Dear Reinhard
On 2 January 2014 14:12, Reinhard Hössinger
<reinhard.hoessin...@boku.ac.at> wrote:
> I want to estimate an equation system with 3 nonlinear continuous
> equations and one discrete choice (using multinomial logit for the
> latter). For the nonlinear continuous equations, function nlsystemfit
> {systemfit} seems to be appropriate. But what's about the discrete
> choice? Its error component has a logistic distribution. Can it still be
> incorporated in the equation system? Or can/must the error component be
> treated in some way to be incorporated? I cannot find a reference to
> discrete choices in the systemfit package description nor somewhere
> else.
I guess that the econometric estimation of this type of model has not
been implemented in any ready-to-use software package. Hence, you
probably need to derive the log-likelihood function of this model,
implement this function (and preferably also the derivatives/gradients
with respect to the parameters), and maximize this log-likelihood
function, e.g. with the R package "maxLik" (http://maxlik.org/).
Best regards,
Arne
--
Arne Henningsen
http://www.arne-henningsen.name
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
