[R] Pseudo R2 for Tobit Regression

[Achim Zeileis]

On Sat, 16 Aug 2008, ldotero wrote:

>
> Dear All:
> I need some guidance in calculating a goodness-of-fit statistic for a Tobit
> Regression model.
> To develop the Tobit regression, I used the tobit() method from the AER
> package, which is basically a simpler interface to the survreg() method.
>
> I've read about pseudo R2 and C-index and was wondering if there is a
> package that calculates this for me.  Also, is there a reason to select
> pseudo R2 over C-Index, or viceversa?

McFadden's pseudo-R^2 can be computed via
   fm <- tobit(...)
   fm0 <- update(fm, . ~ 1)
   1 - as.vector(logLik(fm)/logLik(fm0))
using accessor functions or via computing on the internal structure via
   1 - fm$loglik[2]/fm$loglik[1]

hth,
Z

> I will appreciate any suggestions.
>
> Many thanks.
> -- 
> View this message in context: http://www.nabble.com/Pseudo-R2-for-Tobit-Regression-tp19014572p19014572.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> R-help at 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.
>
>