Skip to contents

Generate SSVS Parameters

Source: R/simulate-param.R
sim_ssvs_var.Rd

[Deprecated] This function generates parameters of VAR with SSVS prior.

Usage

sim_ssvs_var(
  bayes_spec,
  p,
  dim_data = NULL,
  include_mean = TRUE,
  minnesota = FALSE,
  mn_prob = 1,
  method = c("eigen", "chol")
)

sim_ssvs_vhar(
  bayes_spec,
  har = c(5, 22),
  dim_data = NULL,
  include_mean = TRUE,
  minnesota = c("no", "short", "longrun"),
  mn_prob = 1,
  method = c("eigen", "chol")
)

Arguments

bayes_spec

A SSVS model specification by set_ssvs().

p

VAR lag

dim_data

Specify the dimension of the data if hyperparameters of bayes_spec have constant values.

include_mean

Add constant term (Default: TRUE) or not (FALSE)

minnesota

Only use off-diagonal terms of each coefficient matrices for restriction. In sim_ssvs_var() function, use TRUE or FALSE (default). In sim_ssvs_vhar() function, no (default), short type, or longrun type.

mn_prob

Probability for own-lags.

method

Method to compute \(\Sigma^{1/2}\).

har

Numeric vector for weekly and monthly order. By default, c(5, 22).

Value

List including coefficients.

VAR(p) with SSVS prior

Let \(\alpha\) be the vectorized coefficient of VAR(p). $$(\alpha \mid \gamma)$$ $$(\gamma_i)$$ $$(\eta_j \mid \omega_j)$$ $$(\omega_{ij})$$ $$(\psi_{ii}^2)$$

VHAR with SSVS prior

Let \(\phi\) be the vectorized coefficient of VHAR. $$(\phi \mid \gamma)$$ $$(\gamma_i)$$ $$(\eta_j \mid \omega_j)$$ $$(\omega_{ij})$$ $$(\psi_{ii}^2)$$

References

George, E. I., & McCulloch, R. E. (1993). Variable Selection via Gibbs Sampling. Journal of the American Statistical Association, 88(423), 881-889.

George, E. I., Sun, D., & Ni, S. (2008). Bayesian stochastic search for VAR model restrictions. Journal of Econometrics, 142(1), 553-580.

Ghosh, S., Khare, K., & Michailidis, G. (2018). High-Dimensional Posterior Consistency in Bayesian Vector Autoregressive Models. Journal of the American Statistical Association, 114(526).

Koop, G., & Korobilis, D. (2009). Bayesian Multivariate Time Series Methods for Empirical Macroeconomics. Foundations and Trends® in Econometrics, 3(4), 267-358.