Update how-to-machine-learning-interpretability-automl.md

Author: Aniththa

articles/machine-learning/how-to-machine-learning-interpretability-automl.md

@@ -39,7 +39,7 @@ Retrieve the explanation from the `best_run`, which includes explanations for en
 You can use `ExplanationClient` to download the engineered feature explanations from the artifact store of the `best_run`. To get the explanation for the raw features set `raw=True`.
 
 ```python
-from azureml.contrib.interpret.explanation.explanation_client import ExplanationClient
+from azureml.explain.model._internal.explanation_client import ExplanationClient
 
 client = ExplanationClient.from_run(best_run)
 engineered_explanations = client.download_model_explanation(raw=False)
@@ -53,7 +53,7 @@ When you compute model explanations and visualize them, you're not limited to an
 ### Retrieve any other AutoML model from training
 
 ```python
-automl_run, fitted_model = local_run.get_output(metric='r2_score')
+automl_run, fitted_model = local_run.get_output(metric='accuracy')
 ```
 
 ### Set up the model explanations
@@ -67,7 +67,7 @@ Use `automl_setup_model_explanations` to get the engineered and raw feature expl
 The `automl_explainer_setup_obj` contains all the structures from above list.
 
 ```python
-from azureml.train.automl.runtime.automl_explain_utilities import AutoMLExplainerSetupClass, automl_setup_model_explanations
+from azureml.train.automl.runtime.automl_explain_utilities import automl_setup_model_explanations
 
 automl_explainer_setup_obj = automl_setup_model_explanations(fitted_model, X=X_train, 
                                                              X_test=X_test, y=y_train, 
@@ -82,16 +82,16 @@ To generate an explanation for AutoML models, use the `MimicWrapper` class. You
 - Your workspace
 - A LightGBM model, which acts as a surrogate to the `fitted_model` automated ML model
 
-The MimicWrapper also takes the `automl_run` object where the raw and engineered explanations will be uploaded.
+The MimicWrapper also takes the `automl_run` object where the engineered explanations will be uploaded.
 
 ```python
 from azureml.explain.model.mimic.models.lightgbm_model import LGBMExplainableModel
 from azureml.explain.model.mimic_wrapper import MimicWrapper
 
 # Initialize the Mimic Explainer
-explainer = MimicWrapper(ws, automl_explainer_setup_obj.automl_estimator, LGBMExplainableModel,
+explainer = MimicWrapper(ws, automl_explainer_setup_obj.automl_estimator, LGBMExplainableModel, 
                          init_dataset=automl_explainer_setup_obj.X_transform, run=automl_run,
-                         features=automl_explainer_setup_obj.engineered_feature_names,
+                         features=automl_explainer_setup_obj.engineered_feature_names, 
                          feature_maps=[automl_explainer_setup_obj.feature_map],
                          classes=automl_explainer_setup_obj.classes)
 ```
@@ -101,27 +101,8 @@ explainer = MimicWrapper(ws, automl_explainer_setup_obj.automl_estimator, LGBMEx
 You can call the `explain()` method in MimicWrapper with the transformed test samples to get the feature importance for the generated engineered features. You can also use `ExplanationDashboard` to view the dashboard visualization of the feature importance values of the generated engineered features by automated ML featurizers.
 
 ```python
-from azureml.contrib.interpret.visualize import ExplanationDashboard
-engineered_explanations = explainer.explain(['local', 'global'],              
-                                            eval_dataset=automl_explainer_setup_obj.X_test_transform)
-
+engineered_explanations = explainer.explain(['local', 'global'], eval_dataset=automl_explainer_setup_obj.X_test_transform)
 print(engineered_explanations.get_feature_importance_dict())
-ExplanationDashboard(engineered_explanations, automl_explainer_setup_obj.automl_estimator, automl_explainer_setup_obj.X_test_transform)
-```
-
-### Use Mimic Explainer for computing and visualizing raw feature importance
-
-You can call the `explain()` method in MimicWrapper again with the transformed test samples and setting `get_raw=True` to get the feature importance for the raw features. You can also use `ExplanationDashboard` to view the dashboard visualization of the feature importance values of the raw features.
-
-```python
-from azureml.contrib.interpret.visualize import ExplanationDashboard
-
-raw_explanations = explainer.explain(['local', 'global'], get_raw=True, 
-                                     raw_feature_names=automl_explainer_setup_obj.raw_feature_names,
-                                     eval_dataset=automl_explainer_setup_obj.X_test_transform)
-
-print(raw_explanations.get_feature_importance_dict())
-ExplanationDashboard(raw_explanations, automl_explainer_setup_obj.automl_pipeline, automl_explainer_setup_obj.X_test_raw)
 ```
 
 ### Interpretability during inference