% Local Regression and Likelihood, Figure 2.5. % % Using different smoothing parameters. load ethanol; alp = [0.2 0.4 0.6 0.8]; f=['a','b','c','d']; for (i=1:4) figure('Name', sprintf( 'figure2_5%c: Smoothing parameters', f(i) ) ); fit = locfit(E,NOx,'alpha',alp(i)); res = residuals(fit); fit2 = locfit(E,res,'alpha',0.2); lfplot(fit2); title(strcat('a = ',num2str(alp(i)))); ylabel('Residual'); hold on; plot([min(E) max(E)], [0 0], ':'); hold off; end;