Prev: Counting the frequency of data in an array
Next: fminunc, fminsearch
From: Gabriel on 24 Jul 2010 15:39
Hi,

I am interested in learning how to reduce the simulation time for a nested for loop so that I can reduce some computational time. The purpose of my code is to create a 2D staggered grid (j,k) where U is defined at (j+1/2,k), V is defined at (j,k+1/2), and h is defined at (j,k). Most of the computational time of the code is taken doing nested for loops like this one:

U=zeros(n,n,3);V=zeros(n,n,3);Z=zeros(n,n,3);PHI=zeros(n,n,3);
for L=1:3
for j=1:n-1
for k=1:n-1
U(j+1,k,L)=0.5*(h(j+1,k,L)+h(j,k,L))*u(j+1,k,L);
V(j,k+1,L)=0.5*(h(j,k+1,L)+h(j,k,L))*v(j,k+1,L);
end
end
end

I wanted to know if there is a way to vectorize this loop. Thanks.
From: Matt J on 24 Jul 2010 16:02


U=convn(h,[0.5;.5],'valid').*u(2:end,:,:);
V=convn(h,[0.5,.5],'valid').*v(:,2:end,:);
 | 
Pages: 1
Prev: Counting the frequency of data in an array
Next: fminunc, fminsearch