Excluding*data entries

[Patrick Mulcahy]

Hi,

I have something that is confusing me. I have a dataset and want to merge it with another, but before I can do that I have to drop all entries which are the same personID, same varA, but different varB.

I want to table 1 to become table 2.

Is there a way to do this?

I have been trying to use this explanation, https://www.stata.com/support/faqs/d...-observations/ but have not been able to get it to work yet.

Thanks.

Table 1

PersonID	varA	varB
1	a	ab
1	a	ab
2	c	cd
3	a	ab
3	a	ac
4	z	zy
4	z	zy
5	z	zy
5	z	zx

Table 2

PersonID	varA	varB
1	a	ab
1	a	ab
2	c	cd
4	z	zy
4	z	zy

[Andrew Musau]

* Example generated by -dataex-. To install: ssc install dataex
clear
input byte personid str1 vara str2 varb
1 "a" "ab"
1 "a" "ab"
2 "c" "cd"
3 "a" "ab"
3 "a" "ac"
4 "z" "zy"
4 "z" "zy"
5 "z" "zy"
5 "z" "zx"
end
 
egen group= group(vara)
bys group (personid): gen fob=_n==1
bys personid: egen tokeep = max(fob)
keep if tokeep

Result:

. l

     +-----------------------------------------------+
     | personid   vara   varb   group   fob   tokeep |
     |-----------------------------------------------|
  1. |        1      a     ab       1     1        1 |
  2. |        1      a     ab       1     0        1 |
  3. |        2      c     cd       2     1        1 |
  4. |        4      z     zy       3     0        1 |
  5. |        4      z     zy       3     1        1 |
     +-----------------------------------------------+