Date:         Tue, 18 Jun 2002 13:24:54 -0400
Reply-To:     Steve Bloom <steve@WHITEHURST-ASSOCIATES.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Steve Bloom <steve@WHITEHURST-ASSOCIATES.COM>
Subject:      Re: Something about missing data treatment
Comments: To: YCF <s007422@CUHK.EDU.HK>
In-Reply-To:  <aemqpp$2psj$1@ijustice.itsc.cuhk.edu.hk>
Content-Type: text/plain; charset="iso-8859-1"

It looks like your data is fixed with, so use column input...


data temp;
input @3 date yymmdd10.
         var2 $ 15-21
         var3    24
         var4    25-26
;
cards   ;
  2001-04-09  s007552  25104278                     6
  2001-04-09  s001123  1  UC  7
  2001-04-09  a554413  2  NA 5
;
run;


-----Original Message-----
From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU]On Behalf Of YCF
Sent: Tuesday, June 18, 2002 7:28 PM
To: SAS-L@LISTSERV.UGA.EDU
Subject: Something about missing data treatment


Hello all,
   Now, I have a dataset which contains 5 columns:
  2001-04-09  s007552  25104278                     6
  2001-04-09  s001123  1  UC  7
  2001-04-09  a554413  2  NA 5

However, for line 1, the variable in column 4 is missing, and there is no
"." here. Is it possible to define a "length of space" between the columns?
That means if the length of space between two variables is greater than a
certain number, say, 8, then it will treat as missing data for this column.
I can't use @18, @23 etc, as the pattern is quite irregular.

Thanks for your help!