Date:         Fri, 7 Mar 1997 13:13:38 -0500
Reply-To:     YAJUN DU <YDU@DRAFT.DRAFTNET.COM>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         YAJUN DU <YDU@DRAFT.DRAFTNET.COM>
Subject:      Re: proc transpose(2)

     Hi, Frank.Thank you for your response. Before I posted my first
     question about your procedure, I followed your instructions: added
     n=_N_, however, it would not help for solving the problem. Would you
     like give us more explanation?
 
     Regards
 
     ______________________________ Forward Header
     __________________________________ Subject: Re: Proc Transpose and
     Missing Values (A) (A)
     Author:  uunet!SWOV.NL!poppe at INTERNET Date:    3/6/97 5:04 AM
 
 
     Answer in short: read my first answer again.
 
     Somebody (maybe the same person who asked the original question on
     PROC TRANSPOSE ,
     although this time using another mail address):
 
     "YDU@draft.draftnet.COM" 97.03.04, 12:09 >>> YDU>
     YDU>      Dear Frank Poppe, I tested your procedure: YDU>
     YDU>      proc transpose  data=a   out=trans1; YDU>      by item;
     YDU>      id date;
     YDU>      run;
     YDU>
     YDU> and the following are what the SAS says YDU>
     YDU>
     YDU> NOTE: No variables to transpose.
     YDU> NOTE: The data set WORK.TRANS1 has 0 observations and 5
     variables. YDU> NOTE: The PROCEDURE TRANSPOSE used 0.16 seconds.
     YDU>
     YDU>
     YDU> Did I miss something?
 
     Yes.
     You missed the following part of my first answer:
 
     >The only trouble with this code is that there now is no variable left
     to fill th e
     >columns. I added a N=_N_; statement to the first datastep, which
     makes N the var iable
     >PROC TRANSPOSE uses to fill the columns (no VAR statement is
     necessary, only whe n you
     >do not want all the variables used). This also shows how the original
     observatio ns
 
     Good luck.
 
     Frank