Fixes UnboundLocalError: local variable 'cv_pipeline' referenced before assignment when error in automl search

docs/source/api_reference.rst

@@ -179,6 +179,7 @@ Transformers are components that take in data as input and output transformed da
     OneHotEncoder
     PerColumnImputer
     SimpleImputer
+    Imputer
     StandardScaler
     RFRegressorSelectFromModel
     RFClassifierSelectFromModel

docs/source/release_notes.rst

@@ -14,6 +14,7 @@ Release Notes
         * Removed incorrect parameter passed to pipeline classes in `_add_baseline_pipelines` :pr:`941`
         * Added universal error for calling `predict`, `predict_proba`, `transform`, and `feature_importances` before fitting :pr:`969`, :pr:`994`
         * Made `TextFeaturizer` component and pip dependencies `featuretools` and `nlp_primitives` optional :pr:`976`
+        * Fixed UnboundLocalError for`cv_pipeline` when automl search errors :pr:`996`
     * Changes
         * Moved `get_estimators ` to `evalml.pipelines.components.utils` :pr:`934`
         * Modified Pipelines to raise `PipelineScoreError` when they encounter an error during scoring :pr:`936`

evalml/automl/automl_search.py

@@ -540,6 +540,7 @@ def _compute_cv_scores(self, pipeline, X, y):
             X_train, X_test = X.iloc[train], X.iloc[test]
             y_train, y_test = y.iloc[train], y.iloc[test]
             objectives_to_score = [self.objective] + self.additional_objectives
+            cv_pipeline = None
             try:
                 X_threshold_tuning = None
                 y_threshold_tuning = None

evalml/pipelines/prediction_explanations/__init__.py

@@ -1,2 +1,2 @@
 # flake8:noqa
-from .explainers import explain_prediction
+from .explainers import explain_prediction

evalml/tests/automl_tests/test_automl.py

@@ -888,3 +888,16 @@ def test_catch_keyboard_interrupt(mock_fit, mock_score, mock_input,
     automl.search(X, y)
 
     assert len(automl._results['pipeline_results']) == number_results
+
+
+@patch('evalml.automl.automl_search.train_test_split')
+@patch('evalml.pipelines.BinaryClassificationPipeline.score')
+@patch('evalml.pipelines.BinaryClassificationPipeline.fit')
+def test_error_during_train_test_split(mock_fit, mock_score, mock_train_test_split, X_y_binary):
+    X, y = X_y_binary
+    mock_score.return_value = {'Log Loss Binary': 1.0}
+    mock_train_test_split.side_effect = RuntimeError()
+    automl = AutoMLSearch(problem_type='binary', objective='accuracy_binary', max_pipelines=2, optimize_thresholds=True)
+    automl.search(X, y)
+    for pipeline in automl.results['pipeline_results'].values():
+        assert np.isnan(pipeline['score'])