Re: benchmarking the query planner

Alvaro Herrera <alvherre@commandprompt.com> writes:
> Is there any way to "merge" the statistics?  i.e. if a full table scan
> is done to compute precise statistics, and later a regular analyze scan
> is done, then perhaps instead of clobbering the previous stats, you
> merge them with the new ones, thus not completely losing those previous
> ones.

Seems like a pretty hard problem unless you store a whole lot more
statistics state than we do now (which of course would create its own
costs).  How would you know which portion of the old stats to not
believe anymore?
        regards, tom lane