FEA add link for HTML representation

skrub/_gap_encoder.py

@@ -19,17 +19,12 @@
 
 from . import _dataframe as sbd
 from ._on_each_column import RejectColumn, SingleColumnTransformer
-from ._repr import (
-    _HTMLDocumentationLinkMixin,
-    doc_link_module,
-    doc_link_template,
-    doc_link_url_param_generator,
-)
+from ._repr import _SkrubHTMLDocumentationLinkMixin
 from ._utils import unique_strings
 
 
 class GapEncoder(
-    _HTMLDocumentationLinkMixin, TransformerMixin, SingleColumnTransformer
+    _SkrubHTMLDocumentationLinkMixin, TransformerMixin, SingleColumnTransformer
 ):
     """Constructs latent topics with continuous encoding.
 
@@ -185,10 +180,6 @@ class GapEncoder(
     The higher the value, the bigger the correspondence with the topic.
     """
 
-    _doc_link_module = doc_link_module
-    _doc_link_template = doc_link_template
-    _doc_link_url_param_generator = doc_link_url_param_generator
-
     def __init__(
         self,
         n_components=10,

skrub/_repr.py

@@ -5,11 +5,11 @@
 
 sklearn_version = parse_version(sklearn.__version__)
 
-# TODO: remove when scikit-learn 1.6 is the minimum supported version
-# TODO: subsequently, we should remove the inheritance from _HTMLDocumentationLinkMixin
-# for each estimator then.
+# TODO: remove when scikit-learn 1.6 is the minimum supported version and only import
+# We have this fix due to the following bug:
+# https://github.com/scikit-learn/scikit-learn/pull/29774
 if sklearn_version > parse_version("1.6"):
     from sklearn.utils._estimator_html_repr import _HTMLDocumentationLinkMixin
 else:
 
     class _HTMLDocumentationLinkMixin:
@@ -20,12 +20,12 @@
 
         @property
         def _doc_link_template(self):
             sklearn_version = parse_version(sklearn.__version__)
             if sklearn_version.dev is None:
                 version_url = f"{sklearn_version.major}.{sklearn_version.minor}"
             else:
                 version_url = "dev"
             return getattr(
                 self,
                 "__doc_link_template",
                 (
@@ -36,7 +36,7 @@
 
         @_doc_link_template.setter
         def _doc_link_template(self, value):
             setattr(self, "__doc_link_template", value)
 
         def _get_doc_link(self):
             """Generates a link to the API documentation for a given estimator.
@@ -52,10 +52,10 @@
                 `""`) is returned.
             """
             if self.__class__.__module__.split(".")[0] != self._doc_link_module:
                 return ""
 
             if self._doc_link_url_param_generator is None:
                 estimator_name = self.__class__.__name__
                 # Construct the estimator's module name, up to the first private
                 # submodule. This works because in scikit-learn all public estimators
                 # are exposed at that level, even if they actually live in a private
@@ -66,10 +66,10 @@
                         self.__class__.__module__.split("."),
                     )
                 )
                 return self._doc_link_template.format(
                     estimator_module=estimator_module, estimator_name=estimator_name
                 )
             return self._doc_link_template.format(
                 **self._doc_link_url_param_generator()
             )
 
@@ -82,22 +82,28 @@
 
 
 def doc_link_url_param_generator(estimator):
     from skrub import __version__
 
     skrub_version = parse_version(__version__)
     if skrub_version.dev is None:
         version_url = f"{skrub_version.major}.{skrub_version.minor}"
     else:
         version_url = "dev"
     estimator_name = estimator.__class__.__name__
     estimator_module = ".".join(
         itertools.takewhile(
             lambda part: not part.startswith("_"),
             estimator.__class__.__module__.split("."),
         )
     )
     return {
         "version": version_url,
         "estimator_module": estimator_module,
         "estimator_name": estimator_name,
     }
+
+
+class _SkrubHTMLDocumentationLinkMixin(_HTMLDocumentationLinkMixin):
+    _doc_link_template = doc_link_template
+    _doc_link_module = doc_link_module
+    _doc_link_url_param_generator = doc_link_url_param_generator

skrub/_table_vectorizer.py

@@ -17,12 +17,7 @@
 from ._datetime_encoder import DatetimeEncoder
 from ._gap_encoder import GapEncoder
 from ._on_each_column import SingleColumnTransformer
-from ._repr import (
-    _HTMLDocumentationLinkMixin,
-    doc_link_module,
-    doc_link_template,
-    doc_link_url_param_generator,
-)
+from ._repr import _SkrubHTMLDocumentationLinkMixin
 from ._select_cols import Drop
 from ._to_datetime import ToDatetime
 from ._to_float32 import ToFloat32
@@ -116,7 +111,9 @@ def _check_transformer(transformer):
     return clone(transformer)
 
 
-class TableVectorizer(_HTMLDocumentationLinkMixin, TransformerMixin, BaseEstimator):
+class TableVectorizer(
+    _SkrubHTMLDocumentationLinkMixin, TransformerMixin, BaseEstimator
+):
     """Transform a dataframe to a numerical (vectorized) representation.
 
     Applies a different transformation to each of several kinds of columns:
@@ -411,10 +408,6 @@ class TableVectorizer(_HTMLDocumentationLinkMixin, TransformerMixin, BaseEstimat
     ValueError: Column 'A' used twice in 'specific_transformers', at indices 0 and 1.
     """  # noqa: E501
 
-    _doc_link_module = doc_link_module
-    _doc_link_template = doc_link_template
-    _doc_link_url_param_generator = doc_link_url_param_generator
-
     def __init__(
         self,
         *,