Provides fit statistics (pseudo R-squared values) and the Fagerland, Hosmer and Bonfi (2008) specification test for Multinomial Logistic Regression models.

mnlfit(obj, permute = FALSE)

Arguments

obj

An object of class multinom

permute

Logical indicating whether to check all base categories for the Fagerland et. al. specification test.

Value

A list with elements:

result

Fit statistics.

permres

The results of the base category permutation exercise.

References

Fagerland M. W., D. W. Hosmer and A. M. Bonfi. 2008. ‘Multinomial goodness-of-fit tests for logistic regression models.’ Statistics in Medicine. 27(21): 4238-4253.

Examples

library(nnet) data(france) mnl.mod <- multinom(vote ~ age + male + retnat + lrself, data=france)
#> # weights: 35 (24 variable) #> initial value 872.315349 #> iter 10 value 655.272636 #> iter 20 value 559.902797 #> iter 30 value 551.176433 #> final value 551.169697 #> converged
mnlfit(mnl.mod)
#> Warning: 42% of expected counts < 5
#> $result #> Estimate p-value #> Fagerland, Hosmer and Bonfi 45.0880444 0.06232676 #> Count R2 0.5977860 NA #> Count R2 (Adj) 0.3292308 NA #> ML R2 0.5990507 NA #> McFadden R2 0.3100393 NA #> McFadden R2 (Adj) 0.2737367 NA #> Cragg-Uhler(Nagelkerke) R2 0.6322150 NA #> #> attr(,"class") #> [1] "mnlfit"