Date: Mon, 30 Dec 2002 10:55:49 -0700
Reply-To: Jack Hamilton <JackHamilton@FIRSTHEALTH.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Jack Hamilton <JackHamilton@FIRSTHEALTH.COM>
Subject: Re: the order of variables shown in a table
Content-Type: text/plain; charset=us-ascii
I'd just like to point out this item on the SASWare Ballot:
8 provide a data set option to reorder variables in a data set
Anyone who thinks that a general way of specifying the order of
variables in a data set is important should vote for this item.
The ballot doesn't specify how they're going to implement this if it's
accepted, but the software architectss have probably read, at least in
synopsis, the design considerations we've discussed on SAS-L.
My proposal was an ORDER= data set option which could be placed on
either an output data set or an input data set. This method has the
advantage of allowing different data sets output from a single data step
to have different orders, which an ORDER statement in the data step
itself would not provide.
--
JackHamilton@FirstHealth.com
Manager, Technical Development
METRICS Department, First Health
West Sacramento, California USA
>>> "David L. Cassell" <Cassell.David@EPAMAIL.EPA.GOV> 12/27/2002 4:31
PM >>>
Jack already gave a thorough discussion about the use (and
abuse) of RETAIN. I have a different concern. Are you
interested in re-ordering:
(1) the variables in a data set,
(2) the variables in a PROC PRINT or other output,
(3) the variables as displayed in the SAS Viewer, or
(4) the variables as displayed in PROC FSLIST
?
The answer to your question depends on your answer to this
question. Unless your answer is #1, the RETAIN solution is
probably not your best option. This question has come up in
SAS-L many times, in many forms, and many experts have written
about it. So you can always look in the archives of SAS-L to
find a slew of responses on each of these.
HTH,
David
--
David Cassell, CSC
Cassell.David@epa.gov
Senior computing specialist
mathematical statistician
