fix indexing error when creating a filtered error analysis tree view …

…with a dataset that contains categoricals
microsoft · Apr 5, 2023 · ea4f305 · ea4f305
1 parent dd7d6d5
commit ea4f305
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 2 deletions.
diff --git a/erroranalysis/erroranalysis/_internal/surrogate_error_tree.py b/erroranalysis/erroranalysis/_internal/surrogate_error_tree.py
@@ -330,7 +330,7 @@ def get_surrogate_booster_local(filtered_df, analyzer, is_model_analyzer,
         else:
             string_indexed_data = analyzer.string_indexed_data
         for idx, c_i in enumerate(analyzer.categorical_indexes):
-            input_data[:, c_i] = string_indexed_data[row_index, idx]
+            input_data[:, c_i] = string_indexed_data[:, idx]
     dataset_sub_features = input_data[:, indexes]
 
     categorical_info = get_categorical_info(analyzer,

diff --git a/erroranalysis/tests/test_surrogate_error_tree.py b/erroranalysis/tests/test_surrogate_error_tree.py
@@ -66,12 +66,24 @@ def test_surrogate_error_tree_iris(self, analyzer_type):
     def test_surrogate_error_tree_int_categorical(self, analyzer_type):
         X_train, X_test, y_train, y_test, categorical_features = \
             create_adult_census_data()
-
         model = create_kneighbors_classifier(X_train, y_train)
 
         run_error_analyzer(model, X_test, y_test, list(X_train.columns),
                            analyzer_type, categorical_features)
 
+    @pytest.mark.parametrize('analyzer_type', [AnalyzerType.MODEL,
+                                               AnalyzerType.PREDICTIONS])
+    def test_surrogate_error_tree_categorical_filtered(self, analyzer_type):
+        X_train, X_test, y_train, y_test, categorical_features = \
+            create_adult_census_data()
+        model = create_kneighbors_classifier(X_train, y_train)
+        filters = [{'arg': [40],
+                    'column': 'Age',
+                    'method': 'less and equal'}]
+        run_error_analyzer(model, X_test, y_test, list(X_train.columns),
+                           analyzer_type, categorical_features,
+                           filters=filters)
+
     def test_large_data_surrogate_error_tree(self):
         # validate tree trains quickly for large data
         X_train, y_train, X_test, y_test, _ = \