How to avoid double for-loop

I have the following code

    for n = 1:N
        for l = 1:N
            phase = exp(1i * (rand*0.5*pi-0.25*pi));
            Efield_x(n,l) = Efield_x(n,l) .* phase;
        end
    end

with N=2^10 . Each matrix element should be multiplied with a different random number from -0.25*pi to 0.25*pi . Although the computer is running ("busy") I am waiting for 15 minutes now. How can we avoid this slow double for-loop to fasten to process significantly? By the way, I have 8 GB Ram.

Answer 1

[Star Strider]

This seems to work:

N = 2^10;
Efield_x = ones(N);                         % Create Data
phase = exp(1i * (rand(N)-0.5)*0.25*pi);
Efield_x = Efield_x .* phase;

[Lucius]

you mean

    ...(rand(N)-0.5)*0.5*pi);

This works, thanks.

[Star Strider]

My pleasure.

You’re correct on ‘phase’, with 0.5 rather than 0.25 as the multiplier to produce ±0.25*pi.