求解stata一比一配对循环

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

想要为660个测试公司找到最接近的660个一比一控制样本。控制条件为年份year，行业industry，和ROA。要找到最就近的ROA值的公司。应该要用滚动的去配对，并且要把配对样本另外储存或者用一个新变量做标记。

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏1 回帖

关键词：Stata tata Industry Dust year industry 行业 样本

method1: check the joinby in Stata
use sample.dta, clear
ren roa roa_sample
merge 1:m year industry using match.dta, unmatched(master)
gen diff = abs(roa_sample - roa)
gsort year industry diff
by year industry: keep if _n == 1

Method2
use sample, clear
gen id = _n
local record = `c(N)'
forvalue i = 1 / `record' {
use sample, clear
keep if id == `i'
local a = year[1]
local b = industry[1]
local c = roa[1]

use match, clear
keep if year = `a'
keep if industry = `b'
gen diff = abs( roa - `c')
sort diff
keep if _n == 1
gen id = `i'
tempfile f`i'
save `f`i''
}

use `f1', clear
forvalue i = 2  / `record' {
append using `f`i''

占座

ctx5518 发表于 2013-7-13 14:01
method1: check the joinby in Stata
use sample.dta, clear
ren roa roa_sample

谢谢哈，我之前找了一个例子改了一下
你的这个没试过，但看起来应该是可信的
谢谢你的回复哦