topic Re: Combining data sets. in SAS Programming

Combining data sets.

attjooo — Tue, 29 Apr 2014 08:31:56 GMT

I have three data sets ONE, TWO and THREE.

They have the same variables: ID, VAR1, VAR2, VAR3 and VAR4.

Each ID can have multiple records. Each ID can be member of more than one of the data sets: 1, 2 or all three. If so, the ID:s records are the same in all the data sets where the ID occurs.

I want to have each ID:s all records in one data set: RESULT.

If I first sort ONE, TWO and THREE by ID, will the following code produce what I want, or can it be done even simpler?

DATA RESULT;

SET ONE TWO THREE;

BY ID;
RUN;

Re: Combining data sets.

Kurt_Bremser — Tue, 29 Apr 2014 08:42:17 GMT

Your solution will give you work.result sorted by id and (implicitly) by occurence in one, two, three.

Re: Combining data sets.

LinusH — Tue, 29 Apr 2014 08:44:06 GMT

Or do the sort after the data - set step.

And you wish to keep all records, including duplicates?

Re: Combining data sets.

attjooo — Tue, 29 Apr 2014 10:35:52 GMT

I want duplicates only if the duplicates were there from the start.

Your question indicates that my code example create duplicates.

Re: Combining data sets.

LinusH — Tue, 29 Apr 2014 11:17:31 GMT

Not creates, preserves rather...

Re: Combining data sets.

Kurt_Bremser — Tue, 29 Apr 2014 11:53:47 GMT

The question is, should dataset result have x records where x is the sum of all records of one,two,three, regardless of the contents, or should records be eliminated if two identical records are found in different datasets?

Re: Combining data sets.

attjooo — Tue, 29 Apr 2014 12:57:43 GMT

I give an example:

ID = 1 have a total of 2 reccords in data set ONE.

ID = 1 does not occur in data set TWO..

ID = 1 occurs also in data set THREE with the same 2 records as in data set ONE.

In data set RESULT, ID = 1 should have the same 2 records as in data set ONE and THREE, and NOT those 2 records twice.

Re: Combining data sets.

Kurt_Bremser — Tue, 29 Apr 2014 13:00:18 GMT

Is it possible to have completely identical records in one of the datasets?

Re: Combining data sets.

LinusH — Tue, 29 Apr 2014 13:04:17 GMT

Then you need to add a PROC SORT NODUPRECS after the data - set step.

Re: Combining data sets.

Reeza — Tue, 29 Apr 2014 14:26:50 GMT

What about using an update statement instead?