By element average of multiple matrices

Hi,

I have 40 matrices of 781x981 in dimension. I need to calculate the by element mean of all these matrices in a new, 41st matrix.

I.e., the mean of element 1,1 of all 40 matrices is returned as element 1,1 in the new, 41st matrix. The same is then done for all elements.

In additionto the mean, I need to do the same for the standard deviation also.

Is there a function in Matlab that can do this? Or do I have to write some kind of a loop??

Thanks, Bryan

Answer 1

[Wayne King]

Yes, concatenate the matrices into a 3-D array and use mean(x,3)

   x = randn(4,4,10);
   mean(x,3)

and

     std(x,[],3)

To concatenate, you can use cat()

    x = randn(4,4);
    y = randn(4,4);
    z = cat(3,x,y);
    mean(z,3)
    std(z,[],3)

[Jan]

... and std(x, 0, 3);

Answer 2

[Sean de Wolski]

Stack your 40 matrices into one three-dimensional matrix, then take the mean along the 3rd dimension

meanMatrix = mean(stackedMatrix,3); %doc mean for more info.

To stack your matrices look at these two FAQs:

• A1,A2...An
• Process Sequence of Files

[Jan]

It would be even better to avoid something like A1, A2, ... but store all matrices in a 3D array directly, or at least in a cell. Example:

C = cell(1, 10);

for i=1:10; C{i} = rand(4, 4); end

D = cat(3, C{:});

[Bryan]

thank you all for your answers, i did not think of this approach

[Belinda Finlay]

I am very new to MATLAB and am attempting a do a similar calacuation; however, I am having trouble stacking my matrices (as suggested by Sean), is someone able to guide me.

Thanks

Belinda

[Yash Ajay Garje]

Hi Belinda,

I do it like this:

A(:,:,1) = ['you could enter your 1st matrix here']

A(:,:,2) = ['you could enter your 2nd matrix here']

A(:,:,3) = ['you could enter your 3rd matrix here']

....

A(:,:,n) = ['you could enter your nth matrix here']

As a result, you would have created a matrix 'A' with the third dimension as 'n'. You could also loop this if required. Hope this helps.

Best,

Yash