Mike Harding <mvh@ix.netcom.com> writes:
> Any way to adjust n_distinct to be more accurate?

You could try increasing the statistics target for the relevant columns.
What does pg_stats show for the "numdistinct" estimates of the columns
you are grouping over, and does that have anything to do with reality?

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?

http://archives.postgresql.org