Splitting arrays using loops

would there be a way to create a loop for splitting an 8x8 array into seperate 4x4 arrays and then find the mean of each of the 4x4 arrays and place each of the mean values into a new array ?

[Dyuman Joshi]

Do you want (fixed) 4 4x4 array, like top left, top right, bottom left and bottom right?

or do you want (sliding) 4x4 arrays like (row 1-4, col 1-4); (row 1-4, col 2-5) and so on?

[Stephen23]

A = reshape(1:64,8,8);
M = mean(permute(reshape(A,4,2,4,2),[2,4,1,3]),3:4)

M = 2×2

14.5000 46.5000 18.5000 50.5000

F = @(s) mean(s.data(:));
M = blockproc(A,[4,4],F)

M = 2×2

14.5000 46.5000 18.5000 50.5000

B = conv2(A,ones(4,4));
M = B(4:4:end,4:4:end)/16

M = 2×2

14.5000 46.5000 18.5000 50.5000

[Bruno Luong]

No need for permute, squeeze does the job just fine and perhaps faster (no data moving around)

A = reshape(1:64,8,8);

M = squeeze(mean(reshape(A,[4 2 4 2]),[1 3]))

M = 2×2

14.5000 46.5000 18.5000 50.5000

[Stephen23]

"No need for permute, squeeze does the job just fine"

Not really, because SQUEEZE is fragile, unlike PERMUTE. It all looks "fine" ... until one day the user has data which consists of one row of blocks and are then astonished when the output has the wrong orientation:

A = reshape(1:32,4,8)

A = 4×8

1 5 9 13 17 21 25 29 2 6 10 14 18 22 26 30 3 7 11 15 19 23 27 31 4 8 12 16 20 24 28 32

M = squeeze(mean(reshape(A,[4 1 4 2]),[1 3])) % oops, wrong output

M = 2×1

8.5000 24.5000

M = mean(permute(reshape(A,4,1,4,2),[2,4,1,3]),3:4) % aaah, much better

M = 1×2

8.5000 24.5000

M = permute(mean(reshape(A,4,1,4,2),[1,3]),[2,4,1,3]) % also this

M = 1×2

8.5000 24.5000

SQUEEZE is just like LENGTH: used only by people who like hidden bugs in their code.

[Bruno Luong]

OP stated clearly he wants average on 8 x 8 matrix not 4 x 8.

But granted if one doesn't like squeeze, in this block average problem use reshape rather than permute for efficienty

A = reshape(1:32,4,8)

A = 4×8

1 5 9 13 17 21 25 29 2 6 10 14 18 22 26 30 3 7 11 15 19 23 27 31 4 8 12 16 20 24 28 32

M = reshape(mean(reshape(A,[4 1 4 2]),[1 3]), [1 2])

M = 1×2

8.5000 24.5000

Answer 1

[Chetan Bhavsar]

I have already answered you same apart from making mean value as new array. here is how you can do it.

A = reshape(1:64, [8,8]);
mean_values = zeros(2,2);

% Loop to extract 4x4 matrices and compute mean
for row = 1:2
    for col = 1:2
        r_idx = (row-1)*4 + 1 : row*4;
        c_idx = (col-1)*4 + 1 : col*4;
        
        submatrix = A(r_idx, c_idx);
        mean_values(row, col) = mean(submatrix(:));
    end
end


disp(mean_values);

   14.5000   46.5000
   18.5000   50.5000