Find unique rows comparing 2 colums?

 % code
a = [1,200,400;
     2,222,410;
     3,300,409;
     4, 90, 100;
     5, 320, 420;
     6, 200, 400;
     7, 222, 410;
     8, 350, 405;
     9, 210, 110;
     10, 50, 80;]

if you look at the matrix, the first row at column 2 and 3 has values the same as the 6th row. e.g. a(1,2:3) = a(6,2:3). I m trying to return 'a' where it doesn't include duplicates. so the first, second, sixth and seventh row should only have one instance. Previously I just used a nested for loop to take each row and compare it to all other rows but on a larger scale this is computationally expensive. I was trying out the unique function with rows but it only seems to compare individual row elements to one another rather than compare 2 columns of one row to the other rows.

Answer 1

[Thorsten]

 unique(a(:, 2:end), 'rows')

[Ronan]

thank you for that, although is there a way to keep its original order?

[Star Strider]

Add the ‘setOrder’ argument 'stable' to keep the original order:

unique(a(:, 2:end), 'rows', 'stable')

[Ronan]

my version of matlab doesn't support stable but i found a work around for anybody else wanting to know.

    % code
  a = [1,200,400;
     2,222,410;
     3,300,409;
     4, 90, 100;
     5, 320, 420;
     6, 200, 400;
     7, 222, 410;
     8, 350, 405;
     9, 210, 110;
     10, 50, 80;]
[b,m,n] = unique(a(:, 2:end), 'rows')
values = zeros(length(b),3)
values(:,2:3) = b
values(:,1) = m
Bop = sortrows(values)