Well since David Cassell seems to be preoccupied and you have some solutions, Ill ask the Caseell Question here... Why do you want to create non-atomic variables? IF it is for reporting purposes okay I can live with that. However, if you plan on doing any calculations with these values you may as well chunk it in the trash as the more you have to do with these types of variables the harder and harder the code will be to create, debug, maintain, and the time to execute the code will increase.Toby Dunn Compromise is like telling a lie, it gets easier and easier. Each compromise you make, that becomes your standard. Perfection doesnt exist, once you reach it, its not perfect anymore. It means something else.> Date: Wed, 3 Oct 2007 06:59:19 -0700> From: ram.sasha@GMAIL.COM> Subject: how to flag duplicates ?> To: SAS-L@LISTSERV.UGA.EDU> > I have a dataset in the following format:> > id var1> 1 xy> 1 bc> 1 cn> 2 lm> 2 lc> 3 cm> 3 cn> 4 fr> 5 gt> 6 uy> > I want the above dataset in 2 different formats:> format 1: concatenate all the var1 values into one variable> id var1> 1 xy,bc,cn> 2 lm ,lc> 3 cm,cn> 4 fr> 5 gt> 6 uy> > > format 2: requires a dup flag in the duplicate entries> > id var1 duplicates> 1 xy> 1 bc 1> 1 cn 1> 2 lm> 2 lc 1> 3 cm> 3 cn 1> 4 fr> 5 gt> 6 uy> > > Thanks for the help> Ram _________________________________________________________________ Windows Live Hotmail and Microsoft Office Outlook – together at last.  Get it now. http://office.microsoft.com/en-us/outlook/HA102225181033.aspx?pid=CL100626971033