", [10, 100])
+@pytest.mark.parametrize("weighted", [True, False])
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_partial_fit_single_option_on_random_data(
+    dataframe, queue, num_batches, option, row_count, column_count, weighted, dtype
+):
+    result_option, function, tols = option
+    fp32tol, fp64tol = tols
+    seed = 77
+    gen = np.random.default_rng(seed)
+    X = gen.uniform(low=-0.3, high=+0.7, size=(row_count, column_count))
+    X = X.astype(dtype=dtype)
+    X_split = np.array_split(X, num_batches)
+    if weighted:
+        weights = gen.uniform(low=-0.5, high=+1.0, size=row_count)
+        weights = weights.astype(dtype=dtype)
+        weights_split = np.array_split(weights, num_batches)
+    incbs = IncrementalBasicStatistics(result_options=result_option)
+
+    for i in range(num_batches):
+        X_split_df = _convert_to_dataframe(
+            X_split[i], sycl_queue=queue, target_df=dataframe
+        )
+        if weighted:
+            weights_split_df = _convert_to_dataframe(
+                weights_split[i], sycl_queue=queue, target_df=dataframe
+            )
+            result = incbs.partial_fit(X_split_df, sample_weight=weights_split_df)
+        else:
+            result = incbs.partial_fit(X_split_df)
+
+    res = getattr(result, result_option)
+    if weighted:
+        weighted_data = np.diag(weights) @ X
+        gtr = function(weighted_data)
+    else:
+        gtr = function(X)
+
+    tol = fp32tol if res.dtype == np.float32 else fp64tol
+    assert_allclose(gtr, res, atol=tol)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("num_batches", [2, 10])
+@pytest.mark.parametrize("row_count", [100, 1000])
+@pytest.mark.parametrize("column_count", [10, 100])
+@pytest.mark.parametrize("weighted", [True, False])
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_partial_fit_multiple_options_on_random_data(
+    dataframe, queue, num_batches, row_count, column_count, weighted, dtype
+):
+    seed = 42
+    gen = np.random.default_rng(seed)
+    X = gen.uniform(low=-0.3, high=+0.7, size=(row_count, column_count))
+    X = X.astype(dtype=dtype)
+    X_split = np.array_split(X, num_batches)
+    if weighted:
+        weights = gen.uniform(low=-0.5, high=+1.0, size=row_count)
+        weights = weights.astype(dtype=dtype)
+        weights_split = np.array_split(weights, num_batches)
+    incbs = IncrementalBasicStatistics(result_options=["mean", "max", "sum"])
+
+    for i in range(num_batches):
+        X_split_df = _convert_to_dataframe(
+            X_split[i], sycl_queue=queue, target_df=dataframe
+        )
+        if weighted:
+            weights_split_df = _convert_to_dataframe(
+                weights_split[i], sycl_queue=queue, target_df=dataframe
+            )
+            result = incbs.partial_fit(X_split_df, sample_weight=weights_split_df)
+        else:
+            result = incbs.partial_fit(X_split_df)
+
+    res_mean, res_max, res_sum = result.mean, result.max, result.sum
+    if weighted:
+        weighted_data = np.diag(weights) @ X
+        gtr_mean, gtr_max, gtr_sum = (
+            expected_mean(weighted_data),
+            expected_max(weighted_data),
+            expected_sum(weighted_data),
+        )
+    else:
+        gtr_mean, gtr_max, gtr_sum = (
+            expected_mean(X),
+            expected_max(X),
+            expected_sum(X),
+        )
+
+    tol = 3e-4 if res_mean.dtype == np.float32 else 1e-7
+    assert_allclose(gtr_mean, res_mean, atol=tol)
+    assert_allclose(gtr_max, res_max, atol=tol)
+    assert_allclose(gtr_sum, res_sum, atol=tol)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("num_batches", [2, 10])
+@pytest.mark.parametrize("row_count", [100, 1000])
+@pytest.mark.parametrize("column_count", [10, 100])
+@pytest.mark.parametrize("weighted", [True, False])
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_partial_fit_all_option_on_random_data(
+    dataframe, queue, num_batches, row_count, column_count, weighted, dtype
+):
+    seed = 77
+    gen = np.random.default_rng(seed)
+    X = gen.uniform(low=-0.3, high=+0.7, size=(row_count, column_count))
+    X = X.astype(dtype=dtype)
+    X_split = np.array_split(X, num_batches)
+    if weighted:
+        weights = gen.uniform(low=-0.5, high=+1.0, size=row_count)
+        weights = weights.astype(dtype=dtype)
+        weights_split = np.array_split(weights, num_batches)
+    incbs = IncrementalBasicStatistics(result_options="all")
+
+    for i in range(num_batches):
+        X_split_df = _convert_to_dataframe(
+            X_split[i], sycl_queue=queue, target_df=dataframe
+        )
+        if weighted:
+            weights_split_df = _convert_to_dataframe(
+                weights_split[i], sycl_queue=queue, target_df=dataframe
+            )
+            result = incbs.partial_fit(X_split_df, sample_weight=weights_split_df)
+        else:
+            result = incbs.partial_fit(X_split_df)
+
+    if weighted:
+        weighted_data = np.diag(weights) @ X
+
+    for option in options_and_tests:
+        result_option, function, tols = option
+        fp32tol, fp64tol = tols
+        res = getattr(result, result_option)
+        if weighted:
+            gtr = function(weighted_data)
+        else:
+            gtr = function(X)
+        tol = fp32tol if res.dtype == np.float32 else fp64tol
+        assert_allclose(gtr, res, atol=tol)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("weighted", [True, False])
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_fit_multiple_options_on_gold_data(dataframe, queue, weighted, dtype):
+    X = np.array([[0, 0], [1, 1]])
+    X = X.astype(dtype=dtype)
+    X_df = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
+    if weighted:
+        weights = np.array([1, 0.5])
+        weights = weights.astype(dtype=dtype)
+        weights_df = _convert_to_dataframe(weights, sycl_queue=queue, target_df=dataframe)
+    incbs = IncrementalBasicStatistics(batch_size=1)
+
+    if weighted:
+        result = incbs.fit(X_df, sample_weight=weights_df)
+    else:
+        result = incbs.fit(X_df)
+
+    if weighted:
+        expected_weighted_mean = np.array([0.25, 0.25])
+        expected_weighted_min = np.array([0, 0])
+        expected_weighted_max = np.array([0.5, 0.5])
+        assert_allclose(expected_weighted_mean, result.mean)
+        assert_allclose(expected_weighted_max, result.max)
+        assert_allclose(expected_weighted_min, result.min)
+    else:
+        expected_mean = np.array([0.5, 0.5])
+        expected_min = np.array([0, 0])
+        expected_max = np.array([1, 1])
+        assert_allclose(expected_mean, result.mean)
+        assert_allclose(expected_max, result.max)
+        assert_allclose(expected_min, result.min)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("num_batches", [2, 10])
+@pytest.mark.parametrize("option", options_and_tests)
+@pytest.mark.parametrize("row_count", [100, 1000])
+@pytest.mark.parametrize("column_count", [10, 100])
+@pytest.mark.parametrize("weighted", [True, False])
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_fit_single_option_on_random_data(
+    dataframe, queue, num_batches, option, row_count, column_count, weighted, dtype
+):
+    result_option, function, tols = option
+    fp32tol, fp64tol = tols
+    seed = 77
+    gen = np.random.default_rng(seed)
+    batch_size = row_count // num_batches
+    X = gen.uniform(low=-0.3, high=+0.7, size=(row_count, column_count))
+    X = X.astype(dtype=dtype)
+    X_df = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
+    if weighted:
+        weights = gen.uniform(low=-0.5, high=1.0, size=row_count)
+        weights = weights.astype(dtype=dtype)
+        weights_df = _convert_to_dataframe(weights, sycl_queue=queue, target_df=dataframe)
+    incbs = IncrementalBasicStatistics(
+        result_options=result_option, batch_size=batch_size
+    )
+
+    if weighted:
+        result = incbs.fit(X_df, sample_weight=weights_df)
+    else:
+        result = incbs.fit(X_df)
+
+    res = getattr(result, result_option)
+    if weighted:
+        weighted_data = np.diag(weights) @ X
+        gtr = function(weighted_data)
+    else:
+        gtr = function(X)
+
+    tol = fp32tol if res.dtype == np.float32 else fp64tol
+    assert_allclose(gtr, res, atol=tol)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("num_batches", [2, 10])
+@pytest.mark.parametrize("row_count", [100, 1000])
+@pytest.mark.parametrize("column_count", [10, 100])
+@pytest.mark.parametrize("weighted", [True, False])
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_fit_multiple_options_on_random_data(
+    dataframe, queue, num_batches, row_count, column_count, weighted, dtype
+):
+    seed = 77
+    gen = np.random.default_rng(seed)
+    batch_size = row_count // num_batches
+    X = gen.uniform(low=-0.3, high=+0.7, size=(row_count, column_count))
+    X = X.astype(dtype=dtype)
+    X_df = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
+    if weighted:
+        weights = gen.uniform(low=-0.5, high=1.0, size=row_count)
+        weights = weights.astype(dtype=dtype)
+        weights_df = _convert_to_dataframe(weights, sycl_queue=queue, target_df=dataframe)
+    incbs = IncrementalBasicStatistics(
+        result_options=["mean", "max", "sum"], batch_size=batch_size
+    )
+
+    if weighted:
+        result = incbs.fit(X_df, sample_weight=weights_df)
+    else:
+        result = incbs.fit(X_df)
+
+    res_mean, res_max, res_sum = result.mean, result.max, result.sum
+    if weighted:
+        weighted_data = np.diag(weights) @ X
+        gtr_mean, gtr_max, gtr_sum = (
+            expected_mean(weighted_data),
+            expected_max(weighted_data),
+            expected_sum(weighted_data),
+        )
+    else:
+        gtr_mean, gtr_max, gtr_sum = (
+            expected_mean(X),
+            expected_max(X),
+            expected_sum(X),
+        )
+
+    tol = 3e-4 if res_mean.dtype == np.float32 else 1e-7
+    assert_allclose(gtr_mean, res_mean, atol=tol)
+    assert_allclose(gtr_max, res_max, atol=tol)
+    assert_allclose(gtr_sum, res_sum, atol=tol)
+
+
+@pytest.mark.parametrize("dataframe,queue", get_dataframes_and_queues())
+@pytest.mark.parametrize("num_batches", [2, 10])
+@pytest.mark.parametrize("row_count", [100, 1000])
+@pytest.mark.parametrize("column_count", [10, 100])
+@pytest.mark.parametrize("weighted", [True, False])
+@pytest.mark.parametrize("dtype", [np.float32, np.float64])
+def test_fit_all_option_on_random_data(
+    dataframe, queue, num_batches, row_count, column_count, weighted, dtype
+):
+    seed = 77
+    gen = np.random.default_rng(seed)
+    batch_size = row_count // num_batches
+    X = gen.uniform(low=-0.3, high=+0.7, size=(row_count, column_count))
+    X = X.astype(dtype=dtype)
+    X_df = _convert_to_dataframe(X, sycl_queue=queue, target_df=dataframe)
+    if weighted:
+        weights = gen.uniform(low=-0.5, high=+1.0, size=row_count)
+        weights = weights.astype(dtype=dtype)
+        weights_df = _convert_to_dataframe(weights, sycl_queue=queue, target_df=dataframe)
+    incbs = IncrementalBasicStatistics(result_options="all", batch_size=batch_size)
+
+    if weighted:
+        result = incbs.fit(X_df, sample_weight=weights_df)
+    else:
+        result = incbs.fit(X_df)
+
+    if weighted:
+        weighted_data = np.diag(weights) @ X
+
+    for option in options_and_tests:
+        result_option, function, tols = option
+        fp32tol, fp64tol = tols
+        res = getattr(result, result_option)
+        if weighted:
+            gtr = function(weighted_data)
+        else:
+            gtr = function(X)
+        tol = fp32tol if res.dtype == np.float32 else fp64tol
+        assert_allclose(gtr, res, atol=tol)
+
+
+def test_warning():
+    basicstat = IncrementalBasicStatistics("all")
+    # Only 2d inputs supported into IncrementalBasicStatistics
+    data = np.array([[0.0], [1.0]])
+
+    basicstat.fit(data)
+    for i in basicstat._onedal_estimator.get_all_result_options():
+        with pytest.warns(
+            UserWarning,
+            match="Result attributes without a trailing underscore were deprecated in version 2025.1 and will be removed in 2026.0",
+        ) as warn_record:
+            getattr(basicstat, i)
+
+        if daal_check_version((2026, "P", 0)):
+            assert len(warn_record) == 0, i
+        else:
+            assert len(warn_record) == 1, i

sklearnex/cluster/__init__.py

@@ -0,0 +1,20 @@
+# ===============================================================================
+# 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.
+# ===============================================================================
+
+from .dbscan import DBSCAN
+from .k_means import KMeans
+
+__all__ = ["DBSCAN", "KMeans"]

sklearnex/cluster/dbscan.py

@@ -0,0 +1,197 @@
+# ===============================================================================
+# 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 numbers
+from abc import ABC
+
+from scipy import sparse as sp
+from sklearn.cluster import DBSCAN as _sklearn_DBSCAN
+from sklearn.utils.validation import _check_sample_weight
+
+from daal4py.sklearn._n_jobs_support import control_n_jobs
+from daal4py.sklearn._utils import sklearn_check_version
+from onedal.cluster import DBSCAN as onedal_DBSCAN
+
+from .._device_offload import dispatch
+from .._utils import PatchingConditionsChain
+
+if sklearn_check_version("1.1") and not sklearn_check_version("1.2"):
+    from sklearn.utils import check_scalar
+
+if sklearn_check_version("1.6"):
+    from sklearn.utils.validation import validate_data
+else:
+    validate_data = _sklearn_DBSCAN._validate_data
+
+
+class BaseDBSCAN(ABC):
+    def _onedal_dbscan(self, **onedal_params):
+        return onedal_DBSCAN(**onedal_params)
+
+    def _save_attributes(self):
+        assert hasattr(self, "_onedal_estimator")
+
+        self.labels_ = self._onedal_estimator.labels_
+        self.core_sample_indices_ = self._onedal_estimator.core_sample_indices_
+        self.components_ = self._onedal_estimator.components_
+        self.n_features_in_ = self._onedal_estimator.n_features_in_
+
+
+@control_n_jobs(decorated_methods=["fit"])
+class DBSCAN(_sklearn_DBSCAN, BaseDBSCAN):
+    __doc__ = _sklearn_DBSCAN.__doc__
+
+    if sklearn_check_version("1.2"):
+        _parameter_constraints: dict = {**_sklearn_DBSCAN._parameter_constraints}
+
+    def __init__(
+        self,
+        eps=0.5,
+        *,
+        min_samples=5,
+        metric="euclidean",
+        metric_params=None,
+        algorithm="auto",
+        leaf_size=30,
+        p=None,
+        n_jobs=None,
+    ):
+        super(DBSCAN, self).__init__(
+            eps=eps,
+            min_samples=min_samples,
+            metric=metric,
+            metric_params=metric_params,
+            algorithm=algorithm,
+            leaf_size=leaf_size,
+            p=p,
+            n_jobs=n_jobs,
+        )
+        self.eps = eps
+        self.min_samples = min_samples
+        self.metric = metric
+        self.metric_params = metric_params
+        self.algorithm = algorithm
+        self.leaf_size = leaf_size
+        self.p = p
+        self.n_jobs = n_jobs
+
+    def _onedal_fit(self, X, y, sample_weight=None, queue=None):
+        if sklearn_check_version("1.0"):
+            X = validate_data(self, X, force_all_finite=False)
+
+        onedal_params = {
+            "eps": self.eps,
+            "min_samples": self.min_samples,
+            "metric": self.metric,
+            "metric_params": self.metric_params,
+            "algorithm": self.algorithm,
+            "leaf_size": self.leaf_size,
+            "p": self.p,
+            "n_jobs": self.n_jobs,
+        }
+        self._onedal_estimator = self._onedal_dbscan(**onedal_params)
+
+        self._onedal_estimator.fit(X, y=y, sample_weight=sample_weight, queue=queue)
+        self._save_attributes()
+
+    def _onedal_supported(self, method_name, *data):
+        class_name = self.__class__.__name__
+        patching_status = PatchingConditionsChain(
+            f"sklearn.cluster.{class_name}.{method_name}"
+        )
+        if method_name == "fit":
+            X, y, sample_weight = data
+            patching_status.and_conditions(
+                [
+                    (
+                        self.algorithm in ["auto", "brute"],
+                        f"'{self.algorithm}' algorithm is not supported. "
+                        "Only 'auto' and 'brute' algorithms are supported",
+                    ),
+                    (
+                        self.metric == "euclidean"
+                        or (self.metric == "minkowski" and self.p == 2),
+                        f"'{self.metric}' (p={self.p}) metric is not supported. "
+                        "Only 'euclidean' or 'minkowski' with p=2 metrics are supported.",
+                    ),
+                    (not sp.issparse(X), "X is sparse. Sparse input is not supported."),
+                ]
+            )
+            return patching_status
+        raise RuntimeError(f"Unknown method {method_name} in {self.__class__.__name__}")
+
+    def _onedal_cpu_supported(self, method_name, *data):
+        return self._onedal_supported(method_name, *data)
+
+    def _onedal_gpu_supported(self, method_name, *data):
+        return self._onedal_supported(method_name, *data)
+
+    def fit(self, X, y=None, sample_weight=None):
+        if sklearn_check_version("1.2"):
+            self._validate_params()
+        elif sklearn_check_version("1.1"):
+            check_scalar(
+                self.eps,
+                "eps",
+                target_type=numbers.Real,
+                min_val=0.0,
+                include_boundaries="neither",
+            )
+            check_scalar(
+                self.min_samples,
+                "min_samples",
+                target_type=numbers.Integral,
+                min_val=1,
+                include_boundaries="left",
+            )
+            check_scalar(
+                self.leaf_size,
+                "leaf_size",
+                target_type=numbers.Integral,
+                min_val=1,
+                include_boundaries="left",
+            )
+            if self.p is not None:
+                check_scalar(
+                    self.p,
+                    "p",
+                    target_type=numbers.Real,
+                    min_val=0.0,
+                    include_boundaries="left",
+                )
+            if self.n_jobs is not None:
+                check_scalar(self.n_jobs, "n_jobs", target_type=numbers.Integral)
+        else:
+            if self.eps <= 0.0:
+                raise ValueError(f"eps == {self.eps}, must be > 0.0.")
+
+        if sample_weight is not None:
+            sample_weight = _check_sample_weight(sample_weight, X)
+        dispatch(
+            self,
+            "fit",
+            {
+                "onedal": self.__class__._onedal_fit,
+                "sklearn": _sklearn_DBSCAN.fit,
+            },
+            X,
+            y,
+            sample_weight,
+        )
+
+        return self
+
+    fit.__doc__ = _sklearn_DBSCAN.fit.__doc__