Mike Orr wrote: > I have a complex query question whose answer I think would help me to > understand subselects and aggregates better. I have a table with four > columns of interest: > > id (int primary key), loc_title (varchar null), loc_value (float > null), loc_unit (varchar null) > > I want the output columns to be: > (1) each distinct value of loc_title, sorted > (2) an id of a record containing that loc_title > (3) the loc_value for the record in column 2 > (4) the loc_unit for the record in column 2 > > I don't care as much how the records for columns 2-4 are chosen. It > could be max(loc_value), min(id), or something else. I just need some > sample records to test my program against. > > Is this something I should be able to do with a single query with a > subselect, or is it too much for one query? I tried a few ways and > none of them were syntactically valid. Sorry to disappoint you, but you won't learn a lot about subselects and aggregates with that: SELECT DISTINCT ON (loc_title) loc_title, id, loc_value, loc_unit FROM mytable ORDER BY loc_title; Yours, Laurenz Albe -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
