How do I average rows which contain the same first to elements?

I'm stuck on this problem for a while. I want to obtain a matrix which is averaged for the rows which contain the same first two elements. I don't want to clear the rows which are not equal for the first two elements. When I have a matrix A:

A=[1     1     1     1     1
   1     2     3     4     5
   1     2     4     4     4
   1     2     3     3     3
   2     2     2     2     2
   2     2     3     3     3]

I want the outcome to be:

 A=[1    1    1     1     1
    1    2   3.33  3.67   4
    2    2   2.5   2.5   2.5]

I hope someone could help me on this.

Answer 1

[Ahmed]

This solution is not very elegant, but should work. There is room for improvement, though.

    uA = unique(A(:,1:2),'rows');

    A = [uA cell2mat(cellfun(@(x)...
         mean(A(all(bsxfun(@eq,A(:,1:2),x),2),3:end),1), ...
         num2cell(uA,2),...
         'UniformOutput',false)) ]

[Glenn Roumen]

Thank you very much!

Answer 2

[Andrei Bobrov]

[~,~,c] = unique(A(:,1:2),'rows');
out = accumarray([repmat(c,size(A,2),1) kron((1:size(A,2))',...
                                  ones(size(A,1),1))],A(:),[],@mean);

or

[c c c] = unique(A(:,1:2),'rows');
[ii,jj] = ndgrid(c,1:size(A,2));
out = accumarray([ii(:) jj(:)],A(:),[],@mean);