In short, I need to get the importance of the variables after a zero-inflated regression, with all my predictors being dichotomous factors. I tried something like this:
library(pscl) library(relaimpo) data("bioChemists", package = "pscl") fm_zip2 <- zeroinfl(art ~ . | ., data = bioChemists) relImportance <- calc.relimp(fm_zip2, type = "lmg", rela = TRUE)
But I only get error messages, such as that the resulting model must be of the class "glm" or like the following one (from the previous code):
Error in calc.relimp.default.intern(object = list(coefficients = list( : If x is NULL, then object must be a data frame or a matrix.
How can I fix this? Thanks!