how to delete rows with some repeated elements?

for example I need to delete one out of two rows whose numbers in column 1&2 are the same. At the same time the deleted row should have a smaller number in column 3. I know the unique command but don't know how to use in this situation. Thanks in advance. I have a=[1,2,3; 1,2,4; 5,6,7; 5,6,8];I want to get b=[1,2,4;5,6,8]

Answer 1

[Stephen23]

One way using sortrows and unique:

>> a = [1,2,3;1,2,4;5,6,7;5,6,8];
>> b = sortrows(a,3);
>> [~,idx] = unique(b(:,1:2),'rows','last');
>> b = b(idx,:)
b =
   1   2   4
   5   6   8

The sortrows call sorts rows by the third column (i.e. smaller values come before larger), then we select the unique rows (only columns 1 & 2), taking the last one each time (which because the rows are sorted will be the larger value in the third column).

[Iris Li]

Thanks again.