To see the observations in a dataset but not in the other

[manya92]

Hi,

I have two datasets, and i want to see what observations are in dataset2 that are not in dataset1.

How to do that ?

 My code is this 

/*proc sql ;
	create table DATASET1 as
	select distinct enrolid, code, (svcdate) as index_cpt format mmddyy10. 
	from derived._02_cpt ;
quit ;/*179*/

/*proc sql ;
	create table DATASET2 as  
	select distinct a.enrolid, min(a.index_diag, b.index_cpt) as index_date 
        format mmddyy10.
	from derived._02_tr2 as a 
	inner join derived._02_cpt2 as b 
	on a.enrolid = b.enrolid 
	where -7 le (a.index_diag-b.index_cpt) le 7
	order by a.enrolid, index_date;
quit ;/*188*/

[PaigeMiller]

In PROC SQL, do a FULL JOIN, the resulting data set will include all observations from both data sets, and the ones that are in one data set but not the other will be obvious.

[PeterClemmensen · Accepted solution]

If your data is simple, you can so something like this

data data1;
a=1;output;
a=2;output;
a=3;output;
run;

data data2;
a=1;output;
a=3;output;
run;

proc sql;
   create table want as
   select * from data1
   where a not in
   (
   select distinct a from data2
   );
quit;

Otherwise, when you compare two data sets, PROC COMPARE is a pretty good place to start.