Hendri --
I think Dirk means -ssc install- which works for -estout- and -outreg-
and -outreg2- among other options (I like -estout- myself). I believe
Dirk intends you to compute one of the residuals possible after
-logit-, all of which are described in -help logit postestimation-,
but you are better off using the -resid- option if that's what you
want. You can see where Stata keeps the estimates by typing -eret li-
after an estimation command, and you can use them in various ways with
-matrix- commands. Ex:
sysuse auto, clear
logit for mpg wei rep78
eret li
mat li e(b)
mat li e(V)
predict yhat
gen dirk=(yhat-for)/ sqrt(yhat*(1-yhat))
predict e1, dev
predict e2, resid
predict e3, rst
su e1 e3 e2 dirk
But note:
logit for mpg
predict yhat2
gen dirk2=(yhat-for)/ sqrt(yhat*(1-yhat))
predict r2, resid
su r2 dirk2
On 04 May 2007 15:49:32 +0100, Dirk Nachbar <[email protected]> wrote:
I think you can install within Stata with
net install outreg
I don't think it does store the variance, use this
logit y x
predict yhat
gen res=(yhat-y)/ sqrt(yhat(1-yhat))
sum res
The opposite of a correct statement is a false statement. But the opposite of a profound truth may well be another profound truth.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
