Combining data sets.

[attjooo]

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;

[Kurt_Bremser]

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

[LinusH]

Or do the sort after the data - set step.

And you wish to keep all records, including duplicates?

[attjooo]

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

Your question indicates that my code example create duplicates.

[LinusH]

Not creates, preserves rather...

[Kurt_Bremser]

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?

[attjooo]

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.

[Kurt_Bremser]

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

[LinusH]

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

[Reeza]

What about using an update statement instead?