logical operations on particular matrix elements

I have the following code:

 numels = 7;
 numpts = 3;

 C = sparse(numpts,numels)

 A = randi(numels,[numpts,1]);
 B = rand(numpts,1);

I want to treat the matrix A as an index for the columns of matrix C and move the values of B to their respective columns.

So if we have:

 A = [3;3;6];
 B = [0.383;0.892;0.192];

Then we should be able to get:

 full(C) =
 0  0  0.383  0  0  0  0
 0  0  0.892  0  0  0  0
 0  0  0  0  0  0.192  0

I thought that

 C(:,A)=B;

might work, but C(:,A) attempts refers to a matrix and not a set of values.

BTW, I want a logical operation, I don't want to use accumarray or something. It is important that it is fast.

Any help is appreciated.

[Christopher]

I've found that I can just find the indices in the sparse matrix and thus use the following:

 newA = (A-1).*numpts+(1:numpts)';
 C(newA)=B;

However, I've found that using larger matrices of C (required for my implementations), execution of C(newA)=B; is EXTREMELY SLOW.

How can I make this operation faster?

Answer 1

[Azzi Abdelmalek]

numels = 7;
numpts = 3;
C = sparse(numpts,numels)
A = [3;3;6];
B = [0.383;0.892;0.192]
idx=sub2ind(size(C),1:numel(A),A')
C(idx)=B
full(C)