hi ... had a similar idea but used a view since that intermediate data set is just a means to an end (also used two ID variables as mentioned by Ksharp in another discussion) ... data new / view=new; set have; by id period; count + (-first.period * count) + 1; var = time; name = 'T'; output; var = conc; name = 'C'; output; run; proc transpose data=new (drop=_name_) out=want; var var; by id period; id name count; run;
... View more
