Thanks Austin and Martin for the discussion.
Austin's code did run for me few hours ago, however, I could not obtain the same results when I ran it again. Any idea? I am using Stata Version 11.
---------- Error message --------
. x2
option eret not allowed
r(198);
---------------------------------
Undesired Results are as follows:
. est sto re
.
. esttab re, scal(estat estatp j jp)
----------------------------
(1)
ys
----------------------------
k 0.00562
(0.73)
cap -0.00124
(-0.87)
_cons 4.644***
(621.14)
----------------------------
N 1031
estat
estatp
j
jp
----------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
Regards,
Kelvin
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Martin Weiss
Sent: Thursday, 29 October 2009 6:23 PM
To: [email protected]
Subject: AW: st: Can I produce a regression output with -outreg2- after -xtoverid2, noisily- ? and temporary names query.
<>
Austin`s code does run for me, and does deliver the desired result. To
appreciate the difference, you have to make -esttab- display the standard
errors as - esttab re, scal(estat estatp j jp) se(%10.9fc)-.
Note it is necessary to reload the -xtoverid2- command from Austin`s site,
as the latest version dated yesterday now returns r-class results that
should enable Kelvin to capture his desired results...
HTH
Martin
-----Ursprüngliche Nachricht-----
Von: [email protected]
[mailto:[email protected]] Im Auftrag von Kelvin Tan
Gesendet: Mittwoch, 28. Oktober 2009 23:51
An: [email protected]
Betreff: RE: st: Can I produce a regression output with -outreg2- after
-xtoverid2, noisily- ? and temporary names query.
Thanks Austin for the codes,
I tried to run the following codes, but I got an error message when I ran
the third last line -x2-. Please see the error message.
. x2
conformability error
r(503);
. est sto re
. esttab re, scal(estat estatp j jp)
----------------------------
(1)
ys
----------------------------
k 0.00562
(0.73)
cap -0.00124
(-0.87)
_cons 4.644***
(621.14)
----------------------------
N 1031
estat
estatp
j
jp
----------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001
The clustered standard errors that I want to capture are from -xtoverid2,
noisily cl(id)-.
I want
----------------------------
k 0.00562
(0.011)
cap -0.00124
(-0.002)
_cons 4.644***
(0.010)
----------------------------
----------------------------------------------------------------------------
--
| Robust
__00000O | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+--------------------------------------------------------------
--
__00000Q | .0056218 .0106049 0.53 0.596 -.0151635
.026407
__00000Z | -.0012407 .0018955 -0.65 0.513 -.0049558
.0024745
__00000M | 4.643618 .0095373 486.89 0.000 4.624925
4.662311
----------------------------------------------------------------------------
--
Regards,
Kelvin
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Austin Nichols
Sent: Thursday, 29 October 2009 2:00 AM
To: [email protected]
Subject: Re: st: Can I produce a regression output with -outreg2- after
-xtoverid2, noisily- ? and temporary names query.
net from http://www-personal.umich.edu/~nicholsa/stata
net install xtoverid2, replace
clear all
prog x2, eclass
version 8.2
tempvar es
tempname b1 b V
g byte `es'=e(sample)
mat `b'=r(b)
mat `V'=r(V)
mat `b1'=e(b)
loc cn: colnames `b1'
mat colnames `b'=`cn'
mat colnames `V'=`cn'
mat rownames `V'=`cn'
loc y=e(depvar)
loc N=e(N)
eret post `b' `V', dep(`y') e(`es') obs(`N')
eret local cmd "xtoverid2"
eret scalar estat=r(estat)
eret scalar estatp=r(estatp)
eret scalar j=r(j)
eret scalar jp=r(jp)
end
webuse abdata
qui xtivreg ys cap (k=emp wage), re
xtoverid2, cl(id) noi
x2
est sto re
esttab re, scal(estat estatp j jp)
On Wed, Oct 28, 2009 at 10:38 AM, Martin Weiss <[email protected]> wrote:
>
> <>
>
>
>
> You will probably have to dig into the code of -xtoverid2- to capture the
> results returned by -ivreg2-. The -xtoverid- commands subject the initial
> estimations to transformations, as in line 454, hence the tempvars...
>
>
>
> HTH
> Martin
>
> Von: [email protected]
> [mailto:[email protected]] Im Auftrag von Kelvin Tan
> Gesendet: Mittwoch, 28. Oktober 2009 14:57
> An: [email protected]
> Betreff: RE: st: Can I produce a regression output with -outreg2- after
> -xtoverid2, noisily- ? and temporary names query.
>
> Thanks Martin and Austin,
>
> -xtivreg, re- produces non-robust standard errors while -xtoverid, noisily
> cl(id)- produces clustered standard errors. I want to capture the
clustered
> standard errors from -xtoverid, noisily cl(id)- not non-robust standard
> errors from -xtivreg,re-.
>
> Martin said:
> -xtoverid2-, a modification of -xtoverid- by Austin, gets the internal
> results it displays with the unintelligble names of tempvars from
-ivreg2-,
> as far as I can tell from my log file. So there could be a case for a
> -saveinternal- option along the lines of the -savefirst- option for
> -ivreg2-...
> Does anyone have any examples that I can change unintelligable names of
> tempvars from -xtivreg, re- to my label names?
>
> Regards,
> Kelvin
> ________________________________________
> From: [email protected] on behalf of Martin Weiss
> Sent: Wed 28/10/2009 11:06 PM
> To: [email protected]
> Subject: AW: st: Can I produce a regression output with -outreg2- after
> -xtoverid2, noisily- ? and temporary names query.
>
> <>
>
>
> But I think Kelvin wanted to capture the standard errors displayed by the
> -noisily- option to -xtoverid2-, and they do differ from the ones returned
> by -xtivreg-...
>
>
>
> HTH
> Martin
>
>
> -----Ursprüngliche Nachricht-----
> Von: [email protected]
> [mailto:[email protected]] Im Auftrag von Austin
Nichols
> Gesendet: Mittwoch, 28. Oktober 2009 14:00
> An: [email protected]
> Betreff: Re: st: Can I produce a regression output with -outreg2- after
> -xtoverid2, noisily- ? and temporary names query.
>
> Kelvin Tan <[email protected]> :
> Note that -xtoverid2- is merely replaying results you already have
> when you specify the -noisily- option. You can add the statistics
> produced by -xtoverid2- like so:
>
> webuse abdata, clear
> xtivreg ys cap (k=emp wage), re
> est sto re
> xtoverid2, cl(id)
> estadd scalar estat=r(estat)
> estadd scalar estatp=r(estatp)
> estadd scalar j=r(j)
> estadd scalar jp=r(jp)
> esttab re, scal(estat estatp j jp)
>
> if you have -estout- and the other packages installed.
> Probably a similar approach works with -outreg2-.
>
> On Wed, Oct 28, 2009 at 4:28 AM, Kelvin Tan <[email protected]>
> wrote:
>> Dear All,
>>
>> I would like to know if I can produce a regression output with -outreg2-
>> in LaTeX after-xtoverid2, noisily cl(id)- command. When I ran the
>> following commands, I only managed to get the "Results 1" from -
>> xtivreg, re - but not "Result 2" from -xtoverid, noisily cl(id)-. I
>> would also like to know if there is a command that I can use to change
>> the temporary names to my variable label names after xtoverid. Thanks in
>> advance.
>>
>> Commands:
>> net install xtoverid2,
>> from(http://www-personal.umich.edu/~nicholsa/stata)
>> webuse abdata, clear
>> xtivreg ys (k = emp wage cap), re
>> xtoverid2, noisily cl(id)
>>
>> outreg2 using myfile, tex bdec(3) tdec(2) alpha(0.001, 0.01, 0.05) label
>> ctitle(test) adds("Endogeneity Pvalue", `r(estatp)', "FE vs RE",
>> `r(j)',"FE vs RE pvalue", `r(jp)' )
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
