scikit-learn-intelex 2025.4.0__py313-none-win_amd64.whl → 2025.6.0__py313-none-win_amd64.whl

{scikit_learn_intelex-2025.4.0.data → scikit_learn_intelex-2025.6.0.data}/data/Lib/site-packages/sklearnex/svm/nusvc.py

@@ -14,10 +14,13 @@
 # limitations under the License.
 # ==============================================================================
 
+from functools import wraps
+
 import numpy as np
 from sklearn.exceptions import NotFittedError
 from sklearn.metrics import accuracy_score
 from sklearn.svm import NuSVC as _sklearn_NuSVC
+from sklearn.utils.metaestimators import available_if
 from sklearn.utils.validation import (
     _deprecate_positional_args,
     check_array,
@@ -26,26 +29,18 @@ from sklearn.utils.validation import (
 
 from daal4py.sklearn._n_jobs_support import control_n_jobs
 from daal4py.sklearn._utils import sklearn_check_version
+from onedal.svm import NuSVC as onedal_NuSVC
 
 from .._device_offload import dispatch, wrap_output_data
 from ..utils._array_api import get_namespace
+from ..utils.validation import validate_data
 from ._common import BaseSVC
 
-if sklearn_check_version("1.0"):
-    from sklearn.utils.metaestimators import available_if
-
-from onedal.svm import NuSVC as onedal_NuSVC
-
-if sklearn_check_version("1.6"):
-    from sklearn.utils.validation import validate_data
-else:
-    validate_data = BaseSVC._validate_data
-
 
 @control_n_jobs(
     decorated_methods=["fit", "predict", "_predict_proba", "decision_function", "score"]
 )
-class NuSVC(_sklearn_NuSVC, BaseSVC):
+class NuSVC(BaseSVC, _sklearn_NuSVC):
     __doc__ = _sklearn_NuSVC.__doc__
 
     if sklearn_check_version("1.2"):
@@ -145,99 +140,27 @@ class NuSVC(_sklearn_NuSVC, BaseSVC):
             sample_weight=sample_weight,
         )
 
-    if sklearn_check_version("1.0"):
-
-        @available_if(_sklearn_NuSVC._check_proba)
-        def predict_proba(self, X):
-            """
-            Compute probabilities of possible outcomes for samples in X.
-
-            The model need to have probability information computed at training
-            time: fit with attribute `probability` set to True.
-
-            Parameters
-            ----------
-            X : array-like of shape (n_samples, n_features)
-                For kernel="precomputed", the expected shape of X is
-                (n_samples_test, n_samples_train).
-
-            Returns
-            -------
-            T : ndarray of shape (n_samples, n_classes)
-                Returns the probability of the sample for each class in
-                the model. The columns correspond to the classes in sorted
-                order, as they appear in the attribute :term:`classes_`.
-
-            Notes
-            -----
-            The probability model is created using cross validation, so
-            the results can be slightly different than those obtained by
-            predict. Also, it will produce meaningless results on very small
-            datasets.
-            """
-            check_is_fitted(self)
-            return self._predict_proba(X)
-
-        @available_if(_sklearn_NuSVC._check_proba)
-        def predict_log_proba(self, X):
-            """Compute log probabilities of possible outcomes for samples in X.
-
-            The model need to have probability information computed at training
-            time: fit with attribute `probability` set to True.
-
-            Parameters
-            ----------
-            X : array-like of shape (n_samples, n_features) or \
-                    (n_samples_test, n_samples_train)
-                For kernel="precomputed", the expected shape of X is
-                (n_samples_test, n_samples_train).
-
-            Returns
-            -------
-            T : ndarray of shape (n_samples, n_classes)
-                Returns the log-probabilities of the sample for each class in
-                the model. The columns correspond to the classes in sorted
-                order, as they appear in the attribute :term:`classes_`.
-
-            Notes
-            -----
-            The probability model is created using cross validation, so
-            the results can be slightly different than those obtained by
-            predict. Also, it will produce meaningless results on very small
-            datasets.
-            """
-            xp, _ = get_namespace(X)
-
-            return xp.log(self.predict_proba(X))
-
-    else:
-
-        @property
-        def predict_proba(self):
-            self._check_proba()
-            check_is_fitted(self)
-            return self._predict_proba
-
-        def _predict_log_proba(self, X):
-            xp, _ = get_namespace(X)
-            return xp.log(self.predict_proba(X))
-
-        predict_proba.__doc__ = _sklearn_NuSVC.predict_proba.__doc__
+    @available_if(_sklearn_NuSVC._check_proba)
+    @wraps(_sklearn_NuSVC.predict_proba, assigned=["__doc__"])
+    def predict_proba(self, X):
+        check_is_fitted(self)
+        return self._predict_proba(X)
+
+    @available_if(_sklearn_NuSVC._check_proba)
+    @wraps(_sklearn_NuSVC.predict_log_proba, assigned=["__doc__"])
+    def predict_log_proba(self, X):
+        xp, _ = get_namespace(X)
+
+        return xp.log(self.predict_proba(X))
 
     @wrap_output_data
     def _predict_proba(self, X):
-        sklearn_pred_proba = (
-            _sklearn_NuSVC.predict_proba
-            if sklearn_check_version("1.0")
-            else _sklearn_NuSVC._predict_proba
-        )
-
         return dispatch(
             self,
             "predict_proba",
             {
                 "onedal": self.__class__._onedal_predict_proba,
-                "sklearn": sklearn_pred_proba,
+                "sklearn": _sklearn_NuSVC.predict_proba,
             },
             X,
         )
@@ -306,23 +229,15 @@ class NuSVC(_sklearn_NuSVC, BaseSVC):
         self._save_attributes()
 
     def _onedal_predict(self, X, queue=None):
-        if sklearn_check_version("1.0"):
-            validate_data(
-                self,
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                ensure_2d=False,
-                accept_sparse="csr",
-                reset=False,
-            )
-        else:
-            X = check_array(
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-            )
+        validate_data(
+            self,
+            X,
+            dtype=[np.float64, np.float32],
+            ensure_all_finite=False,
+            ensure_2d=False,
+            accept_sparse="csr",
+            reset=False,
+        )
 
         return self._onedal_estimator.predict(X, queue=queue)
 
@@ -341,22 +256,14 @@ class NuSVC(_sklearn_NuSVC, BaseSVC):
             return self.clf_prob.predict_proba(X)
 
     def _onedal_decision_function(self, X, queue=None):
-        if sklearn_check_version("1.0"):
-            validate_data(
-                self,
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-                reset=False,
-            )
-        else:
-            X = check_array(
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-            )
+        validate_data(
+            self,
+            X,
+            dtype=[np.float64, np.float32],
+            ensure_all_finite=False,
+            accept_sparse="csr",
+            reset=False,
+        )
 
         return self._onedal_estimator.decision_function(X, queue=queue)
 

{scikit_learn_intelex-2025.4.0.data → scikit_learn_intelex-2025.6.0.data}/data/Lib/site-packages/sklearnex/svm/nusvr.py

@@ -27,16 +27,12 @@ from daal4py.sklearn._utils import sklearn_check_version
 from onedal.svm import NuSVR as onedal_NuSVR
 
 from .._device_offload import dispatch, wrap_output_data
+from ..utils.validation import validate_data
 from ._common import BaseSVR
 
-if sklearn_check_version("1.6"):
-    from sklearn.utils.validation import validate_data
-else:
-    validate_data = BaseSVR._validate_data
-
 
 @control_n_jobs(decorated_methods=["fit", "predict", "score"])
-class NuSVR(_sklearn_NuSVR, BaseSVR):
+class NuSVR(BaseSVR, _sklearn_NuSVR):
     __doc__ = _sklearn_NuSVR.__doc__
 
     if sklearn_check_version("1.2"):
@@ -147,22 +143,14 @@ class NuSVR(_sklearn_NuSVR, BaseSVR):
         self._save_attributes()
 
     def _onedal_predict(self, X, queue=None):
-        if sklearn_check_version("1.0"):
-            X = validate_data(
-                self,
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-                reset=False,
-            )
-        else:
-            X = check_array(
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-            )
+        X = validate_data(
+            self,
+            X,
+            dtype=[np.float64, np.float32],
+            ensure_all_finite=False,
+            accept_sparse="csr",
+            reset=False,
+        )
         return self._onedal_estimator.predict(X, queue=queue)
 
     fit.__doc__ = _sklearn_NuSVR.fit.__doc__

{scikit_learn_intelex-2025.4.0.data → scikit_learn_intelex-2025.6.0.data}/data/Lib/site-packages/sklearnex/svm/svc.py

@@ -14,11 +14,14 @@
 # limitations under the License.
 # ==============================================================================
 
+from functools import wraps
+
 import numpy as np
 from scipy import sparse as sp
 from sklearn.exceptions import NotFittedError
 from sklearn.metrics import accuracy_score
 from sklearn.svm import SVC as _sklearn_SVC
+from sklearn.utils.metaestimators import available_if
 from sklearn.utils.validation import (
     _deprecate_positional_args,
     check_array,
@@ -27,27 +30,19 @@ from sklearn.utils.validation import (
 
 from daal4py.sklearn._n_jobs_support import control_n_jobs
 from daal4py.sklearn._utils import sklearn_check_version
+from onedal.svm import SVC as onedal_SVC
 
 from .._device_offload import dispatch, wrap_output_data
 from .._utils import PatchingConditionsChain
 from ..utils._array_api import get_namespace
+from ..utils.validation import validate_data
 from ._common import BaseSVC
 
-if sklearn_check_version("1.0"):
-    from sklearn.utils.metaestimators import available_if
-
-from onedal.svm import SVC as onedal_SVC
-
-if sklearn_check_version("1.6"):
-    from sklearn.utils.validation import validate_data
-else:
-    validate_data = BaseSVC._validate_data
-
 
 @control_n_jobs(
     decorated_methods=["fit", "predict", "_predict_proba", "decision_function", "score"]
 )
-class SVC(_sklearn_SVC, BaseSVC):
+class SVC(BaseSVC, _sklearn_SVC):
     __doc__ = _sklearn_SVC.__doc__
 
     if sklearn_check_version("1.2"):
@@ -147,99 +142,27 @@ class SVC(_sklearn_SVC, BaseSVC):
             sample_weight=sample_weight,
         )
 
-    if sklearn_check_version("1.0"):
-
-        @available_if(_sklearn_SVC._check_proba)
-        def predict_proba(self, X):
-            """
-            Compute probabilities of possible outcomes for samples in X.
-
-            The model need to have probability information computed at training
-            time: fit with attribute `probability` set to True.
-
-            Parameters
-            ----------
-            X : array-like of shape (n_samples, n_features)
-                For kernel="precomputed", the expected shape of X is
-                (n_samples_test, n_samples_train).
-
-            Returns
-            -------
-            T : ndarray of shape (n_samples, n_classes)
-                Returns the probability of the sample for each class in
-                the model. The columns correspond to the classes in sorted
-                order, as they appear in the attribute :term:`classes_`.
-
-            Notes
-            -----
-            The probability model is created using cross validation, so
-            the results can be slightly different than those obtained by
-            predict. Also, it will produce meaningless results on very small
-            datasets.
-            """
-            check_is_fitted(self)
-            return self._predict_proba(X)
-
-        @available_if(_sklearn_SVC._check_proba)
-        def predict_log_proba(self, X):
-            """Compute log probabilities of possible outcomes for samples in X.
-
-            The model need to have probability information computed at training
-            time: fit with attribute `probability` set to True.
-
-            Parameters
-            ----------
-            X : array-like of shape (n_samples, n_features) or \
-                    (n_samples_test, n_samples_train)
-                For kernel="precomputed", the expected shape of X is
-                (n_samples_test, n_samples_train).
-
-            Returns
-            -------
-            T : ndarray of shape (n_samples, n_classes)
-                Returns the log-probabilities of the sample for each class in
-                the model. The columns correspond to the classes in sorted
-                order, as they appear in the attribute :term:`classes_`.
-
-            Notes
-            -----
-            The probability model is created using cross validation, so
-            the results can be slightly different than those obtained by
-            predict. Also, it will produce meaningless results on very small
-            datasets.
-            """
-            xp, _ = get_namespace(X)
-
-            return xp.log(self.predict_proba(X))
-
-    else:
-
-        @property
-        def predict_proba(self):
-            self._check_proba()
-            check_is_fitted(self)
-            return self._predict_proba
+    @available_if(_sklearn_SVC._check_proba)
+    @wraps(_sklearn_SVC.predict_proba, assigned=["__doc__"])
+    def predict_proba(self, X):
+        check_is_fitted(self)
+        return self._predict_proba(X)
 
-        def _predict_log_proba(self, X):
-            xp, _ = get_namespace(X)
-            return xp.log(self.predict_proba(X))
+    @available_if(_sklearn_SVC._check_proba)
+    @wraps(_sklearn_SVC.predict_log_proba, assigned=["__doc__"])
+    def predict_log_proba(self, X):
+        xp, _ = get_namespace(X)
 
-        predict_proba.__doc__ = _sklearn_SVC.predict_proba.__doc__
+        return xp.log(self.predict_proba(X))
 
     @wrap_output_data
     def _predict_proba(self, X):
-        sklearn_pred_proba = (
-            _sklearn_SVC.predict_proba
-            if sklearn_check_version("1.0")
-            else _sklearn_SVC._predict_proba
-        )
-
         return dispatch(
             self,
             "predict_proba",
             {
                 "onedal": self.__class__._onedal_predict_proba,
-                "sklearn": sklearn_pred_proba,
+                "sklearn": _sklearn_SVC.predict_proba,
             },
             X,
         )
@@ -336,23 +259,15 @@ class SVC(_sklearn_SVC, BaseSVC):
         self._save_attributes()
 
     def _onedal_predict(self, X, queue=None):
-        if sklearn_check_version("1.0"):
-            X = validate_data(
-                self,
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                ensure_2d=False,
-                accept_sparse="csr",
-                reset=False,
-            )
-        else:
-            X = check_array(
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-            )
+        X = validate_data(
+            self,
+            X,
+            dtype=[np.float64, np.float32],
+            ensure_all_finite=False,
+            ensure_2d=False,
+            accept_sparse="csr",
+            reset=False,
+        )
         return self._onedal_estimator.predict(X, queue=queue)
 
     def _onedal_predict_proba(self, X, queue=None):
@@ -370,22 +285,14 @@ class SVC(_sklearn_SVC, BaseSVC):
             return self.clf_prob.predict_proba(X)
 
     def _onedal_decision_function(self, X, queue=None):
-        if sklearn_check_version("1.0"):
-            X = validate_data(
-                self,
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-                reset=False,
-            )
-        else:
-            X = check_array(
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-            )
+        X = validate_data(
+            self,
+            X,
+            dtype=[np.float64, np.float32],
+            ensure_all_finite=False,
+            accept_sparse="csr",
+            reset=False,
+        )
         return self._onedal_estimator.decision_function(X, queue=queue)
 
     def _onedal_score(self, X, y, sample_weight=None, queue=None):

{scikit_learn_intelex-2025.4.0.data → scikit_learn_intelex-2025.6.0.data}/data/Lib/site-packages/sklearnex/svm/svr.py

@@ -23,16 +23,12 @@ from daal4py.sklearn._utils import sklearn_check_version
 from onedal.svm import SVR as onedal_SVR
 
 from .._device_offload import dispatch, wrap_output_data
+from ..utils.validation import validate_data
 from ._common import BaseSVR
 
-if sklearn_check_version("1.6"):
-    from sklearn.utils.validation import validate_data
-else:
-    validate_data = BaseSVR._validate_data
-
 
 @control_n_jobs(decorated_methods=["fit", "predict", "score"])
-class SVR(_sklearn_SVR, BaseSVR):
+class SVR(BaseSVR, _sklearn_SVR):
     __doc__ = _sklearn_SVR.__doc__
 
     if sklearn_check_version("1.2"):
@@ -144,22 +140,14 @@ class SVR(_sklearn_SVR, BaseSVR):
         self._save_attributes()
 
     def _onedal_predict(self, X, queue=None):
-        if sklearn_check_version("1.0"):
-            X = validate_data(
-                self,
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-                reset=False,
-            )
-        else:
-            X = check_array(
-                X,
-                dtype=[np.float64, np.float32],
-                force_all_finite=False,
-                accept_sparse="csr",
-            )
+        X = validate_data(
+            self,
+            X,
+            dtype=[np.float64, np.float32],
+            ensure_all_finite=False,
+            accept_sparse="csr",
+            reset=False,
+        )
         return self._onedal_estimator.predict(X, queue=queue)
 
     fit.__doc__ = _sklearn_SVR.fit.__doc__

{scikit_learn_intelex-2025.4.0.data → scikit_learn_intelex-2025.6.0.data}/data/Lib/site-packages/sklearnex/tests/test_common.py

@@ -26,13 +26,16 @@ from contextlib import redirect_stdout
 from multiprocessing import Pipe, Process, get_context
 
 import pytest
+from sklearn.base import BaseEstimator
 from sklearn.utils import all_estimators
 
 from daal4py.sklearn._utils import sklearn_check_version
 from onedal.tests.test_common import _check_primitive_usage_ban
+from sklearnex.base import oneDALEstimator
 from sklearnex.tests.utils import (
     PATCHED_MODELS,
     SPECIAL_INSTANCES,
+    UNPATCHED_MODELS,
     call_method,
     gen_dataset,
     gen_models_info,
@@ -176,6 +179,28 @@ def test_all_estimators_covered(monkeypatch):
     ), f"{uncovered_estimators} are currently not included"
 
 
+def test_oneDALEstimator_inheritance(monkeypatch):
+    """All sklearnex estimators should inherit the oneDALEstimator class, sklearnex-only
+    estimators should have it inherit oneDAL estimator one step before BaseEstimator in the
+    mro.  This is only strictly set for non-preview estimators"""
+    monkeypatch.setattr(pkgutil, "walk_packages", _sklearnex_walk(pkgutil.walk_packages))
+    estimators = all_estimators()  # list of tuples
+    for name, obj in estimators:
+        if "preview" not in obj.__module__ and "daal4py" not in obj.__module__:
+            assert issubclass(
+                obj, oneDALEstimator
+            ), f"{name} does not inherit the oneDALEstimator"
+            # oneDAL estimator should be inherited from before BaseEstimator
+            mro = obj.__mro__
+            assert mro.index(oneDALEstimator) < mro.index(
+                BaseEstimator
+            ), f"incorrect mro in {name}"
+            if not any([issubclass(obj, est) for est in UNPATCHED_MODELS.values()]):
+                assert (
+                    mro[mro.index(oneDALEstimator) + 1] is BaseEstimator
+                ), f"oneDALEstimator should be inherited just before BaseEstimator in {name}"
+
+
 def _fullpath(path):
     return os.path.realpath(os.path.expanduser(path))
 
@@ -406,22 +431,36 @@ def estimator_trace(estimator, method, cache, isolated_trace):
 
 
 def call_validate_data(text, estimator, method):
-    """test that the sklearn function/attribute validate_data is
+    """test that both sklearnex wrapper for validate_data and
+    original sklearn function/method validate_data are
     called once before offloading to oneDAL in sklearnex"""
     try:
         # get last to_table call showing end of oneDAL input portion of code
         idx = len(text["funcs"]) - 1 - text["funcs"][::-1].index("to_table")
-        validfuncs = text["funcs"][:idx]
+        valid_funcs = text["funcs"][:idx]
+        valid_modules = text["modules"][:idx]
     except ValueError:
         pytest.skip("onedal backend not used in this function")
 
-    validate_data = "validate_data" if sklearn_check_version("1.6") else "_validate_data"
+    validate_data_calls = []
+    for func, module in zip(valid_funcs, valid_modules):
+        if func.endswith("validate_data"):
+            validate_data_calls.append({module, func})
 
     assert (
-        validfuncs.count(validate_data) == 1
-    ), f"sklearn's {validate_data} should be called"
+        len(validate_data_calls) == 2
+    ), "validate_data should be called two times: once for sklearn and once for sklearnex"
+    assert validate_data_calls[0] == {
+        "sklearnex.utils.validation",
+        "validate_data",
+    }, "sklearnex's validate_data should be called first"
+    assert (
+        (validate_data_calls[1] == {"sklearn.utils.validation", "validate_data"})
+        if sklearn_check_version("1.6")
+        else (validate_data_calls[1] == {"sklearn.base", "_validate_data"})
+    ), "sklearn's validate_data should be called second"
     assert (
-        validfuncs.count("_check_feature_names") == 1
+        valid_funcs.count("_check_feature_names") == 1
     ), "estimator should check feature names in validate_data"
 
 
@@ -467,11 +506,12 @@ def fit_check_before_support_check(text, estimator, method):
         pytest.skip(f"fitting occurs in {estimator}.{method}")
 
 
-DESIGN_RULES = [n_jobs_check, runtime_property_check, fit_check_before_support_check]
-
-
-if sklearn_check_version("1.0"):
-    DESIGN_RULES += [call_validate_data]
+DESIGN_RULES = [
+    n_jobs_check,
+    runtime_property_check,
+    fit_check_before_support_check,
+    call_validate_data,
+]
 
 
 @pytest.mark.parametrize("design_pattern", DESIGN_RULES)