# Ranking rows in a matrix without changing order of elements in rows

1 view (last 30 days)
J OBrien on 16 Feb 2016
Commented: J OBrien on 16 Feb 2016
Hi there
I have a 4x4 matrix of 1s and 0s, eg A = [1 0 0 1 0; 0 1 1 1 1; 1 0 1 1 0; 0 0 0 1 0]
In my code I am calling each row, so A(1,:),A(2,:) and so forth.
How can I rank each row according to how many 1's are in the row? Without changing the order of the elements in that row. So ideally I would like:
B = [0 1 1 1 1; 1 0 1 1 0; 1 0 0 1 0; 0 0 0 1 0]
I thought about summing each row and then sorting them, but then how would I be able to continue referencing A(1,:), A(2,:)?
Background: I am trying to code a simple genetic algorithm and so I need to crossover the rows that have the most 1's.
Thanks

MHN on 16 Feb 2016
Edited: MHN on 16 Feb 2016
A = [1 0 0 1 0; 0 1 1 1 1; 1 0 1 1 0; 0 0 0 1 0]
B = sum(A')';
[m,n] = sort(B);
now A(n(1),:) shows the raw with the least 1's, A(n(2),:) shows the next raw with the least 1's, and so on. By this method you havent changed the order of A's raw and you also now where is the row with the smallest number of 1's and so on.
J OBrien on 16 Feb 2016
Thank you so very much