Classification: UNCLASSIFIED Caveats: NONE My problem with that is that you are counting rows per region and you have done a left jion on region. That means there will be at least one row per region even if there are 0 compliants. It might yield the same result now, but if you even have a period where a region recieves no complaints they will have a complaint count of 1 instead of 0. That is just my guess based on the logic.
-----Original Message----- From: pgsql-sql-ow...@postgresql.org [mailto:pgsql-sql-ow...@postgresql.org] On Behalf Of Richard Broersma Sent: Tuesday, June 16, 2009 3:50 PM To: Rob Sargent Cc: pgsql-sql@postgresql.org Subject: Re: [SQL] left outer join on more than 2 tables? On Tue, Jun 16, 2009 at 2:40 PM, Rob Sargent<robjsarg...@gmail.com> wrote: > Is there a city without a reference to region? I don't know, but the OP wanted to know complaints by region. > And wouldn't you want to count(cm.id)? Count(cm.id) and Count(*) produce the same result. But I like Count(*) more since it helps to correctly express the idea that we are counting rows per group and not cm.id(s) per group. -- Regards, Richard Broersma Jr. Visit the Los Angeles PostgreSQL Users Group (LAPUG) http://pugs.postgresql.org/lapug -- Sent via pgsql-sql mailing list (pgsql-sql@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-sql Classification: UNCLASSIFIED Caveats: NONE -- Sent via pgsql-sql mailing list (pgsql-sql@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-sql
