Local Regression and Likelihood, Figure 2.5. Using different smoothing parameters.
0001 % Local Regression and Likelihood, Figure 2.5. 0002 % 0003 % Using different smoothing parameters. 0004 0005 load ethanol; 0006 0007 alp = [0.2 0.4 0.6 0.8]; 0008 f=['a','b','c','d']; 0009 for (i=1:4) 0010 figure('Name', sprintf( 'figure2_5%c: Smoothing parameters', f(i) ) ); 0011 fit = locfit(E,NOx,'alpha',alp(i)); 0012 res = residuals(fit); 0013 fit2 = locfit(E,res,'alpha',0.2); 0014 lfplot(fit2); 0015 title(strcat('a = ',num2str(alp(i)))); 0016 ylabel('Residual'); 0017 hold on; 0018 plot([min(E) max(E)], [0 0], ':'); 0019 hold off; 0020 end;