Question

Say I have the following data:

id1   id2    data
----------------------
7     8      test1
7     8      test2
8     7      test1
8     7      test2

For my needs id1 and id2 can be transposed, so this is essentially duplicate data. How can I select only one combination of id1/id2 and all of their unique values for data? So desired output would be:

7   8   test1
7   8   test2

or

8   7   test1
8   7   test2

It really doesn’t matter which.

Answer 1

if really all records are duplicated this way and id1!=id2, could this work?

select distinct id1, id2, data from test where id1 < id2