Why Matlab doesn't export the result?

I import an excel file into Matlab. Then I expect with the code I write in Matlab, it will export the result in column 3 in the original excel. But nothing appears in column 3 of the file sample2.xlsx. Could you help me?

clear all
clc
A=xlsread('Criteria.xlsx');
B=xlsread('sample2.xlsx');
n=0;
x=0;

for h=1:99
  if B(h,2)>50
    n=n+1;
    x= x+B(h,3);
  end
end
x=x/n;

for h=1:99
  if B(h,2)>50
    B(h,4)=x;
  end
end

xlswrite('sample2.xlsx',B)

Answer 1

[Image Analyst]

What is "A" for? You're not even using it.

And column 3 of B has NaNs in it, so x will be a NaN also, which means column 4 is all NaNs. It looks like x is supposed to be the mean of Col 3 of B for rows where B > 50. Is that true? If so, those two for loops can be vectorized:

% Don't read in A - it's not used.
% A=xlsread('Criteria.xlsx');
B=xlsread('sample2.xlsx');  
% Determine which rows to average.
rowsToAverage = B(:,2)>50 & ~isnan(B(:,3));
% Get mean in Column 3
meanOfBigNumbers= mean(B(rowsToAverage, 3));
% Set col 4 = means
B(rowsToAverage, 4) = meanOfBigNumbers;
% Overwrite workbook with new B value.
xlswrite('sample2.xlsx',B)

The xlswrite should work.

[Dung Le]

I have applied your code with adjustment in order to be appropriate with more conditions (instead of only > 50 as mentioned above), here it is:

clear all
clc
A=xlsread('criteria.xlsx');
B=xlsread('panel.xlsx');

for i=1:89
    for j=1:64
        for k=1:3
            for h=1:2

                  rowsToAverage = (B(:,5) == A(i,1)) && (B(:,4) == A(j,2)) && (A(k,4) > B(:,2) > A(k,3)) && (A(h,5) < B(:,1) < A(h,6)) && ~isnan(B(:,3));

                  meanOfBigNumbers= mean(B(rowsToAverage, 3));

                  B(rowsToAverage, 6) = meanOfBigNumbers;
              end
          end
      end
  end

xlswrite('panel.xlsx',B)

Bit it does not work, could you please help me!

[Image Analyst]

You don't need all those loops. That's why I vectorized it. Can you tell me in words what you want to do? Like inspect A and find out if some column in B is more than something in A, and if it is, assign something to something?

[Dung Le]

I have tried your code and find something weird, I asked this in link:

http://www.mathworks.com/matlabcentral/answers/262744-why-two-different-ways-of-expressing-the-same-logical-condition-lead-to-two-different-results

Please help me!

[Image Analyst]

Looks like you've found a solution in your parallel thread because you've accepted an answer there. I still claim you don't need all those for loops. But anyway, at least it's working.

[Walter Roberson]

Dung Le remarks "Thank you your enthusiastic support :)"

Answer 2

[Walter Roberson]

Could you confirm that you want to take the mean of column 3 over the places that column 2 is > 50, and that you want to write copies of that mean to column 4 for those same rows? Like

mask = B(1:99,2) > 50;
x = mean(B(mask,3));
h(mask,4) = x;

[Dung Le]

Yes, but could you explain why my code doesn't work ?

[Image Analyst]

The code in my answer (different than Walter's code in this Answer) does work - I actually tried it. There is stuff in all 4 columns. Where you originally had NaN's there will be a blank cell.

[Dung Le]

Thanks :)