Invalid name with reg and coefplot

[Martin Loriente]

Hi! I am new to this Community, this is my first post. I am trying to create some coefplots with stata using regressions (I am using reghdfestd version 5 so it works) and after this code I find: "__000001- invalid name r(198);

end of do-file r(198); ". I do not have a variable called like that and I am not trying to name one like that. What is the reason for the error?
"gl y_ivs meancact indextab trust_out gl y_gps negrecip

gl fe1ivs ${febaseivs} yearb country year age gl fe1gps ${febasegps} yearb country year age

gl g1 "0 5" gl g2 "6 17" gl g3 "18 25" gl g4 "26 32" gl g5 "33 40" gl g6 "41 50" gl g7 "51 60" gl g8 "61 70" gl g9 "71 200"

u "temp",replace foreach b in ivs gps{ u "temp/b'",replace mer m:1 country year yearb using "temp/2democracy/tot/demAge",keepus(dem1* dem2*) keep(1 3) nogen foreach y in ${y_b'}{

local w = 0 matrix coef = J(9, 3, .) matrix coln coef = coefficients1 lcl1 ucl1 //para guardar coeficiente lowe bound y upper bound matrix rown coef = "Epidemic exposure (0-5)" "Epidemic exposure (6-17)" "Epidemic exposure (18-25)" "Epidemic exposure (26-32)" "Epidemic exposure (33-40)" "Epidemic exposure (41-50)" "Epidemic exposure (51-60)" "Epidemic exposure (61-70)" "Epidemic exposure +71" forv g=1/9 { loc w = `w'+1 loc min=real(word("${g`g'}",1)) loc max=real(word("${g`g'}",2)) preserve drop if !inrange(year-yearb,`min',`max') reghdfestd `y' dem1g`g' if inrange(year-yearb,`min',`max'), a(${fe1`b'}) cl(${cl1}) version(5) mat os = r(table) mat coef[`w',1] = os[1,1] mat coef[`w',2] = os[5,1] mat coef[`w',3] = os[6,1] restore } coefplot (matrix(coef[,1]), ci((coef[,2] coef[,3]))), title("Effect of Exposure to Sickness on `y'",size(medsmall)) graphregion(color(white)) bgcolor(white) horizontal yline(1 2 3 4 5 6 7 8 9, lpattern(shortdash) lcolor(gray)) /// xline(0) ciopts(lwidth(vthick) lcolor(emidblue)) grid(between glpattern(shortdash)) msymbol(O) mfcolor(white) mlabel mlabposition(6) ylabel(,labs(vsmall)) gr export "${savef}/ages`y'.pdf",replace }
}".

[Martin Loriente]

It seems to be related to an intern process of the reghdfestd function renaming some variable because I found "rename meancact __000001" using "set trace on".

[Nick Cox]

Cross-posted at https://www.reddit.com/r/stata/comme...sion_coefplot/

Please note: it's a rule at Reddit and a request here to tell people about cross-posting.

Meanwhile, your code should be pasted between

delimiters to be readable.

[Martin Loriente]

Cross-posted at https://www.reddit.com/r/stata/comme...sion_coefplot/

Please note: it's a rule at Reddit and a request here to tell people about cross-posting.

Meanwhile, your code should be pasted between



delimiters to be readable.

Thanks Nick!

 
gl y_ivs meancact indextab trust_out
gl y_gps negrecip

gl fe1ivs ${febaseivs} yearb country year age 
gl fe1gps ${febasegps} yearb country year age 

gl g1 "0 5"
gl g2 "6 17"
gl g3 "18 25"
gl g4 "26 32"
gl g5 "33 40"
gl g6 "41 50"
gl g7 "51 60"
gl g8 "61 70"
gl g9 "71 200"

u "temp",replace
foreach b in ivs gps{
set trace on
u "temp/`b'",replace
mer m:1 country year yearb using "temp/2democracy/tot/demAge",keepus(dem1* dem2*) keep(1 3) nogen
  foreach y in ${y_`b'}{ 
        
        local w = 0
        
        matrix coef = J(9, 3, .)
        matrix coln coef = coefficients1 lcl1 ucl1 //para guardar coeficiente lower bound y upper bound
        matrix rown coef = "Epidemic exposure (0-5)" "Epidemic exposure (6-17)" "Epidemic exposure (18-25)" "Epidemic exposure (26-32)" "Epidemic exposure (33-40)" "Epidemic exposure (41-50)" "Epidemic exposure (51-60)" "Epidemic exposure (61-70)" "Epidemic exposure +71"
        
        forv g=1/9 { 
            
            loc w = `w'+1    
            loc min=real(word("${g`g'}",1))
            loc max=real(word("${g`g'}",2))
            preserve
            
            drop if !inrange(year-yearb,`min',`max')
        
            reghdfestd `y' dem1g`g' [aw=weight], a(${fe1`b'}) cl(${cl1}) version(5)                
                
            mat os = r(table)
            mat coef[`w',1] = os[1,1]
            mat coef[`w',2] = os[5,1]
            mat coef[`w',3] = os[6,1]
            
            restore
        }
        coefplot (matrix(coef[,1]), ci((coef[,2] coef[,3]))), title("Effect of Exposure to Democracy on `y'",size(medsmall)) graphregion(color(white)) bgcolor(white) horizontal yline(1 2 3 4 5 6 7 8 9, lpattern(shortdash) lcolor(gray)) ///
        xline(0) ciopts(lwidth(vthick) lcolor(emidblue)) grid(between glpattern(shortdash)) msymbol(O) mfcolor(white) mlabel mlabposition(6) ylabel(,labs(vsmall)) 
        gr export "${savef}/ages`y'.pdf",replace        
}
}

[Nick Cox]

Should have been

* code goes here

delimiters.

What is this?

 
 reghdfestd

That's my guess for most suspect.

[Martin Loriente]

Yes, it seems to be a problem with "reghdfestd". It is a function to make regressions standardized with fixed effects that had an update a month ago and did not work anymore, so I was using it 5th version but it seems to make trouble too. I just changed it to "reghdfe". Thanks Nick!