Standard deviation of a 3-d matrix

hi I have a matrix with size(y)=(200*1*4). I want to calculate the std for each (200*1) vector separately. When I used "std(y(y>0)), it gave me an value but I want in fact 4 values for each 200*1 vector size. How can I write the code to give me the std for each inner vector without "for",and ....

Thanks for your help

Answer 1

[Kye Taylor]

If y is 200-by-1-by-4, you can get rid of the singleton dimension using

y2D = squeeze(y);

then compute the 4 different standard deviations using

the4Sigmas = std(y2D);

[Andrea]

thanks, it was so useful. But now another question is :

I want to calculate the std for > 0. the4Sigmas = std(y2D); give me (1 by 4) value for std but when I wrote the4Sigmas = std(y2D(y2D>0)); it gave me gain a single values while I want the std to be like before (1 by 4).

Can you please help me again?

[Kye Taylor]

Sure, but next time, ask a brand new question.

I would then use a for loop

sigmasPositive = zeros(1,4);
for j = 1:size(y2D,2)
  idxOfInterest = y2D(:,j) > 0;
  sigmasPositive(j) = std(y2D(idxOfInterest,j));
end

or you can avoid the loop using

z = mat2cell(y2D,size(y2D,1),ones(1,size(y2D,2)));
sigmasPositiveNoLoop = cellfun(@(c)std(c(c>0)),z);

[Andrea]

Thanks a lot. I've already done that. but a little late. It works perfectly.