what wrong with the fallowing codes?

i have a 20*120 matrix for each column in the matrix i need to find the maximum value between all the values , and then sum the remaining values ,then i need to divide the maximum value by the summation of the remaining values. i tried this codes but the results were not correct what is the problem?

s=1:z      %z=120
for i=1:x       %x=20
maximss=max(Pres_W);     %maximum value
InterFss=(sum(Pres_W))-maximss;     %remaining values
SIRk(:,s)=(maximss(:,s))./(InterFss(:,s));
  end
end

Answer 1

[John D'Errico]

Seems pretty straightforward, assuming I interpret your question properly.

% some completely arbitrary matrix
A = rand(20,120);

% maximum value for each column
mval = max(A,[],1);

% sum the remaining values, i.e., not including the maximum
% just compute the overall sum for each column, then subtract the max.
remsum = sum(A,1) - mval;

% divide the max by the remaining sum
mvalscaled = mval./remsum;

Answer 2

[Image Analyst]

Here's the first part:

rows = 4;
columns = 20;
Pres_W = rand(rows, columns) $ Sample data.
% Get max in each column
columnMaxima = max(Pres_W, [], 1)
% Make same size as original matrix
columnMaxima = repmat(columnMaxima, [rows, 1])
% "find the maximum value between all the values"
differences = Pres_W - columnMaxima
maxDifference = max(differences(:))

But then I came to the instruction "sum the remaining values" and I had no idea what the remaining values were, and I couldn't see how your line of code gave the remaining values. Remaining after what???