| Date: | Fri, 3 Sep 1999 08:35:18 -0400 |
|---|
| Reply-To: | Howard Schreier <Howard_Schreier@ITA.DOC.GOV> |
|---|
| Sender: | "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU> |
|---|
| From: | Howard Schreier <Howard_Schreier@ITA.DOC.GOV> |
|---|
| Subject: | Averages (was:<No subject given>) |
|---|
| Content-Type: | text/plain; charset=US-ASCII |
|---|
* Generate test data;
data original;
array v(20) var1-var20;
do id = 'a','b','c';
do visit = 0 to 5;
do varnum = 1 to 20; drop varnum;
v(varnum) = floor(10*ranuni(1));
end;
output;
end;
end;
run;
* Compute averages;
proc summary data=original(where=(visit=0 or visit=2));
by id;
var var1-var20;
output out=average mean=;
run;
* (signed) Howard_Schreier@ita.doc.gov ;
----- Original Message -----
From: HyeRyung Lee <hrlee@WORLDNET.ATT.NET>
> hello, could anyone solve this question for me?
> I have a dataset which has visits from 0 to 5. make a new data sets which
is
> averaged from visit 0 and visit 2. Each has 20 variables with numeric
> values. Is there any simple way to average the values of variable at one
> step?
> For example,
> data one two;
> set a.original;
> if visit=0 output one;
> if visit=1 output two;
> run;
>
> data one;
> set one;
> rename var1=one_var1;
> rename var2=one_var2;
> .
> .
> .
> proc sort;by id;
> run;
>
> data two;
> set two;
> rename var1=two_var1;
> rename var2=two_var2;
------------------------------
| 
