hurdle: Loglikelihood adjustments for pscl::hurdle fits

In chantrics: Loglikelihood Adjustments for Econometric Models

Description

NOTE: Hurdle models are currently not supported.

Details

Adjust the loglikelihood and the standard errors of a fitted pscl::hurdle() model.

Note that the pscl::hurdle() model has to be run with the option x = TRUE in order for the adjustment to execute properly. The functions residuals.chantrics() and fitted.chantrics() are currently disabled for hurdle models. Additionally, sequential anova.chantrics() are not available.

Supported families

Within each family, any link function should work.

• geometric

• poisson

• negbin

• binomial (for the zero mass distribution only)

Examples

# hurdle model from AER, pg. 139-140
library(pscl)
data("RecreationDemand", package = "AER")
rd_hurdle <- hurdle(trips ~ . | quality + income,
  data = RecreationDemand,
  dist = "negbin", x = TRUE
)
summary(rd_hurdle)

# adjust model
# adj_loglik(rd_hurdle)

chantrics documentation built on Sept. 29, 2021, 9:08 a.m.