RE: st: How to include postestimation results (matrix form) in a regression table created by outreg2?

[Nick Cox <n.j.cox@durham.ac.uk>]

Glad you solved your problem. 

Pedantic would be my middle name if it were not John already, but it really is best _not_ to think of -outreg- and -outreg2- as different versions of the same program. That's been incorrect for some years now. They are now quite different programs with broadly similar aims but radically different implementation. 

Nick 
n.j.cox@durham.ac.uk 

Martin Stürmer

Dear Nick,

thank you for this information, especially with respect to the different
versions of outreg.

I have just found a solution by first extracting the matrix, then transform
it to scalars and finally adding them to the table by addstat:

estat bgodfrey, lags(2)
matrix x=r(p)
scalar x1=x[1,1]
scalar x2=x[1,2]
outreg2 using myfile, tex addstat(B-G-test of serial autocorrelation of
order 1,x1,order 2,x2)
scalar drop _all

Nick Cox

You may get a detailed reply from a user of -outreg2- (SSC, as you are asked
to explain). Otherwise the generic advice about support on
-outreg2- given in (for example)

http://www.stata.com/statalist/archive/2010-08/msg01492.html

http://www.stata.com/statalist/archive/2011-10/msg01319.html

still stands. In addition, it may help to be aware that -outreg2- has not
been updated since 2010. That in itself certainly does not rule out there
being a solution for your problem. But it does imply that the discussion in
the documentation of -outreg2- of the quite different program -outreg- is
now seriously out-of-date.

-outreg2- (author Roy Wada) was originally made public as an extension of
-outreg- (SJ, SSC; John Gallup) at a time when it appeared that
-outreg- was not being developed further. But since then John has
extensively rewritten -outreg- and is actively maintaining it.
Evidently neither author wishes to change their program's name.

Nick

On Thu, Jun 14, 2012 at 8:19 AM, Martin Stürmer
<martin.stuermer@uni-bonn.de> wrote:

> sorry for reposting. I would be very grateful, if somebody could help 
> me. I would like to include a matrix of post-estimation results in a 
> table created by outreg2.
>
> In more detail, I run a regression of the type:
> regress x l.x y l.y
>
> I than run a Breusch-Godfrey Test of serial correlation of order 1 and 2:
> estat bgodfrey, lags(2)
>
> I would like to add the two sided p-values of this test to a 
> regression table using outreg2. These values are saved in the 1x2 matrix
r(p).
>
> Using
> outreg2 using myfile, tex addstat(BG-test p-value, r(p)) I obtain the 
> error "type mismatch" due to the fact that r(p) is a 1x2 matrix
>
> How to add this matrix to the regression table?

*
*   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/