Hi Arild, thanks for the advice and sorry for any confusion. My data file has variable columns like this: a b c ran1 ran2 ran3 ran4 ran5

and I am trying to replace the random variable that is included each time the genmodsummary macro is run. In the example above I would want to run the genmodsummary 5 times. The first run would contain ran1, the second ran2, and so on until (in this example) the fifth run contained ran5. So, the first run include the following models: yvariable= GP04, model1 = a ran1, model2 = b ran1, model3 = c ran1, model4 = a, model5 = b, model6 = c, And the 5th run would include the following models: model1 = a ran5, model2 = b ran5, model3 = c ran5, model4 = a, model5 = b, model6 = c , I also want the output from each run of genmodsummary to be saved in a seperate outfile. any help would be very appreciated! thanks jared

On 8/31/05, Arild S <kog@ssb.no> wrote: > > On Wed, 31 Aug 2005 06:23:04 -0400, Jared Wilson <wilsonjaredm@GMAIL.COM> > wrote: > .... > >Basically I want to call a macro (genmodsummary)100 times, replacing the > >ran1 variable with 100 different randomly generated variables (ran1 to > >ran100 in the data sheet) each time. > .... > > Hi, > I'm not sure if I understand you correctly, > but can't you simply nest the %genmodsummary call like this: > > %macro manygenmod(times); > %do i=1 %to &times.; > %genmodsummary( > data= GPTRAIN, > class= , > yvariable= GP04, > model1 = a ran&i., > model2 = b ran&i., > model3 = c ran&i., > model4 = a, > model5 = b, > model6 = c , > distribution= poisson, > link= log, > number_of_models= 6, > sortby=daic, > out=GP&i. > ); > %end; > %mend; > %manygenmod(100); > > You'll get GP1 - GP100 on the output side (I hope..). > > Arild >