SAS Join vs Union

[Vinz867]

Hi, I have this piece of code below:

data data3;
merge data1(in=a) data2(in=b);
by id
if a or b;
run;

When I read this code, I am the impression that this will be like a full outer join. However, when I run this program, it becomes more of a union than a full outer join. Please keep in mind that data1 has 2 columns, including the ID and date while data2 has 3 - ID, date and description. 

I want to understand as to why data3 resulted into a union rather than a full outer join. Is there a specific instance where it will be a full outer join than a union?

Any insight will be helpful. Thanks!

[SASKiwi]

Your example is a match-merge, which is a limited version of a full outer join, not a union. The MERGE statement can only join one-to-many distinct values of the BY variable. If there are repeats in BY values in both the first and second datasets, only the repeats from the first dataset are kept.

[Vinz867 · Accepted solution]

Thank you! I think I understand this now!