convert function

Converts chains from the Bayesian estimation of a multivariate TAR model to a mcmc object.

This function converts the chains obtained from the Bayesian estimation of a multivariate TAR model to a mcmc object to be analyzed with the coda package.

convert(object, type = c("location", "scale", "extra"), regime = 1)

Arguments

• object: an object of the class mtar.
• type: an (optional) character string that allows the user to specify the parameter that corresponds to the chains to convert. The available options are: "location", "scale" and "extra". As default, type is set to "location".
• regime: an (optional) integer value that allows the user to specify the regime that corresponds to the chains to plot. As default, regime is set to 1.

Returns

a mcmc-type object.

Examples

###### Example 1: Returns of the closing prices of three financial indexes
data(returns)
fit1 <- mtar(~ COLCAP + BOVESPA | SP500, row.names=Date, dist="Slash",
             data=returns, ars=list(p=c(1,1,2)), n.burnin=100, n.sim=3000)
location.chains.1 <- convert(fit1,type="location",regime=2)
summary(location.chains.1)
plot(location.chains.1)

###### Example 2: Rainfall and two river flows in Colombia
data(riverflows)
fit2 <- mtar(~ Bedon + LaPlata | Rainfall, row.names=Date, dist="Laplace",
             data=riverflows, ars=list(p=c(5,5,5)), n.burnin=100, n.sim=3000)
location.chains.2 <- convert(fit2,type="location",regime=3)
summary(location.chains.2)
plot(location.chains.2)