How to add specific rows to a matrix?

I have a matrix A with 3 columns and multiple rows,

A =

a_1 b_1 2

a_1 b_1 3

a_1 b_1 2

a_2 b_2 6

a_2 b_2 8

a_3 b_3 3

a_3 b_3 9

a_3 b_3 5

a_3 b_3 4

... ... ...

For each couple of parameters (a_i, b_i) in A that appears less four times, I woud like to create a matrix B that contains additional rows "a_i b_i 0" for this couple of parameters, such that

B =

a_1 b_1 2

a_1 b_1 3

a_1 b_1 2

a_1 b_1 0 %added rows

a_2 b_2 6

a_2 b_2 8

a_2 b_2 0 %added rows

a_2 b_2 0 %added rows

a_3 b_3 3

a_3 b_3 9

a_3 b_3 5

a_3 b_3 4

... ... ...

Note that the value a_i and b_i ae not unique, but each couple (a_i, b_i) would be different from (a_j, b_j).

Thanks in advance!

[Image Analyst]

Can you give an actual matrix with values?

TUTORIAL: How to ask a question (on Answers) and get a fast answer

I'm thinking that using findgroups() to find groups in the left column would be a part of the solution. If the length of any group is less than 4 you'd have to insert row(s).

[Rodrigues Bitha]

Here is an example of the matrix A

A =

3.8 6.4 2.0

3.8 6.4 5.0

3.8 6.4 9.0

3.8 6.5 4.0

3.8 6.5 3.0

3.9 6.4 1.0

3.9 6.4 8.0

3.9 6.4 7.0

3.9 6.4 4.0

And I would like to have this matrix B as a final result

A =

3.8 6.4 2.0

3.8 6.4 5.0

3.8 6.4 9.0

3.8 6.4 0.0 % added rows

3.8 6.5 4.0

3.8 6.5 3.0

3.8 6.5 0.0 % added rows

3.8 6.5 0.0 % added rows

3.9 6.4 1.0

3.9 6.4 8.0

3.9 6.4 7.0

3.9 6.4 4.0

Answer 1

[Voss]

A =[
    3.8    6.4    2.0
    3.8    6.4    5.0
    3.8    6.4    9.0
    3.8    6.5    4.0
    3.8    6.5    3.0
    3.9    6.4    1.0
    3.9    6.4    8.0
    3.9    6.4    7.0
    3.9    6.4    4.0
    ];
needed_rows = 4;

[uA,~,jj] = unique(A(:,[1 2]),'rows','stable');
n_uA = size(uA,1);

B = [repelem(uA,needed_rows,1) zeros(needed_rows*n_uA,1)];
for ii = 1:n_uA
    idx = jj == ii;
    B(needed_rows*(ii-1)+(1:nnz(idx)),3) = A(idx,3);
end

disp(B);

    3.8000    6.4000    2.0000
    3.8000    6.4000    5.0000
    3.8000    6.4000    9.0000
    3.8000    6.4000         0
    3.8000    6.5000    4.0000
    3.8000    6.5000    3.0000
    3.8000    6.5000         0
    3.8000    6.5000         0
    3.9000    6.4000    1.0000
    3.9000    6.4000    8.0000
    3.9000    6.4000    7.0000
    3.9000    6.4000    4.0000

[Rodrigues Bitha]

Thanks a lot!

Answer 2

[Image Analyst]

Not sure why you didn't try findgroups() like I suggested. Or maybe you have by now. You might have gotten something like this:

A =[
    3.8    6.4    2.0
    3.8    6.4    5.0
    3.8    6.4    9.0
    3.8    6.5    4.0
    3.8    6.5    3.0
    3.9    6.4    1.0
    3.9    6.4    8.0
    3.9    6.4    7.0
    3.9    6.4    4.0
    ];
g = findgroups(A(:, 1))

g = 9×1

1 1 1 1 1 2 2 2 2

neededRows = 8;
for k = 1 : max(g)
    theseRows = g == k;
    numInThisGroup = sum(theseRows);
    if numInThisGroup < neededRows
        lastRow = find(theseRows, 1, 'last');
        rowsToAdd = neededRows - numInThisGroup;
        % Insert rows.
        A = [A(1:lastRow,:); repmat(A(lastRow, :), [rowsToAdd, 1]); A(lastRow+1:end, :)];
        % Need to update g now.
        g = findgroups(A(:, 1));
    end
end
A

A = 16×3

3.8000 6.4000 2.0000 3.8000 6.4000 5.0000 3.8000 6.4000 9.0000 3.8000 6.5000 4.0000 3.8000 6.5000 3.0000 3.8000 6.5000 3.0000 3.8000 6.5000 3.0000 3.8000 6.5000 3.0000 3.9000 6.4000 1.0000 3.9000 6.4000 8.0000

[Rodrigues Bitha]

Thanks a lot!