Derivation
variable: repgp (table[s]: css_seh, gp_gphhold, caps_register)
variable: cluster (table[s]: css_seh, gp_gphhold, caps_register)
variable: hhlistno (table[s]: css_seh, gp_gphhold, caps_register)
variable: house (table[s]: NCD_combined_data_all_final [derived table])
variable: ghhid (table[s]: css_seh, gp_gphhold)
egen crshse=concat(repgp cluster hhlistno) if repgp~=""
gen crshse_ghhid=crshse /*create household identifier mainly crshse, but some only ghhid*/
replace crshse_ghhid=ghhid if crshse==""