evgam: An R Package for Generalized Additive Extreme Value Models

evgam: An R Package for Generalized Additive Extreme Value Models

: generalized extreme value distribution, generalized Pareto distribution, point process, generalized additive model, Laplace’s method, R

This article introduces the R package evgam. The package provides functions for
fitting extreme value distributions. These include the generalized extreme value and
generalized Pareto distributions. The former can also be fitted through a point process
representation. Package evgam supports quantile regression via the asymmetric Laplace
distribution, which can be useful for estimating high thresholds, sometimes used to discriminate between extreme and non-extreme values. The main addition of package evgam
is to let extreme value distribution parameters have generalized additive model forms, the
smoothness of which can be objectively estimated using Laplace’s method. Illustrative
examples fitting various distributions with various specifications are given. These include
daily precipitation accumulations for part of Colorado, US, used to illustrate spatial models, and daily maximum temperatures for Fort Collins, Colorado, US, used to illustrate
temporal models

Benjamin D. Youngman

https://www.jstatsoft.org/article/view/v103i03

University of Exeter

July 2022

Fajar bagus W

PDF

English

Text