Date:         Tue, 26 Nov 2002 10:00:35 -0600
Reply-To:     "Suzanne D. McCoy" <smccoy@LUCIDAN.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Suzanne D. McCoy" <smccoy@LUCIDAN.COM>
Subject:      Re: using proc sql to insert blank rows into a data set
Comments: To: jsdavid1@HOTMAIL.COM
In-Reply-To:  <df56059d.0211260737.74ec0ad4@posting.google.com>
Content-Type: text/plain; charset=iso-8859-1

> Listers,
>
>
> I am working with a data set that has the following structure.
>
>
> var1     var2     var3    var4
> 4        alpha    2       1
> 7        alpha    6       3.7
> 2.8      zap      16      29
> 8        alpha    3       13
> 87       alpha    36      45
> 23       zap      22      21
>
>
> I would like for the data set to have the following form:
>
>
> var1     var2     var3    var4
> 4        alpha    2       1
> 7        alpha    6       3.7
> <*  N number of blank rows here, i.e., insertion point  *>
> 2.8      zap      16      29
> <*  N number of blank rows here, i.e., insertion point    *>
> 8        alpha    3       13
> 87       alpha    36      45
> <*  N number of blank rows here, i.e., insertion point    *>
> 23       zap      22      21
>
>
> Is this possible using proc sql?  I have seen some sophisticated
> techniques using a data step in the archives.  However, I will be
> dealing with multiple insertion points in the same data set.  Perhaps
> there is another strategy to tackle this problem?  Any help or
>suggestions would be appreciated.
>
> Best,
>
> JSD


Use case statements for conditional processing in proc sql that will set
the values to missing for the applicable variables.  As far as other
strategies go, why do you need totally blank insertion points stored in a
dataset?
--
Suzanne D. McCoy
Lucid Analytics Corp.
"Intelligence Unleashed"