Storing maximum values from each iteration

hi guys. below is my code. i'm trying to save the maximum values at each iteration of the for loop h1 = 0.02:0.001:0.03 (that is [maxUd1,indexUd1] at h1 = 0.02, [maxUd1,indexUd1] at h1 = 0.021,...., [maxUd1,indexUd1] at h1 = 0.03). how do i go about this please?

for h1 = 0.02:0.001:0.03;      
c1 = 0.01;                      
X1 = 0.3;
Ud1 = zeros(11,11,11,11);
for a1 = 0:0.1:1;
 for d1 = 0:0.1:1  
     Ud1(pA1,pd11,pd21,pd31) = h1*(d1/(a1+d1)) * (1-X1) - c1*d1 
     [maxUd1,indexUd1] = max(Ud1(:));
   end
 end
end

Answer 1

[Walter Roberson]

h1_values = 0.02:0.001:0.03;
a1_values = 0:0.1:1;
d1_values = 0:0.1:1;
num_h1 = length(h1_values);
num_a1 = length(a1_values);
num_d1 = length(d1_values);

%for efficiency the first index should be the inner loop index  
maxUd1 = zeros(num_d1, num_a1, num_h1);
indexUd1 = maxUd1;

for h1_idx = 1 : num_h1
  h1 = h1_values(h1_idx);
  c1 = 0.01;                      
  X1 = 0.3;
  Ud1 = zeros(11,11,11,11);
  for a1_idx = 1 : num_a1
    a1 = a1_values(a1_idx);
    for d1_idx = 1 : num_d1
      d1 = d1_values(d1_idx);
      %why are you storing what appears to be a scalar into an array at constant indices??
      Ud1(pA1,pd11,pd21,pd31) = h1*(d1/(a1+d1)) * (1-X1) - c1*d1;
      [maxUd1(d1_idx, a1_idx, h1_idx), indexUd1(d1_idx, a1_idx, h1_idx)] = max(Ud1(:));
    end
  end
end

%now that we completed the arrays, flip them around so the first index is the
%h1 index and the last one is the d1 index
maxUd1 = permute(maxUd1, [3 2 1]);
indexUd1 = permute(indexUd1, [3 2 1]);

[John BG]

To Chiamaka Agwuegbo

this is John BG ( <mailto:[e-mail] [e-mail]> )

would you please define range and start values for the following indices

pA1
pd11
pd21
pd31

thanks in advance

[Walter Roberson]

The ranges of those is really only relevant if they are not positive integers, or if X is not a scalar.

I think that part of the code has simply been miswritten, or copied from somewhere else without understanding.

[Walter Roberson]

By the way: unless X is a vector or array, I would write all of this a quite different way -- probably by using ndgrid(), doing all of the arithmetic computations simultaneously, and then using max() on the appropriate dimension of the result.

[Chiamaka Agwuegbo]

Thank you very much. It worked. I apologize for the late reply. My system crashed.