Problem in using function handle?

I have written a function 'PLSR_A', which works fine. The function is as follows

[beta,yfit,q,RMSE] = PLSR_A(X,Y,npc)

I want to create a function handle in matlab 'crossval' function which will compute the cross-validation error for leave one out. Therefore, I defined a function handle as follows

fun = @(xT,yT,xt,yt) ((yt-(([1 xt])*PLSR_A(xT,yT,npc)))^2);

But when I compute fun with different 'npc' values it gives same output, however, using the function PLSR_A directly, gives different outputs for different 'npc' (which is correct).

I have read the documentation for function handle but couldn't find the problem, can anyone help?

Answer 1

[Adam]

npc is a fixed input to that function handle, it will be set at the time the function handle is created as a constant. If you want it to be one of your input variables you need:

fun = @(xT,yT,xt,yt,npc) ((yt-(([1 xt])*PLSR_A(xT,yT,npc)))^2);

[Abhinav]

Thanks a lot Adam!