gk (version 0.6.0)

isValid: Check validity of g-and-k or g-and-h parameters

Description

Check whether parameter choices produce a valid g-and-k or g-and-h distribution.

Usage

isValid(
  g,
  k_or_h,
  c = 0.8,
  model = c("gk", "generalised_gh", "tukey_gh", "gh"),
  initial_z = seq(-1, 1, 0.2)
)

Value

Logical vector denoting whether each parameter combination is valid.

Arguments

g

Vector of g parameters.

k_or_h

Vector of k or h parameters.

c

Vector of c parameters.

model

Which model to check: "gk", "generalised_gh" or "tukey_gh". For backwards compatibility, "gh" acts the same as "generalised_gh".

initial_z

Vector of initial z values to use in each optimisation.

Details

This function tests whether parameter choices provide a valid distribution. Only g k and c parameters need be supplied as A and B>0 have no effect. The function operates by numerically minimising the derivative of the quantile function, and returning TRUE if the minimum is positive. It is possible that a local minimum is found, so it is recommended to use multiple optimisation starting points, and to beware that false positive may still result!

Examples

isValid(0:10, -0.5)
isValid(0:10, 0.5, c=0.9, model="generalised_gh")
isValid(0:10, 0.5, model="tukey_gh")