Bayesian Estimation for Parsimonious Threshold Autoregressive Models in R
The Newsletter of the R Project
8 Pages Posted: 27 May 2009 Last revised: 21 Aug 2009
Date Written: May 27, 2009
Abstract
BAYSTAR provides Bayesian MCMC methods for iteratively sampling to provide parameter estimates and inference for the two-regime SETAR model. A convenient user interface for importing data from a file or specifying true values for simulated data is easy to apply for analysis. Parameter inferences are summarized to an easily readable format. Simultaneously, the checking of convergence can be done by monitoring the MCMC trace plots and autocorrelograms.
Keywords: Asymmetry; MCMC method; two-regime SETAR model; BAYSTAR package
JEL Classification: c11, c15, c22, c51, c52
Suggested Citation: Suggested Citation