scikit-learn Cookbook , Second Edition - Second Edition

How to do it...

1. Import both LinearRegression and TheilSenRegressor. Score the estimators using the original line as the testing set, y_truth:

from sklearn.linear_model import LinearRegression, TheilSenRegressorfrom sklearn.metrics import r2_score, mean_absolute_errornamed_estimators = [('OLS ', LinearRegression()), ('TSR ', TheilSenRegressor())]for num_index, est in enumerate(named_estimators):    y_pred = est[1].fit(x_vals.reshape(-1, 1),y_noisy).predict(x_vals.reshape(-1, 1))    print (est[0], "R-squared: ", r2_score(y_truth, y_pred), "Mean Absolute Error", mean_absolute_error(y_truth, y_pred))    plt.plot(x_vals, y_pred, label=est[0])('OLS ', 'R-squared: ', 0.17285546630270587, 'Mean Absolute Error', 44.099173357335729) ('TSR ', 'R-squared: ', ...