Thank you for the advice..
Ideally I would like to write a function, but I am just learning Postgres
and not familiar with creating functions like "greatest_int".
In this case, the two values represent to different data streams. For the
information I need, I only need the greatest of the two.
To work around this, I've started using two separate SELECT statements with
the comparison as part of the WHERE command. I also just tried the SELECT ..
UNION SELECT .. command, and it appears to have worked.
Thanks,
Greg
-----Original Message-----
From: David Hartwig [mailto:daybee@bellatlantic.net]
Sent: Wednesday, December 02, 1998 8:25 PM
To: Greg Youngblood
Cc: 'PostgreSQL SQL List'
Subject: Re: [SQL] Please help: How to determine largest of two numbers
in a query?
My $.03 worth, without really knowing what these values represent:
There is no such "standard" query that can conditionally select between two
columns. Fortunately PostgreSQL will allow you to create function - of one
does not already exist.
The function would be something like:
greater_int(integer a, integer b)
It takes two integers arguments and returns the greater integer.
Then you do something like:
CREATE VIEW myview AS
SELECT yyymmdd, key, greater_int(value1, value2) FROM mytable;
Or just do the SELECTs directly.
This may not be the most efficient way to accomplish your goal. I would
also
take a good look at may table design. Value1 and value2 MAY be a repeating
group and you may want to normalize a little more. Or not.
