Code doesn't terminate in the way that I want

I have a loop which changes the value of a vector at each iteration of the loop. I don't want the values in the loop to get greater or smaller than predefined values. The code I'm using for this is:

    if (min(u_new)<0.02 || max(u_new)>0.98)
        return;
    end

This doesn't seem to work. I thought that this would be a relatively simple thing to prgram but it's being stubburn.

Answer 1

[Max Alger-Meyer]

So I'm not totally sure what you're asking the code to do, but using 'return' is going to end your script or function. If you just want to exit the loop, use 'break' instead. Alternatively, if you just want the loop to skip over the values that would fall outside of your desired range and keep iterating, you'd want to flip the signs in the 'if' condition and change the values in the body of the if statement, once the criteria are met. If you post additional code so I can see what you are trying to do I can help more.

[Matthew Hunt]

So that didn't work either. The loop in question is:

for i=1:m
    gamma=-2*(dr/D)*(0.5*alpha+beta/R_ps(n)); %For computational efficency.
    c=gamma*w.*q(i)';
    norm(c)
    u_new=B_1*u_old(:,i)+c;
    u(:,i)=u_new;
    if (min(u_new)<0.02 || max(u_new)>0.98)
        break;
    end
end

[Image Analyst]

Looks like it should work. At the end of the loop, just before the if statement, put this:

fprintf('u_new = %f.\n', u_new);

Do you see any values outside the range? If you still have trouble, attach a mat file with m, alpha, beta, dr, D, w, r_ps, etc. in it so we can run your code.

[Jan]

@Matthew Hunt: Just a hint: There is a comment "For computational efficency". It is inefficient to calculate a constant repeatedly inside a loop. Move the line:

gamma = -2 * (dr / D) * (0.5 * alpha + beta / R_ps(n));

in front of the loop.

Is q a vector? Then q(i) is a scalar and there is no need to transpose it. But if q contains imaginary values, it matters that the ' operator is the conjugate transposition.

[Matthew Hunt]

I want the code to stop at that point.

[Max Alger-Meyer]

@Matthew Hunt Another thing that I notice is that you are calling 'norm(c)', but you're not setting that value to a variable, so that line isn't doing anything. Do you perhaps mean to do something like this? If this isn't the issue, I think we'd need a .mat file to test the code, or at least a more detailed description of the error that you're getting.

c=gamma*w.*q(i)';
norm_c = norm(c);
u_new=B_1*u_old(:,i)+norm_c;

[Matthew Hunt]

The norm(c) was just to output the value of norm of c so I could see the change in the value of c as time progressed.

[Jan]

@Matthew Hunt: And it does break the loop, if the condition is met:

min(u_new)<0.02 || max(u_new)>0.98

If the loop is not left, the condition is not true. You wrote "This doesn't seem to work", but of course it works exactly as expected. The problem is, that your u_new has other values than you expect. You can check this by your own or provide the input values.