How to merge part of a dataset with another dataset

[m1986MM]

Hello everyone,

I want to merge two datasets, having a common variable. However, I don't want them to merge completely. For example, the first dataset is like this:

Table 1
StudentID	Date	Test_Score1	TEST_score2	test_score3
*	*	*	*	*

The other dataset contains these variables:

Table 2
StudentID	Field_of_interest	Undergradute_field	Master_field
$	$	$	$

The studentID is the common variable. But I want to add just one variable (Field_of_interest) from table 2 to table one. Is there any way to not merge two tables completely?

I appreciate any suggestion.

[PGStats · Accepted solution]

The simplest way is with proc SQL :

proc sql;

create table Table_3 as

select Table_1.*, Table_2.Field_of_interest

from Table_1 left join Table_2

  on Table_1.StudentID=Table_2.StudentID;

quit;

PG

[m1986MM]

Thanks. But this code will merge all the datasets together.

[Reeza]

Then you haven't tried it...

[m1986MM]

Yes, that's right.

[m1986MM]

Hi again,

I have a question about the mechanism of this code and appreciate if you can help me. When I merge the datasets by using proc sql, for example in this specific code, those observations in table 1 with the missing value of table 2 (Field_of _interest) would be deleted or not?

Thanks.

[PGStats]

All observations from Table 1 will be included, even those without a matching record in Table 2. If you want to keep only matching records, replace left join with inner join.

PG

[m1986MM]

Thanks.

[naveen20jan]

Hi , you can use below part of  code too.

data want  ;
merge t1  (in = a)
      t2 (in = b  keep = StudentID Field_of_interest );
by StudentID ;
if a ;
run;