rDynMix {FitDynMix}

Simulating a dynamic lognormal-Pareto mixture

Description

This function fits a dynamic mixture by Approximate Maximum Likelihood and by standard maximum likelihood. Currently only implemented for the lognormal - generalized Pareto case, with Cauchy or exponential weight.

Usage

rDynMix(nreps, x, weight)

Arguments

nreps

integer: number of observations sampled from the mixture.

x

numerical vector: if weight = 'cau', values of mu_c, \tau, \mu, \sigma, \xi, \beta; if weight = 'exp', values of \lambda, \mu, \sigma, \xi, \beta.

weight

'cau' or 'exp': name of weight distribution.

Details

This function simulates a dynamic lognormal-GPD mixture using the algorithm of Frigessi et al. (2002, p. 221).

Value

ysim (nreps x 1) vector: nreps random numbers from the lognormal-GPD dynamic mixture.

References

Frigessi A, Haug O, Rue H (2002). “A Dynamic Mixture Model for Unsupervised Tail Estimation without Threshold Selection.” Extremes, 3(5), 219–235.

Examples

ysim <- rDynMix(100,c(1,2,0,0.5,0.25,3),'cau')