advice on using duplicates drop command

[Joonho Kim]

Hello,

I would like to make use of "duplicates drop" command to drop out certain rows,
For example, my original dataset as follows


Disno statea namea stateb nameb strtyr
173 365 RUS 2 USA 1958
173 710 CHN 2 USA 1958
608 365 RUS 2 USA 1958
608 365 RUS 2 USA 1958
1124 90 GUA 2 USA 1958
2187 731 PRK 2 USA 1958
2187 731 PRK 2 USA 1958
2187 731 PRK 2 USA 1958
2187 731 PRK 2 USA 1958
2215 365 RUS 2 USA 1958
2216 365 RUS 2 USA 1958
2857 339 ALB 2 USA 1958


I would like to remain the rows as below

Disno statea namea stateb nameb strtyr
173 365 RUS 2 USA 1958
173 710 CHN 2 USA 1958
608 365 RUS 2 USA 1958
1124 90 GUA 2 USA 1958
2187 731 PRK 2 USA 1958
2857 339 ALB 2 USA 1958

Although Disno 173 has a duplicate, since it has a different combination like statea 365 and state 710, I would like to remain these two rows.
My purpose is to remain Disno depending on a specific statea and stateb combination.

If I use "duplicates drop Disno, force", one of 173 would be dropped out.

Instead, I tried to make a distinct row identifier like "gen dyad_year=statea*1000000000+stateb*10000+ strtyr" and used
"duplicates drop dyad_year, force", but it dropped out many unwanted rows.

Are there any tips to make use duplicates more efficiently?

Thank you

[Rich Goldstein]

you did not use -dataex- making it too much work (at least for me) to enter your data; please see the FAQ

on the other hand, it appears that just using

duplicates drop

is all you need to do

however, I see that the two lines starting "2215" and "2216" in your first grouping do not appear in the second and I don't understand why they are dropped (they will not be dropped with the code I provided above)

[Joonho Kim]

Thank you for the answer!