Vectorization of nested for loops

hi, This is Pinaki. I am having a problem regarding the code execution time. The following code is taking almost 54 % of total run time to execute. Pls help me to victorize the following codes.

      for m = 1:nbus
    for n = 1:nbus
        if m ~= n
            Sij(m,n) = Vm(m)*conj(Iij(m,n))*BMVa;
        end
    end
end

Thanks, Pinaki

Answer 1

[Guillaume]

Assuming Vm is a column vector (if not, change it so it is) and BmVa is a scalar:

assert(iscolumn(Vm), 'Vm must be a column vector');
assert(isscalar(BMVa), 'BMVa must be scalar');
Sijall = bsxfun(@times, Vm, conj(Iij) * BMVa);

This will calculate values for Sij even on the diagonal. If Sij is just zeros and you want that diagonal to stay 0:

Sij = Sijall .* ~eye(size(Sijall))

If it has values on the diagonal that you want to preserve:

Sij = Sij .* eye(size(Sij)) + Sijall .* ~eye(size(Sij))

[PINAKI DHAR]

Thank you very much !!!!!!!!!!