Eliminating a FOR loop

I am trying to eliminate a FOR loop to improve speed. The problem I am working with involves evaluating each element of an array with each element of a matrix. Below is a simplified example of what I am trying to do.

I have two matrices:

A = [a11 a12 a13;a21 a22 a23; a31 a32 a33];

B = [b1 b2 b3 b4];

What I would like to do is evaluate each element of B against each element of A. For example

for i=1:1:4

 [I,J] = find(B(i)==A(:,:));

end

Is there a way to do this without a FOR loop.

Thanks in advance,

Scott

[Guillaume]

Your example overwrites I and J on each pass of the loop, so it's not clear what you want out. Did you mean:

 [I{i} J{i}] = find(B(i)==A);

Answer 1

[Matt J]

L=bsxfun(@eq,A(:),B(:).');
 L=reshape(L,[size(A),length(B)]);

[I,J,~]=ind2sub(size(L),find(L));

[Scott]

Matt,

Thank you very much. Your answer was exactly what I was looking for.

Answer 2

[Iain]

The function you might be looking for is probably

 intersect

Answer 3

[Pierre Benoit]

[~,idx] = ismember(B,A);
[I,J] = ind2sub(size(A),idx);

Answer 4

[Guillaume]

You can use arrayfun or bsxfun to replace a loop. It's not necessary faster but may make the intent of the code clearer. In this case, assuming you want I and J as cell arrays, not so:

 [I, J] = arrayfun(@(b) ind2sub(size(A), find(b==A)), B, 'uni', false);

Answer 5

[Andrei Bobrov]

A = randi(10,5,6);
B = randi(10,4,1);

[i1,j1] = find(bsxfun(@eq,A,reshape(B,1,1,[])));
out_ij = [i1, rem(j1 - 1,size(A,2)) + 1];