Date:         Sat, 12 Jul 1997 07:46:34 -0400
Reply-To:     Lucy Liu <lucyl@HAWAII.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Mike Schexanyder <IRMIKE@LSUMVS.SNCC.LSU.EDU>
Organization: Louisiana State University
Subject:      Re: macro assistance?

In article <l03102809afdf0b294d93@[164.111.164.4]>,
"jerry mcgwin, jr." <mcgwin@EYES.UAB.EDU> writes:
 
>proc reg outest=reg_parm; model yy = xx0  xx1  xx2  xx3;
>next i want to use the values of the parameter estimates as values in
>another procedure. for example:
>proc nlin;
>parms b0 = 7.70  b1=3.81 b2=-3.30 b3=1.17 b4=-1.78
>model y = b0 + b1*xx0 + b2*xx1 + b3*xx2 + b4*xx3;
>where the values of b0--b4 in the PARMS statement are the coefficients from
>the previous procedure but the actual dataset used for NLIN is the SAME as
>that used for REG. i think that some MACRO programming could accomplish
>this. any advice? TIA.
>
CALL SYMPUT assigns the value of a variabe to a macro variable.
  CALL SYMPUT("b1", parm1);
 model y = b0 + &b1*xx0 + &b2*xx1 + &b3*xx2 + &b4*xx3;
--
 Mike Schexnayder                       irmike@llsu.edu
 Budget and Planning
 311 Thomas Boyd Hall                   voice  (504) 388-1231
 Louisiana State University             fax    (504) 388-5963
 Baton Rouge, Louisiana  70803