[R-sig-teaching] multinomial logit models

[Robert Glaß]

Dear R community,

Is anybody aware of how to do diagnostics for multinomial logit models?
In the binomial case (for GLM's), I have already used the functions 
provided by J. Fox in his car package.
I estimate my model with multinom (nnet).
Now I would like to do an analysis of the residuals, influence measures, 
nonlinearity and collinearity among the predictors.
How is generally controlled for the results in MNL using R?
Thank you very much in advance

Best regards,

Robert Glass