


Computes and plots the -2*AIC for local fits with different smoothing parameters. Usage: g=aicplot(alpha,varargin) The first argument to aicplot(), alpha, should be a matrix with one or two columns (first column = nearest neighbor component, second column = constant component). Each row of this matrix is, in turn, passed as the 'alpha' argument to aic() (and locfit()). The results are stored in a matrix, and aic score ploted against the degrees of freedom.


0001 function g=aicplot(alpha,varargin) 0002 % Computes and plots the -2*AIC 0003 % for local fits with different smoothing parameters. 0004 % 0005 % Usage: g=aicplot(alpha,varargin) 0006 % 0007 % The first argument to aicplot(), alpha, should be a matrix with one 0008 % or two columns (first column = nearest neighbor component, second 0009 % column = constant component). Each row of this matrix is, in turn, 0010 % passed as the 'alpha' argument to aic() (and locfit()). The results 0011 % are stored in a matrix, and aic score ploted against the degrees of 0012 % freedom. 0013 0014 k = size(alpha,1); 0015 z = zeros(k,4); 0016 0017 for i=1:k 0018 z(i,:) = aic(varargin{:},'alpha',alpha(i,:)); 0019 end; 0020 0021 plot(z(:,3),z(:,4)); 0022 xlabel('Fitted DF'); 0023 ylabel('AIC'); 0024 0025 g = [alpha z]; 0026 return;