Date: Sat, 14 Apr 2007 21:54:56 -0700
Reply-To: David L Cassell <davidlcassell@MSN.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: David L Cassell <davidlcassell@MSN.COM>
Subject: Re: PROC TRANSPOSE, reorg data
In-Reply-To: <BAY123-F2531B33EBE5FB9FE58F0CDE5D0@phx.gbl>
Content-Type: text/plain; format=flowed
tobydunn@HOTMAIL.COM replied:
>David ,
>
>I see our own David Cassell hasnt seen this thread probably due to the fact
>that he is cleaning his crystal ball again so I will step in knowing full
>well I will get plenty of nay sayer emails chastizing me about my advice on
>this but here it goes.
>
>
>Dont molest your data set in this manner. What you will end up with is a
>very short but very wide data set. Which is bad for processing speed and
>bad for coding. Basically you will be holding information which should be
>in your data set as a variables value in you variable names. This just
>makes the job of doing any type of codeing which is data driven and/or
>maintainable to any reasonable degree impossible.
>
>Now I would change your data structure but not in the way that you have
>asked but rather in the opposite direction and make it longer and thinner.
>Since I have no clue what V1 V2 V3 actually are in your data set I will
>have
>to use generic names. Ideally you would want to make these names
>meaningfull.
>
>
>zipcode store_id Num V
>00001 1 1 A
>00001 1 2 B
>00001 1 3 C
>00001 2 1 A
>00001 2 2 B
>00001 2 3 C
>
>
>
>
>
>
>Toby Dunn
I predicted that he'd say that, so I just didn't need to write in. :-)
Let me add that requests to have this sort of display in output can
often be solved using PROC REPORT or PROC TABULATE to do your
heavy lifting.
HTH,
David
--
David L. Cassell
mathematical statistician
Design Pathways
3115 NW Norwood Pl.
Corvallis OR 97330
_________________________________________________________________
MSN is giving away a trip to Vegas to see Elton John. Enter to win today.
http://msnconcertcontest.com?icid-nceltontagline
