aae3c2048e8fb54452087b2fa9ef66cb133a8a23,auto_ml/predictor.py,Predictor,ml_for_analytics,#Predictor#Any#Any#Any#Any#Any#,87

Before Change

                trained_feature_names = self.trained_pipeline.named_steps["dv"].get_feature_names()

                if model_name == "LogisticRegression":
                    print("trained_feature_names")
                    print(trained_feature_names)
                    trained_coefficients = self.trained_pipeline.named_steps["final_model"].model.coef_[0]
                    print("trained_coefficients")
                    print(trained_coefficients)

After Change

                        summary_tuple = (feature_name, trained_coefficients[col_idx], potential_impact)
                        feature_summary.append(summary_tuple)

                    sorted_feature_summary = sorted(feature_summary, key=lambda x: x[2])

                    for summary in sorted_feature_summary:
                        print(summary[0] + ": " + str(summary[1]))
                        print("The potential impact of this feature is: " + str(summary[2]))


        else:
            ppl.fit(X, y)
            self.trained_pipeline = ppl

Instances