unique pairs - frequency appearance

Hello, how can I make a table of unique pairs frequency appearance according to a specific column (In the example - Column A)? example in the image:

10 & 10 - mark +1 in the correct cell in the table.
10& 30 - mark +1 in the correct cell in the table

later in "2": 
10&20 - mark + in the correct cell in the table.
et cetera

Can you help me bulid the algoritem for a table like this? Thank you!

Answer 1

[Guillaume]

There are two parts to your question. First, the hard bit is going from your table to a list of pairs. The second bit, much easier, is building the histograms of the pair.

One possible way to build the list of pairs:

data = [1 10;1 10;1 30;2 10;2 20; 3 30;3 20;3 10;4 10;4 20;4 30]

%the accumarray below assumes that column 1 is always integer positive from 1. 
%column 1 does not need to be ordered:
groupeddata = accumarray(data(:, 1), data(:, 2), [], @(v) {v}); 
%nchoosek to generate pairs, sort(..., 2) to remove the ordering, unique(...) to remove duplicates:
pairpergroup = cellfun(@(v) unique(sort(nchoosek(v, 2), 2), 'rows'), groupeddata, 'UniformOutput', false);
allpairs = vertcat(pairpergroup{:})

One possible way to build the histogram:

[pair, ~, row] = unique(allpairs, 'rows');
pairdist = [pair, histcounts(row, [unique(row); Inf])']