[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: AW: AW: Problems with outreg2 after xtlogit

From   Roy Wada <>
To   <>
Subject   RE: st: AW: AW: Problems with outreg2 after xtlogit
Date   Wed, 12 Aug 2009 12:32:48 -0700

> Sorry, "r(se)" is indeed left behind, in a separate -return list- that is
> not mentioned in the help file. So the code below should get you closer to
> your destination:
> outreg2 using t3.doc, adds ("rho", `e(rho)', "sigma u", `e(sigma_u)', 
> "se rho", `r(se)', "log likelihood", `e(ll)') bdec(3) alpha(0.001, 0.01, 0.05) replace
> Martin

It looks like one of those things whether a local macro needs to be evaluated or 
not. For example, all of these work fine.
di r(se)
di `r(se)'

di e(rho)
di `e(rho)'
But this is inconsistent in a consistent way. Strictly speaking they should not 
yield the same result, since e(rho) is different from `e(rho)', but being able 
to evaluate e-class locals would make it consistent with regular locals. As most 
of us are aware, sometimes one or the other will not work for some usage (such as
loops, etc) under some versions of Stata.
Apparently there is something in outreg2 that will let it handle either versions 
of e-class locals (with or without being evaluated) but not for r-class locals. 
This is a little inconsistent, and probably should be fixed. In the meantime 
the workaround suggested by Martine should work fine.
Express your personality in color! Preview and select themes for Hotmail®.
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index