[R] fitting multinomial logistic regression

[Achim Zeileis]

On Sun, 6 Jun 2010, suman dhara wrote:

> Sir, I want to fit a multinomial logistic regression in R.I think 
> mlogit() is the function for doing this. mlogit () is in packege 
> globaltest.But, I can not install this package. I use the following:

To add to some of the previously given advice:
   - The "globaltest" package provides an mlogit() function that estimates
     a multinomial logit model using (as far as I can see) sum contrasts
     for the coefficients.
   - The "nnet" package provides the function multinom() for the
     multinomial logit model using treatment contrasts for the
     coefficients.
   - The "mlogit" package provides a different function mlogit() that
     provides a wider range of multinomial logit models using not only
     individual-specific variables (as the previous two functions) but
     also alternative-specific variables.

If you have alternative-specific variables, you should probably look at 
the "mlogit" package. If not, then I would personally prefer the 
multinom() function in "nnet" because it provides a few more extractor 
functions and is described in the MASS book by Venables & Ripley.

In addition to that, there are more packages for different flavors of 
multinomial models, including the multinom family in the "VGAM" package, 
among others...

hth,
Z

> install.packages("globaltest")
>
> Can you help me?
>
> Regards,
> Suman Dhara
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> 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.
>