Re: Group By?

Here's what I want to do.

Table control contains values (mon and valves) that are associated by
numbers inserted into the associated column.

I want to transfer the serial _id number of the items associated by the
value '1' into the appropriate columns of the first row of the table
auto_control. All items associated with the value '2' into the second row -
etc. etc.

Is this best accomplished by a 'group by' command or subset???

Bob
Control

device_ID type association
serial varchar int4

1 mon 1
2 valve 2
3 valve 1
4 mon 2
5 valve 1

Auto_control

loop_id mon valve valve
serial int4 int4 int4
1 1 3 5
2 2 4

----- Original Message -----
From: "Guy Rouillier" <guyr(at)masergy(dot)com>
To: "Postgre General" <pgsql-general(at)postgresql(dot)org>
Sent: Monday, November 21, 2005 4:25 PM
Subject: Re: [GENERAL] Group By?

Converted your message to plain text as preferred on most mailing lists.

Bob Pawley wrote:
> I want to take the serial ID of several values in different rows in
> one table and insert them into a single row of another table.
>
> Would the 'group by' command be the best way to do this?

Could you provide an actual example? The wording of your question is a
little vague and an example might help solicit an answer to the actual
problem. For example, are these serial ID values all in a the same
column in the source table? Or is each one in a different column? And
what is the selection criteria that brings these results together?

--
Guy Rouillier

---------------------------(end of broadcast)---------------------------
TIP 3: Have you checked our extensive FAQ?

http://www.postgresql.org/docs/faq