Count number of column pairs occurrence in 3d array

I have a 3d array like the example below:

A(:,:,1) = [5.5 2.2; 5.3 2.5; 5.3 2.1; 5.4 2.6]

A(:,:,2) = [5.2 2.9; 5.4 2.6; 5.4 2.6; 5.5 2.2]

I want to calculate the number of occurrences of all column pairs. So for example [5.4 2.6] should equal 3 and [5.5 2.2] should equal 2 while the rest of all pairs will equal 1.

I want the numbers to be stored in a new array at the same location as its representative pair, like this:

B(:,:,1) = [2 1 1 3]

B(:,:,2) = [1 3 3 2]

How can I do this as efficiently as possible? I should add that the real array is much larger so I really need an automated solution.

[Jan]

Do you mean a 2D matrix as output:

B = [2 1 1 3; ...
     1 3 3 2]

? Or maybe its transposed?

[Hampus Alfredsson]

No, A has "n" different sheets of pairs. I want B to have the same amount of sheets since it sounds simpler to relate between pair and number of occurrences that way. I solved the problem using several nested for-loops, but I feel that there is an easier way to do it.

Answer 1

[Jan]

A(:,:,1) = [5.5 2.2; 5.3 2.5; 5.3 2.1; 5.4 2.6];
A(:,:,2) = [5.2 2.9; 5.4 2.6; 5.4 2.6; 5.5 2.2];

[s1, s2, s3] = size(A);
B = reshape(permute(A, [1,3,2]), s1*s3, []);
[C, ~, iB]  = unique(B, 'rows', 'stable');            % Find unique pairs
[N, ~, bin] = histcounts(iB,'BinMethod','integers');  % Count the pairs
R = reshape(N(bin), s1, [])                            % Create output matrix

[Hampus Alfredsson]

Good solution @Jan Simon. I might use this depending on how my continued coding turns out. Thank you!