Avoid using for loops as it results in ram crash and slow down

My code is:

      for i = 1:n
          for ii = 1:i-1
              for iii = 1:ii-1
                  for iv = 1:iii-1
                    A = data(l([i ii iii iv 27]), 1:3);
      %             C(i,ii) = norm(A)*norm(inv(A));
      %             C(i,ii,iii) = norm(A)*norm(inv(A));
                    C(i,ii,iii, iv, 1) = cond(A);
  %                   tmp = svd(A, 0); 

%                   C(i,ii,iii,iv) = 0; 
    %               vif{i,ii,iii} = diag(inv(A'*A));
    %               vif1{i,ii,iii} = diag(inv(A*A'));
                end
            end
        end
        i
    end

The loop test for only 4 different wavelengths and here I am fixing the the 5th by using 27 in "A = data(l([i ii iii iv 27]), 1:3)", I am interested in adding the fifth as a loop or maybe 6th as well but the problem my ram is crashing and slow down the code a lot.

Any suggestion on how to solve this problem?

Regards,

[Kirby Fears]

Be sure to pre-allocate memory for the variables (like C) that you fill inside of your loop.

[Eng. Fredius Magige]

Hi for i=1:n, mean (i=n:-1:1, might the correct) has to start with 1 it imply in for ii, has output of 1-1=0

[Muhannad]

Thanks for the pre-allocate memory, I have already done that I just did not include it in the code. the problem is that I am creating a matrix of size 200*200*200*200*200 which is huge. Therefore I was wondering if there is a better way to deal with this situation.

Thanks for the replay,

[Walter Roberson]

Only allocate 200*200*200*200*1 as the maximum you use in the final dimension is 1.

[Muhannad]

Yes as I stated I am only using 4, but I want to use more for more wavelengths. Any suggestion on how can I do that without crashing the RAM?