| From: | "Scott Marlowe" <smarlowe(at)qwest(dot)net> |
|---|---|
| To: | "Sean Davis" <sdavis2(at)mail(dot)nih(dot)gov> |
| Cc: | "Postgres" <pgsql-novice(at)postgresql(dot)org> |
| Subject: | Re: Where clause efficiency using "IN" |
| Date: | 2004-10-22 15:42:16 |
| Message-ID: | 1098459735.21035.97.camel@localhost.localdomain |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-novice |
On Fri, 2004-10-22 at 09:07, Sean Davis wrote:
> Just a quick general question: Can someone comment on using where
> clauses like:
>
> (sample = 2 OR sample = 3 OR sample = 4)
>
> as compared to
>
> sample in (2,3,4)
>
> in terms of efficiency?
I believe they are pretty much the same, in that the planner will turn
the in clause you have there into a series of ors. However, if it is a
subselect, then the planner can use a hashed aggregate method, as long
as the dataset can fit into memory, as defined by some setting, (I think
it's sort_mem, but I'm not completely sure).
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2004-10-22 15:56:36 | Re: Where clause efficiency using "IN" |
| Previous Message | Sean Davis | 2004-10-22 15:07:45 | Where clause efficiency using "IN" |