I have 4 separate datasets (A, B, C and D) with a different number of records in each.   A   B   C   D a1   b1   c1   d1 a2   b2   c2   d2 a3   b3   c3   d3 a4   b4       d4 a5   b5         a6                 I need to produce a single dataset which would show all the possible distinct combinations, pulling 2 records at a time from the A dataset, and one at a time from the other datasets.  Something like : Part1 Part2 Part3 Part4 Part5 a1 a2 b1 c1 d1 a1 a2 b2 c1 d1 a1 a2 b1 c2 d1 a1 a2 b1 c1 d2 ... ... ... ... ...   Obviously this is a simplified version of the problem I have.  I'm not familiar with SAS/IML but after exploring the possible solutions, IML articles keep showing up as a great option when dealing with combinations.  Is it ?  I've read a few of Rick Wicklin's articles on dealing with combinations, but they usually show a single list of items to work from.  In my case, I have 4 list of items (A, B, C and D) and their lenght are all different, so they wouldn't form a typical matrix.  Would a DATA step be a better solution ?       ... View more