Well, very good responses recieved from Reeza, ChrisNZ, Ron_Fehd_macro_maven, and Quentin.
Each of them shed very intereting insights and comments.
It was found that my dataset A has n=xxxx,xxxx number of UNIQUE records not observed in dataset B. Similarly, dataset B has n=xxx,xxx number of Unique records not observed in dataset A.
So, I learned that Proc Compare should not be employed, but data step or SQL is a best solution. Case closed.
Thanks to all who have responded.
Girish Patel
... View more