BVAR: bv_priors – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

bv_priors

Prior settings

Description

Function to provide priors and their parameters to bvar. Used for adjusting the parameters treated as hyperparameters, the Minnesota prior and adding various dummy priors through the ellipsis parameter. Note that treating psi (psi) as a hyperparameter in a model with many variables may lead to very low acceptance rates and thus hinder convergence.

Usage

bv_priors(hyper = "auto", mn = bv_mn(), ...)

Arguments

`hyper`	Character vector. Used to specify the parameters to be treated as hyperparameters. May also be set to `"auto"` or `"full"` for an automatic / full subset. Other allowed values are the Minnesota prior's parameters `"lambda"`, `"alpha"` and `"psi"` as well as the names of additional dummy priors included via ....
`mn`	List of class `"bv_minnesota"`. Options for the Minnesota prior, set via `bv_mn`.
`...`	Optional lists of class `bv_dummy` with options for dummy priors. Must be assigned a name in the function call. Created with `bv_dummy`.

Value

Returns a named list of class bv_priors with options for bvar.

Examples

# Extend the hyperparameters to the full Minnesota prior
bv_priors(hyper = c("lambda", "alpha", "psi"))
# Alternatively
# bv_priors("full")

# Add a dummy prior via `bv_dummy()`

# Re-create the single-unit-root prior
add_sur <- function(Y, lags, par) {
  sur <- if(lags == 1) {Y[1, ] / par} else {
    colMeans(Y[1:lags, ]) / par
  }
  Y_sur <- sur
  X_sur <- c(1 / par, rep(sur, lags))

  return(list("Y" = Y_sur, "X" = X_sur))
}
sur <- bv_dummy(mode = 1, sd = 1, min = 0.0001, max = 50, fun = add_sur)

# Add the new prior
bv_priors(hyper = "auto", sur = sur)

BVAR

Hierarchical Bayesian Vector Autoregression

v1.0.1

GPL-3 | file LICENSE

Authors

Nikolas Kuschnig [aut, cre] (<https://orcid.org/0000-0002-6642-2543>), Lukas Vashold [aut] (<https://orcid.org/0000-0002-3562-3414>), Michael McCracken [dtc], Serena Ng [dtc]

Initial release

2020-09-26