文章预览
前情回顾 清洗数据 cd C : \ Download
import exc AF_CFEATUREPROFILE , first clear
drop in 1 / 2
g year = substr ( Acc , 1 , 4 )
keep if regexm ( Acc , "-12-" )
duplicates drop Stkc y , force
destring * , replace
egen n = rowtotal ( * Att * )
keep Stkc y n
merge 1 : 1 S y using 行业代码_2022 , nogen
bys S : fillmissing IndustryCode
g i = substr ( IndustryCode , 1 , 1 )
replace i = IndustryCode if i == "C"
replace n = 0 if mi ( n )
drop I *
bys i y : egen c = count ( n )
drop if c == 1
bys i y : egen r = rank ( n )
bys i y : egen max_r = max ( r )
bys i y : egen min_r = min ( r )
g InfoAuth = 1 - ( r - min_r ) / ( max_r - min_r )
keep S y In
la var I 企业信息可靠性
drop if mi ( I )
drop if I == 1 | I == 0
save 企业信
………………………………