然後稍微檢查一下 tcl、tck 的版本與路徑

				% Example on how to use LDS on the g50c dataset								load g50c;								for i=1:10				  Xl = X(idxLabs(i,:),:)';  				  Xu = X(idxUnls(i,:),:)';				  Yl = y(idxLabs(i,:));				  Yu = y(idxUnls(i,:));				  				  opt.C = 0.1; % For this dataset, the classes overlap and it's				               % better to have a small value of C				  rho = 1;     % Anyvalue smaller than 3 is good (cf figure 5)				  Yp = lds(Xl,Xu,Yl,rho,opt);				  				  te(i) = mean( Yp.*Yu < 0 );				  fprintf('    -> Test error after %d split(s): %f\n\n',i,mean(te));				end;