RJMCMC move for AR order selection of mixture autoregressive models

Computes probabilities for deciding whether the AR order should be increased or decreased by 1 at each iteration in Bayesian analysis of mixture autoregressive models.

Usage

bx_dx(method = c("Ratio", "Poisson", "NULL"), par, pk)

Arguments

method

the method used for updating probabilities. If "Ratio" or "Poisson", probabilities depend on par and decrease as pk increases. If "NULL", probabilities are set to 0.5. See Examples for details.

par

tuning parameter for calculating updating probabilities.

pk

autoregressive order of the selected component.

Value

A list of 2 elements:

bx

The probability of increasing the autoregressive order by 1.

dx

The probability of decreasing the autoregressive order by 1, calculated as 1 - bx.

References

Ravagli D, Boshnakov GN (2020). “Bayesian analysis of mixture autoregressive models covering the complete parameter space.” 2006.11041, https://arxiv.org/abs/2006.11041.

Author

Davide Ravagli

Note

This function is for use within Choose_pk.