extra.parameter {PLreg}

Procedure to Select the Extra Parameter for PLreg Objects

Description

The extra.parameter function is used to select the extra parameter of some power logit models. It provides plots of -2logLik and the Upsilon measure (see Queiroz and Ferrari (2022)) versus \zeta, the extra parameter.

Usage

extra.parameter(object, lower, upper, grid = 10, graph = TRUE)

Arguments

object

fitted model object of class "PLreg".

lower

a numeric value representing the lower limit of the interval for the extra parameter.

upper

a numeric value representing the upper limit of the interval for the extra parameter.

grid

a positive integer representing the number of points in the plots. Default is grid=10. If grid is less than 10, then grid=10.

graph

logical. If graph = TRUE the plots are shown, if graph = FALSE the plots are not shown. Default is graph = TRUE.

Value

extra.parameter returns a list with five objects:

zeta.Ups

The selected zeta based on the Upsilon measure.

zeta.loglik

The selected zeta based on -2logLik.

zeta.values

The values of zeta used in the graphs.

Upsilon.values

-2logLik evaluated at each value of zeta.

loglik.values

Upsilon measure evaluated at each value of zeta.

References

Queiroz, F. F. and Ferrari, S. L. P. (2022). Power logit regression for modeling bounded data. arXiv:2202.01697.

See Also

PLreg

Examples

data("bodyfat_Aeolus")

#Initial model with zeta = 2
fit1 <- PLreg(percentfat ~ days + sex + year, data = bodyfat_Aeolus,
             family = "PE", zeta = 2)
summary(fit1)
# Choosing the best value for zeta

extra.parameter(fit1, lower = 1, upper = 4, grid = 15)