| Re: merge IR and KR: duplicate line number in IR [message #23360 is a reply to message #23359] |
Fri, 27 August 2021 22:45  |
bun_2019fall
Messages: 9
Registered: August 2021
|
Member |
|
|
Hi Bruno,
Thank you for much for looking into the file, and I really appreciated it!
The "caseid" solution is almost perfect, except for one file "DRIR21FL":
use "V:\DHS raw survey/DRIR21DT/DRIR21FL.DTA", clear
*duplicates by caseid
duplicates tag caseid, generate(dup)
sort caseid
tab dup /*4 duplicates*/
*complete duplicates
duplicates tag, gen(dup1)
tab dup1 /*no duplicates*/
The "caseid" itself has 4 duplicates in this dataset, out of all DHS IR files that I have downloaded as of July 2021. I further checked, these observations are not complete duplicates. That said, to proceed with data analysis, I wonder if I should just drop the 4 observations? I wonder if DHS has any best practice regarding this issue? Thank you, again!
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
