/*File to fix the IHIS year and quarter data back to its original format and generate new NHISID variables*/

clear
clear matrix
clear mata

set logtype text

set mem 8g
set more off

log using FixIHIS_pre1967.txt, replace

use ihis_00006.dta, clear
preserve
keep if year>1967
save after1967.dta, replace
restore
keep if year<1968

gen imp_year=substr(nhispid,1,4)
destring imp_year, replace
gen imp_quarter=substr(nhispid,5,1)
destring imp_quarter, replace
gen imp_psurandr=substr(nhispid,6,3)
gen imp_week=substr(nhispid,9,2)
gen imp_segment=substr(nhispid,11,2)
gen imp_hhid=substr(nhispid,13,2)
gen imp_person=substr(nhispid,15,2)

gen year2=year
replace year2=year+1 if imp_quarter>2
*gen quarter2=imp_quarter
gen quarter2=1 if imp_quarter==3
replace quarter2=2 if imp_quarter==4
replace quarter2=3 if imp_quarter==1
replace quarter2=4 if imp_quarter==2

gen str year3=string(year2, "%04.0f") 
gen str quarter3=string(quarter2, "%01.0f")

egen nhispid2=concat(year3 quarter3 imp_psurandr imp_week imp_segment imp_hhid imp_person)
replace nhispid=nhispid2

replace year=year2
replace quarter=quarter2

drop imp_* quarter2 year2 year3 quarter3 nhispid2

save pre1967.dta, replace
append using after1967.dta

save newihis.dta, replace

log close