Date:         Wed, 9 May 2007 22:27:03 -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: re-SurveySelect not selecting where sample size exceeds
              sampling units
In-Reply-To:  <OFF4B2414E.76C9050B-ON852572D6.00478D47-852572D6.00487EFE@census.gov>
Content-Type: text/plain; format=flowed

roger.s.clark@CENSUS.GOV replied:
>
>Hi, All;
>
>I have never posted to this list, before.  I copied the address from the
>"reply to" box, so I hope I'm responding to the list and not sending a
>command to the server.
>
>I also encountered this problem when I had to take over coding on a project
>that was behind schedule.  This was one of the errors that the code was
>having.  I searched the online doc and found no solution within proc
>surveyselect.
>
>I solved it by running proc summary to get the number of observations in
>each category.  Then I merged the output dataset from proc summary to my
>dataset.  Then I used a conditional to send either my predetermined sample
>size, or the number of observations (whichever was smaller) to proc
>surveyselect.  Clumsy, but I was in a hurry and needed to use coding that I
>am very familiar with.
>
>I'm sure there must be a more elegant solution, somewhere...
>
>Thanx,
>Roger S. Clark
>Address Coverage Assurance Programs Branch

There is now.  However, when you had to code this up, there
may not have been.  SELECTALL is (I think) a SAS 9.1 feature.

HTH,
David
--
David L. Cassell
mathematical statistician
Design Pathways
3115 NW Norwood Pl.
Corvallis OR 97330

_________________________________________________________________
See what you’re getting into…before you go there
http://newlivehotmail.com/?ocid=TXT_TAGHM_migration_HM_viral_preview_0507