sqaradap {mcsm}R Documentation

Illustration of the dangers of doing adaptive MCMC on a noisy squared AR model

Description

This function constructs a non-parametric proposal after each iteration of the MCMC algorithm, based on the earlier simulations. It shows how poorly this "natural" solution fares.

Usage

sqaradap(T = 10^4, TT = 10^4, scale = 0.5, factor = 1)

Arguments

T Number of primary MCMC iterations
TT Number of further adaptive MCMC iterations
scale Scale of the normal random walk during the first T iterations
factor Factor of the bw.nrd0(xmc) bandwidth estimation

Value

The function produces two graphs showing the lack of proper fit of the resulting sample.

Author(s)

Christian P. Robert and George Casella

References

Chapter 8 of EnteR Monte Carlo Statistical Methods

See Also

sqar

Examples

sqaradap()

[Package mcsm version 1.0 Index]