Re: group by points

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Bruno Wolff III <bruno(at)wolff(dot)to>
Cc: Albe Laurenz <all(at)adv(dot)magwien(dot)gv(dot)at>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: group by points
Date: 2006-05-22 22:38:35
Message-ID: 15608.1148337515@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Bruno Wolff III <bruno(at)wolff(dot)to> writes:
> On Mon, May 22, 2006 at 10:23:28 +0200,
> Albe Laurenz <all(at)adv(dot)magwien(dot)gv(dot)at> wrote:
>> In order to (efficiently) process a GROUP BY clause, you need a
>> total ordering on the data type that you group by, i.e. an ordering
>> such that for any two data x and y you have either x < y or x > x
>> or x = y.

> An equality operator is good enough if the number of unique groups isn't too
> large, so that a hash aggregate plan works efficiently.

Doesn't help for the case at hand, since point_eq isn't marked hashable
either. It would be good to fix things so that the system doesn't
insist on having the sorting option available, though.

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2006-05-22 23:12:02 Re: Porting MSSQL to PGSQL (Was: [OT] MySQL is bad, but THIS bad?)
Previous Message Tom Lane 2006-05-22 22:34:19 Re: pgbench create index anomoly