how to take average of each column of matrices

28 views (last 30 days)
chan
chan on 30 Nov 2021
Commented: Stephen23 on 3 Dec 2021
A = randi(9,[6 3 6]);
U = 3:5;
% do it with a loop
M = zeros(size(A,1),size(A,2),numel(U));
for k = 1:length(U)
M(:,:,k) = A(U(k),1,2) * A(:,:,U(k));
end
%Here i have 3 matrices i.e. M(:,:,1),M(:,:.2),M(:,:,3).
Could someone suggest me a hint on how to take the average of each column of the three matrices. result will be a 6*3 matrix
"sendcnt" can be 3/4/5
when i do M/sendcnt then each cell of the matrix is divided by sendcnt which is not right.
what i want is (1st column of 1st matrix+1st column of 2nd matrix+1st column of 3rd matrix)/sendcnt
similarly for the other 2nd and 3rd column.
  1 Comment
Stephen23
Stephen23 on 3 Dec 2021
"i want mean for column of the 3D matrix. it will be 6*3 matrix"
The mean along column of a 6x3x6 array will return a 6x1x6 array.
I suspect that you might want the mean along the third dimension (aka "page"):
The mean along each "page" will return a 6x3(x1) array.

Sign in to comment.

Answers (2)

Rik
Rik on 30 Nov 2021
It sounds like you want something like this:
A = randi(9,[6 3 6]);
U = 3:5;
% do it with a loop
M = zeros(size(A,1),size(A,2),numel(U));
for k = 1:length(U)
M(:,:,k) = A(U(k),1,2) * A(:,:,U(k));
end
data=mean(M,2); %apply mean over the second dimension (columns)
squeeze(data) % reshape the nx1xm to nxm
ans = 6×3
45.3333 25.0000 4.3333 18.6667 13.3333 4.3333 21.3333 16.6667 4.0000 58.6667 26.6667 4.6667 40.0000 26.6667 6.0000 18.6667 18.3333 5.0000
  3 Comments
Rik
Rik on 3 Dec 2021
That is what mean(M,2) does, however, the result is not 6*3 but 6*1*3. The reason is that you only have 1 column left if you calculate the average of the columns.
Can you give an example of input and expected output?

Sign in to comment.


KSSV
KSSV on 30 Nov 2021
A = randi(9,[6 3 6]);
iwant = mean(A,1)

Tags

Products


Release

R2016a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by