mnlChange2 {DAMisc} | R Documentation |
Calculates average effects of a variable in multinomial logistic regression holding all other variables at observed values.
mnlChange2(obj, varnames, data, diffchange = c("unit", "sd"), n = 1, R = 1500)
obj |
An object of class |
varnames |
A string identifying the variable to be manipulated. |
data |
Data frame used to fit |
diffchange |
A string indicating the difference in predictor values to
calculate the discrete change. |
n |
Number of |
R |
Number of simulations. |
A list with elements:
mean |
Average effect of the variable for each category of the dependent variable. |
lower |
Lower 95 percent confidence bound |
upper |
Upper 95 percent confidence bound |
Dave Armstrong
library(nnet)
data(france)
mnl.mod <- multinom(vote ~ age + male + retnat + lrself, data=france)
mnlChange2(mnl.mod, "lrself", data=france, )