用 loop 來計算統計量與跑迴歸，都不是好 idea。請直接說明你要做什麼？

谢谢黄老师，刚学stata不久，对这块还不是很熟悉。

是在AER上看到的一篇文章的程序，是想通过循环计算出相应的参数值。但我在stata中运行时，总提示在forvalues语句后提示 invalid syntax,不知道问题出在哪里，请黄老师和各位高手指点。
gen phiG=.
gen varzG=.
gen varzB=.
gen cov=.
gen varx=.
gen nG=.
gen nB=.
gen nTot=.

gen phiB=.
gen sigmasqx=.
gen sigmasqy=.
gen sigmasqzG=.
gen sigmasqzB=.
gen beta=.
gen lambdaG=.
gen lambdaB=.

local rownum = 0

local rowsboth = 5

local goodsample = "wbdqtotal>3.5"
local badsample = "wbdqtotal<3.5"

forvalues j = 1(1)`rowsboth' {
        local rownum = `rownum' + 1
        replace phiG = 1.1 - `j'*0.1 if _n==`rownum'
        qui: corr lngdpwdilocallongdiff lndnlongdiff if `goodsample', cov
        replace varzG = r(Var_1) if _n==`rownum'
        replace nG = r(N) if _n==`rownum'
        qui: corr lngdpwdilocallongdiff lndnlongdiff if `badsample', cov
        replace varzB = r(Var_1) if _n==`rownum'
        replace nB = r(N) if _n==`rownum'
        qui: corr lngdpwdilocallongdiff lndnlongdiff, cov
        replace varx = r(Var_2) if _n==`rownum'
        replace cov = r(cov_12) if _n==`rownum'
        replace nTot = r(N) if _n==`rownum'

        replace sigmasqy = varzG * phiG if _n==`rownum'
        replace phiB = sigmasqy / varzB if _n==`rownum'
        replace beta = cov / sigmasqy if _n==`rownum'
        replace sigmasqx = varx - cov if _n==`rownum'
        replace sigmasqzB = varzB - sigmasqy if _n==`rownum'
        replace sigmasqzG = varzG - sigmasqy if _n==`rownum'

        replace lambdaB = sigmasqx*sigmasqy / ( sigmasqx*sigmasqy + sigmasqzB*(beta*beta*sigmasqy + sigmasqx) ) if _n==`rownum'
        replace lambdaG = sigmasqx*sigmasqy / ( sigmasqx*sigmasqy + sigmasqzG*(beta*beta*sigmasqy + sigmasqx) ) if _n==`rownum'
}

local rownum = `rownum' + 1
replace phiB = phiB[2] if _n==`rownum'
local phiB = phiB[2]
qui: corr lngdpwdilocallongdiff lndnlongdiff if `badsample', cov
replace varzB = r(Var_1) if _n==`rownum'
replace nB = r(N) if _n==`rownum'
replace varx = r(Var_2) if _n==`rownum'
replace cov = r(cov_12) if _n==`rownum'
replace nTot = r(N) if _n==`rownum'

replace sigmasqy = varzB * phiB if _n==`rownum'
replace beta = cov / sigmasqy if _n==`rownum'
replace sigmasqx = varx - cov if _n==`rownum'
replace sigmasqzB = varzB - sigmasqy if _n==`rownum'

replace lambdaB = sigmasqx*sigmasqy / ( sigmasqx*sigmasqy + sigmasqzB*(beta*beta*sigmasqy + sigmasqx) ) if _n==`rownum'

keep  phiG phiB beta lambdaG lambdaB sigmasqy sigmasqzB sigmasqzG sigmasqx
order phiG phiB beta lambdaG lambdaB sigmasqy sigmasqzB sigmasqzG sigmasqx

其实很多老手也是会用 loop 来计算 (如同我所说，bad idea)。

老师那您能看出forvalues循环语句这块有什么问题吗？因为想要搞懂论文中的参数是怎么得到的