I know the idea has come up a few times to do cross-column statistics to
improve plans when the data distributions are dependent. I found a
couple references in the archives:
http://archives.postgresql.org/pgsql-hackers/2006-09/msg02118.php
http://archives.postgresql.org/pgsql-hackers/2006-08/msg00924.php
We can already keep stats for a functional index. Is there a reason we
can't keep stats for a multi-column index?
Out of the two situations outlined by Jim Nasby here:
http://archives.postgresql.org/pgsql-hackers/2006-08/msg00948.php
It would not help with joins, but it would help with columns that are
referred to as a group (e.g. a=1 AND b<20).
Regards,
Jeff Davis