Handling Duplicates Observation

[Mutanen Lau]

my data set contains duplicates i want to save the duplicates separately or export it then drop all the duplicates from my data set because i want to work with unique observation. please i need help on how to achieve that.
thanks

[Basharat Hussain]

Kindly check: https://www.stata.com/support/faqs/d...-observations/
-------------------------------------------------------------------------------------------------------------------------
clear
sysuse auto
save "C:\Data\auto.dta", replace
append using "C:\Data\auto.dta"
save "C:\Data\autox2.dta", replace
bysort make: ge dup = cond(_N==1,0,_n)
count if dup == 2
//you can now 'keep' if dup == 2 or 'drop' if dup == 1
keep if dup == 2
tab dup
save "C:\Data\autoex.dta", replace
help cond
--------------------------------------------------------------------------------------------------------------------------

[Roman Mostazir]

Please note you are expected to provide data example using -dataex-, as per the forum rule. Please read through the Forum rules . Assuming you have some sort of unique identification for the observations, here is another way with "egen"..

*Some fake data

 li id fake in 1/16, clean noobs
    id       fake  
     1   .2047095  
     2   .8927587  
     2   .8927587  
     3   .5844658  
     3   .5844658  
     3   .5844658  
     4   .3697791  
     4   .3697791  
     4   .3697791  
     4   .3697791  
     5   .8506309  
     5   .8506309  
     5   .8506309  
     5   .8506309  
     5   .8506309  
     6   .3913819  

so id
egen tagid = tag(id) //identify unique observations

preserve
keep if tagid==0
save duplicates.dta, replace //data with duplicates are saved in current directory
restore
keep if tagid==1
save unique.dta,  replace //data with unique observations are saved in current directory

use unique.dta, clear

li in 1/10, clean noobs

   id       fake   tagid  
     1   .2047095       1  
     2   .8927587       1  
     3   .5844658       1  
     4   .3697791       1  
     5   .8506309       1  
     6   .3913819       1  
     7   .1196613       1  
     8   .7542434       1  
     9   .6950234       1  
    10   .6866152       1

[Basharat Hussain]

Thank you so much.
Apologies for not having used dataex.
Will read rules. Have been away for too long.

[Roman Mostazir]

Thank you so much.
Apologies for not having used dataex.
Will read rules. Have been away for too long.

Hi Basharat, my reply #3 is targeted to the original poster (Mutanen Lau)and not at you. I think the usual forum practice is that unless and otherwise someone is specifically mentioned in the replies, all replies are responses to the question posted by the original poster.

[Mutanen Lau]

Sorry for not using dataex,i will take correction next time. I sincerely appreciate your help.
Thanks

Lau