Vectorizing a loop for cumulative sum with reset

The lines below work fine to compute cumulative sum on a 1D logical array ('s') such that the summation resets to zero whereever the element in 's' is zero. Question is, how can this be vectorised to avoid loop.

r(1) = s(1); %s is the source vector, r is the result
for i = 2:length(s)
    r(i) = (r(i-1) + s(i))*s(i);
end

Thanks.

Answer 1

[Lars Nelson]

Unfortunately, because each value in vector r depends on the previous value in r, I do not think that there is a way to vectorize this code and you will have to use a loop.