Clear Filters
Clear Filters

Calculate mean value from different matrices

13 views (last 30 days)
Hi there,
I have 5 different matrices (10x10000), where I want to calculate the mean value of the individual rows column by column, so that the end result is a single 10x10000 matrix. How can I do this most elegantly with the "mean" command?
  1 Comment
Image Analyst
Image Analyst on 30 Dec 2020
How do you end up with the same size result? If you have a 10x10000 matrix, then taking the mean of rows, going across columns, is like this
rowMeans1 = mean(M1, 2); % Result is a 10 x 1 vector.
and the result is a 10 x 1 vector. So if you had 5 or those, even if you stitched them together you'd have a 10x5 matrix.
overallResults = [rowMeans1, rowMeans2, rowMeans3, rowMeans4, rowMeans5]; % 10x5
Or if you went in the other direction
columnMeans1 = mean(M1, 1); % Result is a 1 x 10000 vector.
Stitching together 5 of them would give a 5x10000 matrix.
overallResults = [rowMeans1; rowMeans2; rowMeans3; rowMeans4; rowMeans5]; % 5x10000

Sign in to comment.

Accepted Answer

Ameer Hamza
Ameer Hamza on 30 Dec 2020
How are 5 matrices available? Is it a cell array? Try something like this
C = {M1, M2, M3, M4, M5};
M = cat(3, C{:});
M_mean = mean(M, 3)
  4 Comments
Image Analyst
Image Analyst on 30 Dec 2020
Edited: Image Analyst on 30 Dec 2020
Of course creating a cell array is not needed. You could just omit that and do
M = cat(3, M1, M2, M3, M4, M5);
M_mean = mean(M, 3);
Just to clarify Mepe, you're not using "column" in the usual sense of a matrix, as in rows and columns, like everyone else does. This is taking the average of all matching coordinates across 5 different matrices, not column-by-column, hence my question above asking for clarification. Evidently Ameer also has a beta copy of the Mind Reading Toolbox.
Ameer Hamza
Ameer Hamza on 30 Dec 2020
I wish I could subscribe to that toolbox :D . I just guessed based on confusing wording in the question and the intended size of the output matrix.

Sign in to comment.

More Answers (0)

Community Treasure Hunt

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

Start Hunting!