cobin (version 1.0.1.3)

glm.cobin: Find the MLE of cobin GLM

Description

Find the maximum likelihood estimate of a cobin generalized linear model with unknown dispersion. This is a modification of stats::glm to include estimation of the additional parameter, lambda, for a cobin generalized linear model, in a similar manner to the MASS::glm.nb. Note that MLE of regression coefficient does not depends on lambda.

Usage

glm.cobin(
  formula,
  data,
  weights,
  subset,
  na.action,
  start = NULL,
  etastart,
  mustart,
  control = glm.control(...),
  method = "glm.fit",
  model = TRUE,
  x = FALSE,
  y = TRUE,
  contrasts = NULL,
  ...,
  lambda_list = 1:70,
  link = "cobit",
  verbose = TRUE
)

Value

The object is like the output of glm but contains additional components, the ML estimate of lambda and the log-likelihood values for each lambda in the lambda_list.

Arguments

formula, data, weights, subset, na.action, start, etastart, mustart, control, method, model, x, y, contrasts, ...

arguments for the stats::glm without family and offset.

lambda_list

(Default 1:70) an integer vector of candidate lambda values. Note that MLE of coefficient does not depends on lambda

link

character, link function. Default cobit. Must be one of "cobit", "logit", "probit", "cloglog", "cauchit".

verbose

logical, if TRUE, print the MLE of lambda.

Details

Since dispersion parameter lambda is discrete, it does not provide standard error of lambda. With cobit link, we strongly encourage Bayesian approaches, using cobin::cobinreg() function.

Examples

 
# \donttest{
requireNamespace("betareg", quietly = TRUE)
library(betareg)# for dataset example
data(GasolineYield, package = "betareg")
# cobin regression, frequentist
out_freq = glm.cobin(yield ~ temp, data = GasolineYield, link = "cobit")
summary(out_freq) 
# cobin regression, Bayesian
out = cobinreg(yield ~ temp, data = GasolineYield, 
               nsave = 10000, link = "cobit")
summary(out$post_save)
plot(out$post_save)
# }