Re: Proposal - improve eqsel estimates by including histogram bucket numdistinct statistics

"Nathan Boley" <npboley@gmail.com> writes:
> If we query on values that aren't in the table, the planner will
> always overestimate the expected number of returned rows because it (
> implicitly ) assumes that every query will return at least 1 record.

That's intentional and should not be changed.  I can't see the value
of allowing fractional-row estimates anyway.  Now if we can *prove*
the query matches no rows, that's a whole nother matter, but statistics
won't help us with that.
        regards, tom lane