Data matrix manipulation in Matlab

Hi, I have a data matrix, and each row is (t, i, row_vector), i.e., the first two elements represent time t and individual i, and the rest is a row vector of information on (t,i). For simplicity, assume row_vector is a real number. I want to create a matrix M s.t. M(t,i) = row_vector (i.e., the real number data on (t,i)) and M(t,i) missing if (t,i) does not turn up in the original data matrix. How do I do this? I'm lost how to do this.. I'd really appreciate any and all help. Thank you very much in advance!

Best, John

[the cyclist]

Suppose your input data matrix is

data = [1 3 4 5 6;
        4 2 7 8 9];

What do you want your output to be?

Do you want it to be a cell array where

M{1,3} = [4 5 6]

and

M{4,2} = [7 8 9]

?

[Jeong Ho]

Dear the cyclist, Yes, that's exactly it! Is there a way of doing it?

Answer 1

[Image Analyst]

Try this:

data = randi(50, 10, 7)
[rows, column] = size(data);
for row = 1 : rows
  t = data(row, 1);
  i = data(row, 3);
  M{t, i} = data(row, 3:end);
end
% Print to command window:
celldisp(M);

Answer 2

[Stephen23]

This can also be solved very neatly using accumarray:

>> data = [1 3 4 5 6; 4 2 7 8 9];
>> A = repmat(data(:,1:2),size(data,2)-2,[]);
>> B = reshape(data(:,3:end),[],1);
>> C = accumarray(A,B,[],@(v){v.'})
C = 
    []              []    [1x3 double]
    []              []              []
    []              []              []
    []    [1x3 double]              []
>> C{1,3}
ans =
     6     5     4
>> C{4,2}
ans =
     7     8     9

This is likely to scale better to larger matrices than using a loop.