I want to have two datasets split evenly from a much larger dataset but before the split, SAS must not split obersvations that are in same zipcodes (criteria). We want to mail an food offer from a dataset we created but with two different valid dates. The first valid date is assigned to the first splited dataset and second valid dates assigned to the second splited dataset. However, we don't want to separate married couple who want to dine together. Our manual logic is to sort by zip codes and split the datasets evenly and right before the split make sure no two address are the same. Below is what our rows of data look likes.
ID
FN
LN
Address
City State
state
ZIP
1111111
Wayne
Smith
111 Left St
BROOKLYN
NY
11234
1111112
Mary
Smith
111 left
BROOKLYN
NY
11234
A dataset with 10,000 rows split evenly:
Dataset 1 = 5,000
Dataset 2 = 5,000
But don't separate same zip codes.
Please help!
... View more