HOW TO TAKE AVERAGE OF ROW AND COLUMN AT SOME POINT

I HAVE A MATRIX

   2 5 3 500 4 5,
   3 4 2 2  3  5,
   4 5 5 1  2   2

code must scan first where 500 exists and take average of that row and column wher 500 exists excluding that 500,

here ans should be like avg row at first 2+5+3+4+5/5 and avg col of that point 2+1/2,

[per isakson]

I assume you mean

M = [ 2 5 3 500 4 5; 3 4 2 2 3 5; 4 5 5 1 2 2 ];

>> whos M
  Name      Size            Bytes  Class     Attributes
  M         3x6               144  double

[abdul wahab aziz]

i will search for 500 first then will replace that 500 with that row and column avg....here i need row avg where 500 exists and its column average but 500 must be excluded while calculating averages or mean

[abdul wahab aziz]

my finals ans would i will replace row_Avg+col_avg/2 with 500,but before that i have to calculat row_Avg and col_avg seperately of that point where 500 exists

Answer 1

[Star Strider]

This works:

M = [2 5 3 500  4  5
     3 4 2   2  3  5
     4 5 5   1  2  2];

[r,c] = find(M == 500);

RowAvg = mean(M(r,M(r,:)~=500));
ColAvg = mean(M(M(:,c)~=500,c));

[abdul wahab aziz]

what does r and c reprsent are these iterations?

[abdul wahab aziz]

what if 500 exists multiple time?

[Star Strider]

The ‘r’ and ‘c’ are the row and column indices of ‘500’ respectively.

In the event of more than one ‘500’, the easiest way to modify my code would be to add a loop:

M = [2 5 3 500  4  5
     3 4 2   2  3  5
     4 5 500 1  2  2];

[r,c] = find(M == 500);

for k1 = 1:size(r,1)
    RowAvg(k1) = mean(M(r(k1),M(r(k1),:)~=500));
    ColAvg(k1) = mean(M(M(:,c(k1))~=500,c(k1)));
end

Here, every ‘[r,c]’ pair defines a specific value for ‘RowAvg’ and ‘ColAvg’.

This works if there is only one ‘500’ in any specific row or column. If there are more than one ‘500’ in any row or column, you have to decide how you want to deal with the second ‘500’.

Answer 2

[Thorsten]

A = [2 5 3 500 4 5; 3 4 2 2 3 5; 4 5 5 1 2 2];
[i, j] = ind2sub(size(A), find(A==500));
m1 = mean(A(i, ~ismember(1:size(A,2), j)));
m2 = mean(A(~ismember(1:size(A,1), i), j));

A(i,j) = (m1 + m2)/2;

or

idx = A == 500
A(idx) = nan;
A(idx) = mean([nanmean(A(:, any(idx, 1))) nanmean(A(any(idx, 2), :), 2)])

[abdul wahab aziz]

THANK YOU THORSTEN YOUR ANSWER SOLVED MY PROBLEM

[abdul wahab aziz]

for i = 1:size(G2,1)

    for j = 1:size(G2,2)

        hidden=G2(i,j);
   if(hidden==100)
K=K+1;
 RowAvg=mean(G1(i, ~ismember(1:size(G1,2), j)));
 ColAvg=mean(G1(~ismember(1:size(G1,1), i), j));

 this is my code i also want to ignore zeroes, while calculating mean? can u add that thing in to it