Row/column-wise logical indexing

I have a matrix

A=[1 2 3;

4 5 6;

7 8 9];

where each row represents a group of numbers. I have a logical matrix

L = [ 0 0 0 1 1 1 1 1 1;

1 1 1 0 0 0 1 1 1;

1 1 1 1 1 1 0 0 0]

where each column represents the groups of A I want to extract.

For example,

A(L(:,1),:) gives me

[4 5 6;

7 8 9]

A(L(:,4),:) gives me

[1 2 3;

7 8 9]

Is there a way to extract all such pairs of rows of A in one vectorized step instead of looping over the columns of L?

Answer 1

[Akira Agata]

How about the following solution?

B = arrayfun(@(k) A(L(:,k),:), 1:size(L,2), 'UniformOutput', false);

Then, your can obtain cell array B where B{1} = A(L(:,1),:), B{2} = A(L(:,2),:), ... B{9} = A(L(:,9),:).

[Angshuman Deb]

Thanks for the answer. But I am not sure if using arrayfun is going to speed up the calculations. Because I want to use this for a matrix L with very large size along the 2nd dimension which is why I was thinking of avoiding to loop over it.