added explanation again

jakedubb · jakedubb · commit be00e66acab8 · 2019-11-20T13:55:18.000-08:00
diff --git a/code/training/train.py b/code/training/train.py
@@ -30,9 +30,15 @@
 from sklearn.linear_model import Ridge
 from sklearn.metrics import mean_squared_error
 from sklearn.model_selection import train_test_split
+
 from sklearn.externals import joblib
 import numpy as np
 
+from interpret.ext.blackbox import TabularExplainer
+from azureml.contrib.explain.model.explanation.explanation_client import (
+ ExplanationClient
+)
+
 parser = argparse.ArgumentParser("train")
 parser.add_argument(
     "--release_id",
@@ -57,6 +63,7 @@
 run = Run.get_context()
 exp = run.experiment
 ws = run.experiment.workspace
+client = ExplanationClient.from_run(run)
 
 X, y = load_diabetes(return_X_y=True)
 columns = ["age", "gender", "bmi", "bp", "s1", "s2", "s3", "s4", "s5", "s6"]
@@ -77,6 +84,19 @@
 preds = reg.predict(data["test"]["X"])
 run.log("mse", mean_squared_error(preds, data["test"]["y"]))
 
+# create an explainer to validate or debug the model
+tabular_explainer = TabularExplainer(reg,
+                                     initialization_examples=X_train,
+                                     features=columns)
+# explain overall model predictions (global explanation)
+# passing in test dataset for evaluation examples
+
+global_explanation = tabular_explainer.explain_global(X_test)
+
+# uploading model explanation data for storage or visualization
+comment = 'Global explanation on of Diabetes Regression'
+client.upload_model_explanation(global_explanation, comment=comment)
+
 with open(model_name, "wb") as file:
     joblib.dump(value=reg, filename=model_name)
 
diff --git a/ml_service/pipelines/build_train_pipeline.py b/ml_service/pipelines/build_train_pipeline.py
@@ -52,7 +52,12 @@ def main():
                         'scikit-learn', 'tensorflow', 'keras'],
         pip_packages=['azure', 'azureml-core',
                       'azure-storage',
-                      'azure-storage-blob'])
+                      'azure-storage-blob', 'azureml-defaults',
+                      'azureml-contrib-interpret',
+                      'azureml-telemetry',
+                      'azureml-interpret',
+                      'sklearn-pandas',
+                      'azureml-dataprep'])
     )
     run_config.environment.docker.enabled = True
 
