Probably simple, but can't seem to make it work.

So, tcouple is an N x Rep matrix with numbers inbetween 0 and 50, where the columns represent scenarios. Now, for every scenario I need to get a vector from t=0:tmax. It's hard to explain. I will give an example. Take the first column of tcouple, then if t=0, I need to have the amount of elements of the first column of tcouple which are greater or equal to 0. if t=1, I need to have the amount of elements of the first column of tcouple which are greater or equal to 1. Etc. So every time from the same column.

I managed to create a m-file for that:

      for i=1:N
        for t=0:tmax
            if tcouple(i,1)>=t
            Mtmatrix(i,t+1)=1;
            else
            Mtmatrix(i,t+1)=0;
            end
        Mt=sum(Mtmatrix);            
        end
    end

Here Mt is the vector from t=0:tmax for scenario 1 (column 1 of tcouple). But now I need to do this for every scenario, so every column of tcouple, and then get a matrix of all the Mt's.

I've been trying to do this, but every time it goes wrong. It's probably easy, but I just don't see how to solve this.

I hope I made my problem clear. Thanks in advance, Marjolein

Answer 1

[Walter Roberson]

Something like,

sum(bsxfun(@ge, tcouple(:,1), 0:tmax))

Answer 2

[Oleg Komarov]

Or to apply on the full matrix:

N   = 30;
REP = 10;

tcouple = randi([0 50],N,REP);
t       = 0:max(tcouple(:));
counts  = histc(tcouple,t);
counts  = flipud(cumsum(flipud(counts)));

Oleg