Date:         Sun, 24 Oct 1999 13:10:46 +0100
Reply-To:     oliver@agyinc.com
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         oliver <oliver@AGYINC.COM>
Organization: Verio
Subject:      Re: ANSWER for:shortcut of using multi-variable in crosscorr=()
Content-Type: text/plain; charset=us-ascii

Hi, Thanks for the help, following is the code which is works.
Enjoy:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
option PAGESIZE=32767 nodate ;

%macro multi;
%do i = 1 %to 10;
COL&i
%end;
%mend;
options mprint;/* this way you will see how the code is generated in the
log*/
LIBNAME oliSAS '/export/home/opm/sas';
*PROC PRINT DATA = oliSAS.BV2TS;
*RUN;
proc arima data=oliSAS.BV2TS;
*identify var=COL1 nlags=2;
*run;
*estimate p=3;
*run;
identify var=COL1 crosscorr=(%multi) nlags=4;

oliver wrote:

> Dear SAS expert,
>
> I try to use some short cut to represent thousands of variable in some
> PROC. However, I find I can not use shortcut like var1-var3 to represent
> var1 var2 var3 in crosscorr=().
>
> What I try to do is get the Crosscorrelations between one and thousand
> of other variable in ARIMA as following:
>
> proc arima data=oliSAS.BVTS;
> identify var=COL1
> crosscorr=(COL1 COL2 COL3 COL4 COL5 COL6 COL7 COL8 COL9 COL10
> ....COL6000)
>
> However SAS do not allow me to do:
>
>  crosscorr=(COL2 - 6000)
> or
> crosscorr=(COL2 - COL6000)
>
> Does anyone know why and how to handle it?
>
> Many thanks
> --
> Oliver

--
Oliver Zhao
~~~~~~~~~~~
Bioinformatic Scientist
AGY Therapeutics, Inc.
Two Corporate Drive
South San Francisco, CA 94080
Phone:(650) 615-4532
FAX:  (650) 615-4544
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
URL link:
AGY's data page:  http://op.agyinc.com
My travel page:   http://travel.to/zhao