I want to create an average 2d field every 10 timesteps within a 3d matrix (6000 x 100 x100) --> (600 x 100 x100)

I have for example a 3d matrix like (6000 x 100 x 100) with the first array being time, or the vector I want to average. How can I create an average every 10 days so that I am left with 600 x 100 x 100, so that 10 timesteps have been averaged to produce one 2d field matrix.

I have tried using accumarray but I can't seem to get it to work...and I have tried the following code (after permuting matrix):

for x = 1:100; x for y = 1:100;

    aa = squeeze(data(x,y,:));
    test(x,y) = nanmean(aa(1:10:end));

    end
end

but then I am not sure what to do next...

Can someone help me?

thanks, Michael

Answer 1

[Guillaume]

Reshape your matrix into a 3d array of size 10x600x100x100 and calculate the average of that along the first dimension:

tdata = reshape(data, 10, 600, 100, 100);
mdata = squeeze(mean(tdata));

[Michael]

great, works perfectly