rename the scikit-learn imports with SK prefix to make them public when scikit-learn not installed

Author: paulbkoch

python/interpret-core/interpret/glassbox/_aplr.py

@@ -17,7 +17,7 @@ class _SeriesType:
         pass
 
 
-from ..utils._scikit import _ClassifierMixin, _RegressorMixin
+from ..utils._scikit import SKClassifierMixin, SKRegressorMixin
 from ..api.base import LocalExplainer, GlobalExplainer
 from ..api.templates import FeatureValueExplanation
 from ..utils._clean_simple import clean_dimensions
@@ -46,7 +46,7 @@ def __init__(self, *args, **kwargs):
 
 
 class APLRRegressor(
-    _RegressorMixin, LocalExplainer, GlobalExplainer, APLRRegressorNative
+    SKRegressorMixin, LocalExplainer, GlobalExplainer, APLRRegressorNative
 ):
     """APLR Regressor."""
 
@@ -341,7 +341,7 @@ def __init__(self, *args, **kwargs):
 
 
 class APLRClassifier(
-    _ClassifierMixin, LocalExplainer, GlobalExplainer, APLRClassifierNative
+    SKClassifierMixin, LocalExplainer, GlobalExplainer, APLRClassifierNative
 ):
     """APLR Classifier."""
 

python/interpret-core/interpret/glassbox/_decisiontree.py

@@ -7,10 +7,10 @@
 
 import numpy as np
 from ..utils._scikit import (
-    _ClassifierMixin,
-    _RegressorMixin,
-    _BaseEstimator,
-    _NotFittedError,
+    SKClassifierMixin,
+    SKRegressorMixin,
+    SKBaseEstimator,
+    SKNotFittedError,
     _is_classifier,
 )
 from ..api.base import LocalExplainer, GlobalExplainer, BaseExplanation
@@ -222,7 +222,7 @@ def _weight_nodes_feature(self, nodes, feature_name):
         return new_nodes
 
 
-class BaseShallowDecisionTree(LocalExplainer, GlobalExplainer, _BaseEstimator):
+class BaseShallowDecisionTree(LocalExplainer, GlobalExplainer, SKBaseEstimator):
     """Shallow Decision Tree (low depth).
 
     Currently wrapper around DecisionTreeClassifier or DecisionTreeRegressor in scikit-learn.
@@ -318,7 +318,7 @@ def predict(self, X):
         """
 
         if not hasattr(self, "n_features_in_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -347,7 +347,7 @@ def explain_global(self, name=None):
         """
 
         if not hasattr(self, "n_features_in_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -398,7 +398,7 @@ def explain_local(self, X, y=None, name=None):
         """
 
         if not hasattr(self, "n_features_in_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -570,7 +570,7 @@ def __sklearn_tags__(self):
         return tags
 
 
-class RegressionTree(_RegressorMixin, BaseShallowDecisionTree):
+class RegressionTree(SKRegressorMixin, BaseShallowDecisionTree):
     """Regression tree with shallow depth."""
 
     def __init__(self, feature_names=None, feature_types=None, max_depth=3, **kwargs):
@@ -619,7 +619,7 @@ def fit(self, X, y, sample_weight=None, check_input=True):
         )
 
 
-class ClassificationTree(_ClassifierMixin, BaseShallowDecisionTree):
+class ClassificationTree(SKClassifierMixin, BaseShallowDecisionTree):
     """Classification tree with shallow depth."""
 
     def __init__(self, feature_names=None, feature_types=None, max_depth=3, **kwargs):
@@ -678,7 +678,7 @@ def predict_proba(self, X):
         """
 
         if not hasattr(self, "n_features_in_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 

python/interpret-core/interpret/glassbox/_ebm/_ebm.py

@@ -18,10 +18,10 @@
 import numpy as np
 from joblib import Parallel, delayed
 from ...utils._scikit import (
-    _BaseEstimator,
-    _ClassifierMixin,
-    _RegressorMixin,
-    _NotFittedError,
+    SKBaseEstimator,
+    SKClassifierMixin,
+    SKRegressorMixin,
+    SKNotFittedError,
     _is_classifier,
     _is_regressor,
 )
@@ -283,7 +283,7 @@ def clean_interactions(interactions, n_features_in):
         return interactions
 
 
-class BaseEBM(LocalExplainer, GlobalExplainer, _BaseEstimator):
+class BaseEBM(LocalExplainer, GlobalExplainer, SKBaseEstimator):
     """Base class for all EBMs.  Do not instantiate directly."""
 
     n_features_in_: int
@@ -1713,7 +1713,7 @@ def to_jsonable(self, detail="all"):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -1735,7 +1735,7 @@ def to_json(self, file, detail="all", indent=2):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -1796,7 +1796,7 @@ def to_excel_exportable(self, file):
         """
 
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -1817,7 +1817,7 @@ def to_excel(self, file):
         """
 
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -1837,7 +1837,7 @@ def _predict_score(self, X, init_score=None):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -1872,7 +1872,7 @@ def eval_terms(self, X):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -1901,7 +1901,7 @@ def explain_global(self, name=None):
             name = gen_name_from_class(self)
 
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2187,7 +2187,7 @@ def explain_local(self, X, y=None, name=None, init_score=None):
         # Values are the model graph score per respective term.
 
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2329,7 +2329,7 @@ def term_importances(self, importance_type="avg_weight"):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2385,7 +2385,7 @@ def monotonize(self, term, increasing="auto", passthrough=0.0):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2486,7 +2486,7 @@ def remove_terms(self, terms):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2540,7 +2540,7 @@ def remove_features(self, features):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2597,7 +2597,7 @@ def sweep(self, terms=True, bins=True, features=False):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2652,7 +2652,7 @@ def scale(self, term, factor):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2688,7 +2688,7 @@ def predict_with_uncertainty(self, X, init_score=None):
         """
 
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2723,7 +2723,7 @@ def predict_with_uncertainty(self, X, init_score=None):
 
     def _multinomialize(self, passthrough=0.0):
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2776,7 +2776,7 @@ def _multinomialize(self, passthrough=0.0):
 
     def _ovrize(self, passthrough=0.0):
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2829,7 +2829,7 @@ def _ovrize(self, passthrough=0.0):
 
     def _binarize(self, passthrough=0.0):
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2878,7 +2878,7 @@ def __sklearn_tags__(self):
         return tags
 
 
-class EBMClassifierMixin(_ClassifierMixin):
+class EBMClassifierMixin(SKClassifierMixin):
     """Mixin class for EBM classifiers.
 
     Provides predict, predict_proba, decision_function, and reorder_classes methods.
@@ -2902,7 +2902,7 @@ def predict_proba(self, X, init_score=None):
         """
 
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2940,7 +2940,7 @@ def decision_function(self, X, init_score=None):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -2974,7 +2974,7 @@ def predict(self, X, init_score=None):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -3013,7 +3013,7 @@ def reorder_classes(self, classes):
 
         """
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 
@@ -3063,7 +3063,7 @@ def reorder_classes(self, classes):
         return self
 
 
-class EBMRegressorMixin(_RegressorMixin):
+class EBMRegressorMixin(SKRegressorMixin):
     """Mixin class for EBM regressors.
 
     Provides the regression predict method.
@@ -3088,7 +3088,7 @@ def predict(self, X, init_score=None):
         """
 
         if not hasattr(self, "bins_"):
-            raise _NotFittedError(
+            raise SKNotFittedError(
                 "This model has not been fitted yet. Call 'fit' first."
             )
 

python/interpret-core/interpret/glassbox/_ebm/_research/_group_importance.py

@@ -8,7 +8,7 @@
 
 import numpy as np
 import plotly.express as px
-from ....utils._scikit import _NotFittedError
+from ....utils._scikit import SKNotFittedError
 
 
 def compute_group_importance(term_list, ebm, X, contributions=None):
@@ -24,7 +24,7 @@ def compute_group_importance(term_list, ebm, X, contributions=None):
         float: term_list's group importance
     """
     if not hasattr(ebm, "bins_"):
-        raise _NotFittedError("This model has not been fitted yet. Call 'fit' first.")
+        raise SKNotFittedError("This model has not been fitted yet. Call 'fit' first.")
 
     if contributions is None:
         contributions = ebm.eval_terms(X)
@@ -116,7 +116,7 @@ def append_group_importance(
         EBMExplanation: A global explanation with the group importance appended to it
     """
     if not hasattr(ebm, "bins_"):
-        raise _NotFittedError("This model has not been fitted yet. Call 'fit' first.")
+        raise SKNotFittedError("This model has not been fitted yet. Call 'fit' first.")
 
     if global_exp is not None:
         if global_exp.explanation_type != "global":