First observation from panel data

[Salome-Joelle Gass]

I have the following panel data of one individual in the study. WHO stage was recorded multiple times, but I only want to identify the first record of WHO stage, make a new variable for this first WHO stage and copy that stage to all observations of the panel data for the individual. The data looks as follows:

* Example generated by -dataex-. For more info, type help dataex
clear
input double id byte who_stage float visit_lab_date
829 .     .
829 4 20898
829 .     .
829 .     .
829 .     .
829 . 20919
829 .     .
829 .     .
829 .     .
829 .     .
829 3 20954
829 .     .
829 .     .
829 . 21014
829 .     .
829 .     .
829 . 21133
829 .     .
829 .     .
829 .     .
829 1 21336
829 .     .
829 1 21336
829 .     .
829 . 21417
829 . 21417
829 . 21417
829 . 21417
829 . 21417
829 1 21427
829 .     .
829 .     .
829 . 21427
829 .     .
829 .     .
829 .     .
829 . 21623
829 .     .
829 1 21623
829 . 21735
829 .     .
829 1 21735
829 .     .
829 .     .
829 1 21882
829 .     .
829 1 21882
829 .     .
829 . 21910
829 .     .
829 .     .
829 . 21939
829 .     .
829 . 21970
829 . 21999
829 . 21999
829 . 21999
829 . 21999
829 . 21999
829 1 22092
end
format %td visit_lab_date

Thanks in advance for all your time and help!

[Carlo Lazzaro]

Salome_Joelle:
do you mean:

. bysort id ( visit_lab_date ): gen wanted= who_stage[1]

[Salome-Joelle Gass]

I used exactly that, but for some cases (e.g. id 42) it didn't pick up on this as shown below:

* Example generated by -dataex-. For more info, type help dataex
clear
input long rowo byte who_stage float(visit_lab_date who_stage_first)
42 . 21369 .
42 . 21369 .
42 . 21381 .
42 . 21381 .
42 . 21381 .
42 . 21381 .
42 . 21381 .
42 . 21381 .
42 . 21381 .
42 . 21573 .
42 . 21573 .
42 1 21573 .
42 . 21573 .
42 . 21573 .
42 1 21753 .
42 . 21753 .
42 1 21934 .
end
format %td visit_lab_date

There was 3 instances of a WHO stage being recorded but all the observations of the new variable are missing

[Andrew Musau]

Maybe you want:

bys rowo (visit_lab_date): gen firstob= sum(sum(!missing(who_stage)))==1
bys rowo: egen wanted= max(cond(firstob, who_stage, .))