bayesfit: Traces from MCMC output

In tsxtreme: Bayesian Modelling of Extremal Dependence in Time Series

Description

Test or show objects of class "bayesfit".

Usage

is.bayesfit(x)

Arguments

x	an arbitrary R object.

Details

Default plot shows samples of residual densities (which==1), residual distribution with credible interval (5\% and 95\% posterior quantiles; which==2), and joint posterior distribution of α and β (which==3) for each lag successively. which can be any composition of 1,2 and 3.

Value

An object of class "bayesfit" is a list containing MCMC traces for:

a,b	Heffernan-Tawn parameters.
sd,mean,w	standard deviations, means and weights of the mixture components.
prec	precision parameter of the Dirichlet process.
ci	auxiliary variable; components' indices for each observation.
noo	number of observations in each mixture component.
noc	number of non-empty components in the mixture.
prop.sd	standard deviations of the proposal distributions for a and b.

And len, the length of the traces, i.e., the number of iterations saved.

See Also

bayesparams, stepfit

tsxtreme documentation built on April 24, 2021, 1:07 a.m.