y_predicted_knn5_es = algorithm.knn_regressor(x_train, y_train, x_test, 5) rmse, y_predicted_knn5 = compare_test(y_test, y_predicted_knn5_es) print('RMSE KNN(5) %.3f' % (rmse)) # KNN10 y_predicted_knn10_es = algorithm.knn_regressor(x_train, y_train, x_test, 10) rmse, y_predicted_knn10 = compare_test(y_test, y_predicted_knn10_es) print('RMSE KNN(10) %.3f' % (rmse)) # SGD y_predicted_sgd_es = algorithm.sgd_regressor(x_train, y_train, x_test) rmse, y_predicted_sgd = compare_test(y_test, y_predicted_sgd_es) print('RMSE SGD %.3f' % (rmse)) # Lasso y_predicted_la_es = algorithm.lasso(x_train, y_train, x_test, normalize=False) rmse, y_predicted_la = compare_test(y_test, y_predicted_la_es) print('RMSE Lasso %.3f' % (rmse)) # titles = ['Y', 'ElasticNet', 'ElasticNet Future', 'KNN5', 'KNN10'] # y_future_en = y_future_en[1] # data = [y_hat_predicted, y_predicted_en, y_future_en, y_predicted_knn5, y_predicted_knn10] titles = ['Y', 'ElasticNet', 'KNN5', 'KNN10', 'SGD', 'Lasso'] data = [y_hat_predicted, y_predicted_en, y_predicted_knn5, y_predicted_knn10, [], y_predicted_la] date_test = date[split + 1:] print('Length date test:' + str(len(date_test))) print('Length data test:' + str(len(y_test))) misc.plot_lines_graph('Stationary, Test Data ', date_test, titles, data)
y_hat_predicted = y_test rmse = compare(y_test, y_hat_predicted) print('RMSE NoPredic %.3f' % (rmse)) # Dummy y_predicted_dummy = x_test[:, -1] rmse = compare(y_test, y_predicted_dummy) print('RMSE Dummy %.3f' % (rmse)) # ElasticNet y_predicted_en, y_future_en = algorithm.elastic_net(x_train, y_train, x_test, y_test,normalize=True) rmse = compare(y_test, y_predicted_en) print('RMSE Elastic %.3f' % (rmse)) # Lasso y_predicted_en = algorithm.lasso(x_train, y_train, x_test, normalize=True) rmse = compare(y_test, y_predicted_en) print('RMSE Lasso %.3f' % (rmse)) titles = ['Y', 'ElasticNet'] data = [y_test, y_predicted_en] date_test = date[split:] print('Length date test:' + str(len(date_test))) print('Length data test:' + str(len(y_test))) print(columns) time_end = datetime.datetime.now() print('End time: %s' % str(time_end.strftime('%Y-%m-%d %H:%M:%S')))
# KNN5 y_hat_predicted = algorithm.knn_regressor(x_train, y_train, x_test, 5) rmse, y_predicted_knn5 = compare_test(test_scaled, y_hat_predicted) print('RMSE KNN(5) %.3f' % (rmse)) # KNN10 y_hat_predicted = algorithm.knn_regressor(x_train, y_train, x_test, 10) rmse, y_predicted_knn10 = compare_test(test_scaled, y_hat_predicted) print('RMSE KNN(10) %.3f' % (rmse)) # SGD y_hat_predicted = algorithm.sgd_regressor(x_train, y_train, x_test) rmse, y_predicted_sgd = compare_test(test_scaled, y_hat_predicted) print('RMSE SGD %.3f' % (rmse)) # Lasso y_hat_predicted = algorithm.lasso(x_train, y_train, x_test, normalize=False) rmse, y_predicted_la = compare_test(test_scaled, y_hat_predicted) print('RMSE Lasso %.3f' % (rmse)) # LSTM y_hat_predicted = algorithm.lstm(x_train, y_train, x_test, batch_size=1, nb_epoch=200, neurons=3) rmse, y_predicted_lstm = compare_test(test_scaled, y_hat_predicted) print('RMSE LSTM %.3f' % (rmse)) titles = ['X', 'Y', 'ElasticNet', 'KNN5', 'KNN10', 'SGD', 'Lasso','LSTM'] data = [raw_values[split:-1], y_predicted_real, y_predicted_en, y_predicted_knn5, y_predicted_knn10, y_predicted_sgd, y_predicted_la,y_predicted_lstm] misc.plot_lines_graph('Stationary - Normalization, Test Data ', date[split:], titles, data)