FIX: Fix html widget docs links (#2299)

* fix html widget doc link

* workaround for older sklearn

* fix test

* attempt at fixing n_jobs failures

* better fix for TSNE

* revert test to 'daal4py'

Author: david-cortes-intel

sklearnex/_utils.py

@@ -16,13 +16,27 @@
 
 import logging
 import os
+import re
 import warnings
 from abc import ABC
 
+import sklearn
+
 from daal4py.sklearn._utils import (
     PatchingConditionsChain as daal4py_PatchingConditionsChain,
 )
-from daal4py.sklearn._utils import daal_check_version
+from daal4py.sklearn._utils import daal_check_version, sklearn_check_version
+
+# Note: if inheriting from '_HTMLDocumentationLinkMixin' here, it then doesn't matter
+# the order of inheritance of classes for estimators when this is later subclassed,
+# whereas if inheriting from something else, the subclass that inherits from this needs
+# to be the first inherited class in estimators in order for it to take effect.
+if sklearn_check_version("1.4"):
+    from sklearn.utils._estimator_html_repr import _HTMLDocumentationLinkMixin
+
+    BaseForHTMLDocLink = _HTMLDocumentationLinkMixin
+else:
+    BaseForHTMLDocLink = ABC
 
 
 class PatchingConditionsChain(daal4py_PatchingConditionsChain):
@@ -128,10 +142,8 @@ def get_hyperparameters(self, op):
 
 
 # This abstract class is meant to generate a clickable doc link for classses
-# in sklearnex that are not part of base scikit-learn. It should be inherited
-# before inheriting from a scikit-learn estimator, otherwise will get overriden
-# by the estimator's original.
-class IntelEstimator(ABC):
+# in sklearnex that are not part of base scikit-learn.
+class IntelEstimator(BaseForHTMLDocLink):
     @property
     def _doc_link_module(self) -> str:
         return "sklearnex"
@@ -141,3 +153,25 @@ def _doc_link_template(self) -> str:
         module_path, _ = self.__class__.__module__.rsplit(".", 1)
         class_name = self.__class__.__name__
         return f"https://uxlfoundation.github.io/scikit-learn-intelex/latest/non-scikit-algorithms.html#{module_path}.{class_name}"
+
+
+# This abstract class is meant to generate a clickable doc link for classses
+# in sklearnex that have counterparts in scikit-learn.
+class PatchableEstimator(BaseForHTMLDocLink):
+    @property
+    def _doc_link_module(self) -> str:
+        return "sklearnex"
+
+    @property
+    def _doc_link_template(self) -> str:
+        if re.search(r"^\d\.\d\.\d$", sklearn.__version__):
+            sklearn_version_parts = sklearn.__version__.split(".")
+            doc_version_url = sklearn_version_parts[0] + "." + sklearn_version_parts[1]
+        else:
+            doc_version_url = "stable"
+        module_path, _ = self.__class__.__module__.rsplit(".", 1)
+        module_path = re.sub("sklearnex", "sklearn", module_path)
+        class_name = self.__class__.__name__
+        # for TSNE, which re-uses daal4py
+        module_path = re.sub(r"daal4py\.", "", module_path)
+        return f"https://scikit-learn.org/{doc_version_url}/modules/generated/{module_path}.{class_name}.html"

sklearnex/cluster/dbscan.py

@@ -26,7 +26,7 @@
 from onedal.cluster import DBSCAN as onedal_DBSCAN
 
 from .._device_offload import dispatch
-from .._utils import PatchingConditionsChain
+from .._utils import PatchableEstimator, PatchingConditionsChain
 
 if sklearn_check_version("1.1") and not sklearn_check_version("1.2"):
     from sklearn.utils import check_scalar
@@ -51,7 +51,7 @@ def _save_attributes(self):
 
 
 @control_n_jobs(decorated_methods=["fit"])
-class DBSCAN(_sklearn_DBSCAN, BaseDBSCAN):
+class DBSCAN(PatchableEstimator, _sklearn_DBSCAN, BaseDBSCAN):
     __doc__ = _sklearn_DBSCAN.__doc__
 
     if sklearn_check_version("1.2"):

sklearnex/cluster/k_means.py

@@ -39,15 +39,15 @@
     from onedal.utils import _is_csr
 
     from .._device_offload import dispatch, wrap_output_data
-    from .._utils import PatchingConditionsChain
+    from .._utils import PatchableEstimator, PatchingConditionsChain
 
     if sklearn_check_version("1.6"):
         from sklearn.utils.validation import validate_data
     else:
         validate_data = _sklearn_KMeans._validate_data
 
     @control_n_jobs(decorated_methods=["fit", "fit_transform", "predict", "score"])
-    class KMeans(_sklearn_KMeans):
+    class KMeans(PatchableEstimator, _sklearn_KMeans):
         __doc__ = _sklearn_KMeans.__doc__
 
         if sklearn_check_version("1.2"):

sklearnex/decomposition/pca.py

@@ -18,6 +18,8 @@
 
 from daal4py.sklearn._utils import daal_check_version
 
+from .._utils import PatchableEstimator
+
 if daal_check_version((2024, "P", 100)):
     import numbers
     from math import sqrt
@@ -50,7 +52,7 @@
         validate_data = _sklearn_PCA._validate_data
 
     @control_n_jobs(decorated_methods=["fit", "transform", "fit_transform"])
-    class PCA(_sklearn_PCA):
+    class PCA(PatchableEstimator, _sklearn_PCA):
         __doc__ = _sklearn_PCA.__doc__
 
         if sklearn_check_version("1.2"):

sklearnex/ensemble/_forest.py

@@ -61,7 +61,7 @@
 from sklearnex._utils import register_hyperparameters
 
 from .._device_offload import dispatch, wrap_output_data
-from .._utils import PatchingConditionsChain
+from .._utils import PatchableEstimator, PatchingConditionsChain
 from ..utils._array_api import get_namespace
 
 if sklearn_check_version("1.2"):
@@ -75,7 +75,7 @@
     validate_data = BaseEstimator._validate_data
 
 
-class BaseForest(ABC):
+class BaseForest(PatchableEstimator, ABC):
     _onedal_factory = None
 
     def _onedal_fit(self, X, y, sample_weight=None, queue=None):

sklearnex/linear_model/linear.py

@@ -27,7 +27,12 @@
 
 from .._config import get_config
 from .._device_offload import dispatch, wrap_output_data
-from .._utils import PatchingConditionsChain, get_patch_message, register_hyperparameters
+from .._utils import (
+    PatchableEstimator,
+    PatchingConditionsChain,
+    get_patch_message,
+    register_hyperparameters,
+)
 
 if sklearn_check_version("1.0") and not sklearn_check_version("1.2"):
     from sklearn.linear_model._base import _deprecate_normalize
@@ -47,7 +52,7 @@
 
 @register_hyperparameters({"fit": get_hyperparameters("linear_regression", "train")})
 @control_n_jobs(decorated_methods=["fit", "predict", "score"])
-class LinearRegression(_sklearn_LinearRegression):
+class LinearRegression(PatchableEstimator, _sklearn_LinearRegression):
     __doc__ = _sklearn_LinearRegression.__doc__
 
     if sklearn_check_version("1.2"):

sklearnex/linear_model/logistic_regression.py

@@ -38,7 +38,7 @@
 
     from .._config import get_config
     from .._device_offload import dispatch, wrap_output_data
-    from .._utils import PatchingConditionsChain, get_patch_message
+    from .._utils import PatchableEstimator, PatchingConditionsChain, get_patch_message
 
     if sklearn_check_version("1.6"):
         from sklearn.utils.validation import validate_data
@@ -65,7 +65,9 @@ def _onedal_gpu_save_attributes(self):
             "score",
         ]
     )
-    class LogisticRegression(_sklearn_LogisticRegression, BaseLogisticRegression):
+    class LogisticRegression(
+        PatchableEstimator, _sklearn_LogisticRegression, BaseLogisticRegression
+    ):
         __doc__ = _sklearn_LogisticRegression.__doc__
 
         if sklearn_check_version("1.2"):

sklearnex/linear_model/ridge.py

@@ -38,15 +38,15 @@
     from onedal.utils import _num_features, _num_samples
 
     from .._device_offload import dispatch, wrap_output_data
-    from .._utils import PatchingConditionsChain
+    from .._utils import PatchableEstimator, PatchingConditionsChain
 
     if sklearn_check_version("1.6"):
         from sklearn.utils.validation import validate_data
     else:
         validate_data = _sklearn_Ridge._validate_data
 
     @control_n_jobs(decorated_methods=["fit", "predict", "score"])
-    class Ridge(_sklearn_Ridge):
+    class Ridge(PatchableEstimator, _sklearn_Ridge):
         __doc__ = _sklearn_Ridge.__doc__
 
         if sklearn_check_version("1.2"):

sklearnex/manifold/t_sne.py

@@ -14,8 +14,13 @@
 # limitations under the License.
 # ===============================================================================
 
+from daal4py.sklearn._n_jobs_support import control_n_jobs
 from daal4py.sklearn.manifold import TSNE
 from onedal._device_offload import support_input_format
 
+from .._utils import PatchableEstimator
+
 TSNE.fit = support_input_format(queue_param=False)(TSNE.fit)
 TSNE.fit_transform = support_input_format(queue_param=False)(TSNE.fit_transform)
+TSNE._doc_link_module = "daal4py"
+TSNE._doc_link_template = PatchableEstimator._doc_link_template

sklearnex/neighbors/common.py

@@ -27,11 +27,11 @@
 from daal4py.sklearn._utils import sklearn_check_version
 from onedal.utils import _check_array, _num_features, _num_samples
 
-from .._utils import PatchingConditionsChain
+from .._utils import PatchableEstimator, PatchingConditionsChain
 from ..utils._array_api import get_namespace
 
 
-class KNeighborsDispatchingBase:
+class KNeighborsDispatchingBase(PatchableEstimator):
     def _fit_validation(self, X, y=None):
         if sklearn_check_version("1.2"):
             self._validate_params()