Merge pull request #318 from TeamHG-Memex/catboost-docs-cleanup

Catboost docs cleanup
TeamHG-Memex · Jun 27, 2019 · 716a4d7 · 716a4d7
2 parents e19139a + aa32308
commit 716a4d7
Show file tree

Hide file tree

Showing 2 changed files with 20 additions and 12 deletions.
diff --git a/README.rst b/README.rst
@@ -42,7 +42,8 @@ It provides support for the following machine learning frameworks and packages:
 * LightGBM_ - show feature importances and explain predictions of
   LGBMClassifier and LGBMRegressor.
 
-* CatBoost_ - show feature importances and explain predictions of CatBoostClassifier and CatBoostRegressor.
+* CatBoost_ - show feature importances of CatBoostClassifier,
+  CatBoostRegressor and catboost.CatBoost.
 
 * lightning_ - explain weights and predictions of lightning classifiers and
   regressors.

diff --git a/eli5/catboost.py b/eli5/catboost.py
@@ -6,8 +6,8 @@
 from eli5.explain import explain_weights
 from eli5._feature_importances import get_feature_importance_explanation
 
-DESCRIPTION_CATBOOST = """CatBoost feature importances; values are numbers 0 <= x <= 1;
-all values sum to 1."""
+DESCRIPTION_CATBOOST = """CatBoost feature importances; 
+values are numbers 0 <= x <= 1; all values sum to 1."""
 
 @explain_weights.register(catboost.CatBoost)
 @explain_weights.register(catboost.CatBoostClassifier)
@@ -20,8 +20,8 @@ def explain_weights_catboost(catb,
                              pool=None
                              ):
     """
-    Return an explanation of an CatBoost estimator (CatBoostClassifier, CatBoost, CatBoostRegressor)
-    as feature importances.
+    Return an explanation of an CatBoost estimator (CatBoostClassifier,
+    CatBoost, CatBoostRegressor) as feature importances.
 
     See :func:`eli5.explain_weights` for description of
     ``top``, ``feature_names``,
@@ -34,13 +34,16 @@ def explain_weights_catboost(catb,
     :param 'importance_type' : str, optional
         A way to get feature importance. Possible values are:
         
-        - 'PredictionValuesChange' - The individual importance values for each of the input features.
-          (default)
-        - 'LossFunctionChange' - The individual importance values for each of the input features for ranking metrics (requires training data to be passed  or a similar dataset with Pool)
+        - 'PredictionValuesChange' (default) - The individual importance
+          values for each of the input features.
+        - 'LossFunctionChange' - The individual importance values for
+          each of the input features for ranking metrics
+          (requires training data to be passed or a similar dataset with Pool)
 
     :param 'pool' : catboost.Pool, optional
-        To be passed if explain_weights_catboost has importance_type set to LossFunctionChange.
-        The catboost feature_importances uses the Pool datatype to calculate the parameter for the specific importance_type.
+        To be passed if explain_weights_catboost has importance_type set
+        to LossFunctionChange. The catboost feature_importances uses the Pool
+        datatype to calculate the parameter for the specific importance_type.
     """
     is_regression = _is_regression(catb)
     catb_feature_names = catb.feature_names_
@@ -69,10 +72,14 @@ def _catb_feature_importance(catb, importance_type, pool=None):
             fs = catb.get_feature_importance(data=pool, type=importance_type)
         else:
             raise ValueError(
-                'importance_type: "LossFunctionChange" requires catboost.Pool datatype to be passed with parameter pool to calculate metric. Either no datatype or invalid datatype was passed'
+                'importance_type: "LossFunctionChange" requires catboost.Pool '
+                'datatype to be passed with parameter pool to calculate '
+                'metric. Either no datatype or invalid datatype was passed'
             )
     else:
         raise ValueError(
-            'Only two importance_type "PredictionValuesChange" and "LossFunctionChange" are supported. Invalid Parameter {} for importance_type'.format(importance_type))
+            'Only two importance_type "PredictionValuesChange" '
+            'and "LossFunctionChange" are supported. Invalid Parameter '
+            '{} for importance_type'.format(importance_type))
     all_features = np.array(fs, dtype=np.float32)
     return all_features/all_features.sum()