-nl- is indicating, inarticulately, that there is a
problem. Perhaps you ended up in a strange corner of
parameter space. Try different starting values. Be
reluctant to believe a fit without reasonable and
indeed finite standard errors.
0 for missing is, I think, a historical hangover
from days in which matrices could not include
missing values, so I support your surmise.
Nick
[email protected]
Harald Badinger
> does anyone know how to interpret standard errors of 0 for part of
> the coefficients in the nonlinear least sqaures output.
>
> Actually, the estimation output does not report standard errors, t-
> and p- values for some of the parameters; however, looking at the
> standard errors using the _se[]command or the e(V) command, the
> values displayed are "0".
> (I guess the opposite is true: that the standard errors are
> extremely large).
>
*
* 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/
