How to create a matrix from multiples of submatrice?

I would like to find a smarter way of building a larger matrix from vectors. The following code works fine:

rv = [1 2 3 4];
alpha =[0.0140;
    0.0100;
    0.0100;
    0.0100];
c = [];
for i = 1:4
    c = [c; diag(alpha).*rv(i)];
end

>> c =
    0.0140         0         0         0
         0    0.0100         0         0
         0         0    0.0100         0
         0         0         0    0.0100
    0.0280         0         0         0
         0    0.0200         0         0
         0         0    0.0200         0
         0         0         0    0.0200
    0.0420         0         0         0
         0    0.0300         0         0
         0         0    0.0300         0
         0         0         0    0.0300
    0.0560         0         0         0
         0    0.0400         0         0
         0         0    0.0400         0
         0         0         0    0.0400

Essentially I build a submatrix from a vector. Then I multiple the diagonal submatrix with each element of a vector. Then I concatenate these vertically.

Is there a computationally more efficient way of doing this?

Thanks!

Answer 1

[madhan ravi]

Z = reshape(rv,1,1,[]) .* diag(Alpha);
Wanted = reshape(Z,numel(rv),[]).'

[Michael Soskind]

Even more confusing but with no tansposes and in one line:

repmat(diag(alpha),4,1).*reshape(repmat(rv, 4,4),[],4)

[madhan ravi]

" and in one line"

Wanted = reshape(reshape(rv,1,1,[]) .* ...
    diag(Alpha),numel(rv),[]).'

"Even more confusing"

Why?

N = 1e6;
tic
for k = 1:N
    Z = reshape(rv,1,1,[]) .* diag(Alpha);
    Wanted = reshape(Z,numel(rv),[]).';
end
% Elapsed time is 4.847994 seconds.
toc
tic
for k = 1:N
    repmat(diag(alpha),4,1) .* reshape(repmat(rv, 4,4),[],4);
end
toc
% Elapsed time is 5.865444 seconds.