To be honest, I am looking at that code and wondering what process could posible warrant a) 5 datasets with essentially the same data, and b) that code to get to it. Makes no sense at all. The essence of by group processing is to add groups within the existing data, then use those rather than creating lots of copies of data, then having to code loops over it etc.
data want;
set have;
<series of if statements>
<assign group value>
run;
proc print data=want;
by group_value;
run;
... View more
