Learn R Programming

BMAmevt (version 1.0.5)

Multivariate Extremes: Bayesian Estimation of the Spectral Measure

Description

Toolkit for Bayesian estimation of the dependence structure in multivariate extreme value parametric models, following Sabourin and Naveau (2014) and Sabourin, Naveau and Fougeres (2013) .

Copy Link

Version

Install

install.packages('BMAmevt')

Monthly Downloads

186

Version

1.0.5

License

GPL (>= 2)

Maintainer

Leo Belzile

Last Published

April 21st, 2023

Functions in BMAmevt (1.0.5)

Diagnostics for the MCMC output in the PB and NL models.

Pairwise Beta (PB) and Nested Asymmetric Logistic (NL) distributions

Image and/or Contour plots of spectral densities in trivariate extreme value models

PB and NL spectral densities on the two-dimensional simplex

Posterior distribution the probability of joint threshold excess, in the NL model.

excessProb.condit.dm

Probability of joint threshold exceedance, in the Dirichlet Mixture model, given a DM parameter.

excessProb.condit.pb

Estimates the probability of joint excess, given a PB parameter.

Discretization grid builder.

excessProb.condit.nl

Probability of joint threshold excess in the NL model

Example of valid Dirichlet mixture parameter for tri-variate extremes.

Logit transformation

marginal.lkl.nl

Marginal likelihoods of the PB and NL models.

Estimates the probability of joint excess (Frechet margins)

Exponent function in the NL model.

Laplace approximation of a model marginal likelihood by Laplace approximation.

Maximum likelihood optimization

Acceptance probability in the MCMC algorithm.

Inverse logit transformation

Marginal model likelihood

Multivariate data set with margins following unit Frechet distribution.

posterior.predictive.nl

Posterior predictive densities in the three dimensional PB, NL and NL3 models

Posterior predictive density on the simplex, for three-dimensional extreme value models.

posterior.predictive3D

Posterior predictive density on the simplex, for three-dimensional extreme value models.

posteriorMCMC.nl

MCMC posterior samplers for the pairwise beta and the negative logistic models.

Default hyper-parameters for the NL model.

Default hyper-parameters for the Pairwise Beta model.

Default MCMC tuning parameter for the Pairwise Beta model.

posteriorDistr.bma

Posterior distribution in the average model

Positive alpha-stable distribution.

Default MCMC tuning parameter for the Nested Asymmetric logistic model.

Density integration on the two-dimensional simplex

Linear coordinate transformations

Dirichlet distribution: random generator

Logarithmic score and \(L^2\) distance between two densities on the simplex (trivariate case).

PB model: proposal distribution

Five-dimensional air quality dataset recorded in Leeds(U.K.), during five winter seasons.

NL3 model: proposal distribution.

Prior parameter distribution for the Pairwise Beta model

Prior parameter distribution for the NL model

MCMC sampler for parametric spectral measures

posteriorWeights

Posterior model weights

Multivariate data set with margins following unit Frechet distribution.

Plots the Dirichlet mixture density on a discretization grid

Generic Monte-Carlo integration of a function under the prior distribution

cons.angular.dat

Angular data set generation from unit Frechet data.

Angular density/likelihood function in the Dirichlet Mixture model.

MCpriorIntFun.nl

Generic Monte-Carlo integration under the prior distribution in the PB and NL models.

Adds graphical elements to a plot of the two dimensional simplex.

ddirimix.grid1D

Univariate projection or marginalization of a Dirichlet mixture density on on [0,1]

BMAmevt-package

Bayesian Model Averaging for Multivariate Extremes

Tri-variate ‘angular’ data set approximately distributed according to a multivariate extremes angular distribution