Question

I have a table like this:

ID | name
=================================
 1 | apple
 1 | bear
 2 | cactus
 2 | dog
 2 | apple
 3 | apple
 3 | dog
==================================

and I wish to get the following unique set of combinations per ID as output from an SQL query:

===================
apple  | bear
apple  | cactus
apple  | dog
cactus | dog
===================

It’s essentially a type of transpose. Ideally, I’d like to expand this to include a count of co-occurances, like this:

===================
1 | apple  | bear
1 | apple  | cactus
2 | apple  | dog
1 | cactus | dog
===================

but this may be above and beyond. I know I can accomplish all of the above using a stored proc; was just curious if I can do this as a query.

Any SQL will do, but MySQL would be preferred if possible.

Thanks for any and all input!

Answer 1

From your example, it looks like you want the first column to contain whatever name comes first alphabetically, so I used a < when joining the table to itself.

select count(*), t1.Name, t2.Name
from 
    tblThings t1 join
    tblThings t2 ON t1.ID = t2.ID AND t1.Name < t2.Name
group by t1.Name, t2.Name

And here’s the query in action to prove I’m not crazy.