python خانه پایتون گروه آموزشی برنامه نویسی

# -------------------------------------------------- # 10. Create the complete machine-learning pipeline # -------------------------------------------------- model = Pipeline( steps=[ ( "preprocessing", preprocessor, ), ( "regressor", LinearRegression(), ), ] ) # -------------------------------------------------- # 11. Train the pipeline # -------------------------------------------------- model.fit( X_train, y_train, ) # -------------------------------------------------- # 12. Make predictions # -------------------------------------------------- y_pred = model.predict(X_test) # -------------------------------------------------- # 13. Evaluate the model # -------------------------------------------------- mae = mean_absolute_error( y_test, y_pred, ) mse = mean_squared_error( y_test, y_pred, ) rmse = np.sqrt(mse) r2 = r2_score( y_test, y_pred, ) print(f"MAE: {mae:,.2f}") print(f"MSE: {mse:,.2f}") print(f"RMSE: {rmse:,.2f}") print(f"R²: {r2:.4f}") # -------------------------------------------------- # 14. Create a prediction report # -------------------------------------------------- results = pd.DataFrame( { "actual_price": y_test, "predicted_price": y_pred, } ) results["residual"] = ( results["actual_price"] - results["predicted_price"] ) results["absolute_error"] = ( results["residual"].abs() ) print("\nPrediction results:") print(results.head(10)) # -------------------------------------------------- # 15. Inspect the fitted preprocessing steps # -------------------------------------------------- fitted_preprocessor = ( model.named_steps["preprocessing"] ) fitted_regressor = ( model.named_steps["regressor"] ) fitted_scaler = ( fitted_preprocessor .named_transformers_["num"] .named_steps["scaler"] ) fitted_encoder = ( fitted_preprocessor .named_transformers_["cat"] .named_steps["onehot"] ) transformed_feature_names = ( fitted_preprocessor .get_feature_names_out() ) print("\nScaler means:") print(fitted_scaler.mean_) print("\nScaler standard deviations:") print(fitted_scaler.scale_) print("\nEncoder categories:") print(fitted_encoder.categories_) print("\nRegression intercept:") print(fitted_regressor.intercept_) print("\nRegression coefficients:") print(fitted_regressor.coef_) # -------------------------------------------------- # 16. Inspect the transformed training matrix # -------------------------------------------------- X_train_transformed = ( fitted_preprocessor.transform(X_train) ) X_train_transformed_df = pd.DataFrame( X_train_transformed, columns=transformed_feature_names, index=X_train.index, ) print("\nTransformed training data:") print(X_train_transformed_df.head()) # -------------------------------------------------- # 17. Compare with NumPy least squares # -------------------------------------------------- X_design = np.column_stack( [ np.ones( X_train_transformed.shape[0] ), X_train_transformed, ] ) beta, *_ = np.linalg.lstsq( X_design, y_train.to_numpy(), rcond=None, ) print("\nNumPy intercept:") print(beta[0]) print("\nNumPy coefficients:") print(beta[1:]) print( "\nIntercepts are close:", np.allclose( beta[0], fitted_regressor.intercept_, ), ) print( "Coefficients are close:", np.allclose( beta[1:], fitted_regressor.coef_, ), ) # -------------------------------------------------- # 18. Predict the price of a new car # -------------------------------------------------- new_car = pd.DataFrame( { "mileage": [50_000], "age": [4], "engine_size": [2.0], "brand": ["B"], "automatic": ["yes"], } ) predicted_price = model.predict( new_car )[0] print( "\nPredicted price for the new car:", f"{predicted_price:,.2f}", )