lfstat (version 0.9.12)

check_distribution: Checks if a Distribution is suited

Description

Most distributions are used for modelling either minima or maxima. Sometimes a better fit can be achieved by reversing the distribution. This functions helps to decide if the reversed distribution is advisable.

Usage

check_distribution(extreme = c("minimum", "maximum"), distribution,
                   def = list(minimum = c(),
                              maximum = c("gev")))

Value

a character vector as long as distribution containing the optimal choice for the given distributions under the constraints of def.

Arguments

extreme

character vector, describing the kind of extreme value to be fitted. Either 'minimum' or 'maximum'.

distribution

character vector of length one. Distribution chosen by the user.

def

a list of length two, containing the elements 'minimum' and 'maximum'.

Examples

# Using the Weibull distribution for minimum values is a good choice
check_distribution(extreme = "minimum", distribution = "wei")

# ... whereas the GEV is meant for maxima.
# Therefore the reversed distribution is suggested.
check_distribution(extreme = "minimum", distribution = "gev")