Skip to contents

R6 Class representing the specification of the BVARGROUPPRIORPANEL model

Source: R/specify_bvargrouppriorpanel.R
specify_bvarGroupPriorPANEL.Rd

The class BVARGROUPPRIORPANEL presents complete specification for the Bayesian Panel Vector Autoregression with county grouping for global prior parameters. The groups can be pre-specified, which requires the argument group_allocation to be provided, or estimated, which requires the argument G for the number of groups to be provided and the argument group_allocation to be left empty.

References

Zellner, Hong (1989). Forecasting international growth rates using Bayesian shrinkage and other procedures. Journal of Econometrics, 40(1), 183–202, doi:10.1016/0304-4076(89)90036-5 .

Public fields

p

a non-negative integer specifying the autoregressive lag order of the model.

G

a non-negative integer specifying the number of country groupings.

estimate_groups

a logical value denoting whether the groups are to be estimated.

prior

an object PriorBSVAR with the prior specification.

data_matrices

an object DataMatricesBVARPANEL with the data matrices.

starting_values

an object StartingValuesBVARGROUPPRIORPANEL with the starting values.

adaptiveMH

a vector of four values setting the adaptive MH sampler for nu: adaptive rate, target acceptance rate, the iteration at which to start adapting, the initial scaling rate

Methods

Public methods

Method new()

Create a new specification of the Bayesian Panel VAR model with country grouping for global prior parameters BVARGROUPPRIORPANEL. The groups can be pre-specified, which requires the argument group_allocation to be provided, or estimated, which requires the argument G for the number of groups to be provided and the argument group_allocation to be left empty.

Usage

specify_bvarGroupPriorPANEL$new(
  data,
  p = 1L,
  exogenous = NULL,
  stationary = rep(FALSE, ncol(data[[1]])),
  type = rep("real", ncol(data[[1]])),
  G = NULL,
  group_allocation = NULL
)

Arguments

data

a list with C elements of (T_c+p)xN matrices with time series data.

p

a positive integer providing model's autoregressive lag order.

exogenous

a (T+p)xd matrix of exogenous variables.

stationary

an N logical vector - its element set to FALSE sets the prior mean for the autoregressive parameters of the Nth equation to the white noise process, otherwise to random walk.

type

an N character vector with elements set to "rate" or "real" determining the truncation of the predictive density to [0, 100] and (-Inf, Inf) (no truncation) for each of the variables.

G

a positive integer specifying the number of country groups. Its specification is required if group_allocation is not provided and the country groups to be estimated.

group_allocation

an argument that can be provided as a numeric vector with integer numbers denoting group allocations to pre-specify the the country groups, in which case they are not estimated, or left empty if the country groups are to be estimated.

Returns

A new complete specification for the Bayesian Panel VAR model BVARPANEL.

Method get_data_matrices()

Returns the data matrices as the DataMatricesBVARPANEL object.

Usage

specify_bvarGroupPriorPANEL$get_data_matrices()

Examples

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$get_data_matrices()

Method get_prior()

Returns the prior specification as the PriorBVARPANEL object.

Usage

specify_bvarGroupPriorPANEL$get_prior()

Examples

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$get_prior()

Method get_starting_values()

Returns the starting values as the StartingValuesBVARPANEL object.

Usage

specify_bvarGroupPriorPANEL$get_starting_values()

Examples

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$get_starting_values()

Method get_type()

Returns the type of the model.

Usage

specify_bvarGroupPriorPANEL$get_type()

Examples

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$get_type()

Method set_global2pooled()

Sets the prior mean of the global autoregressive parameters to the OLS pooled panel estimator following Zellner, Hong (1989).

Usage

specify_bvarGroupPriorPANEL$set_global2pooled(x)

Arguments

x

a vector of four values setting the adaptive MH sampler for nu: adaptive rate, target acceptance rate, the iteration at which to start adapting, the initial scaling rate

Examples

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$set_global2pooled()

Method set_adaptiveMH()

Sets the parameters of adaptive Metropolis-Hastings sampler for the parameter nu.

Usage

specify_bvarGroupPriorPANEL$set_adaptiveMH(x)

Arguments

x

a vector of four values setting the adaptive MH sampler for nu: adaptive rate, target acceptance rate, the iteration at which to start adapting, the initial scaling rate

Examples

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$set_adaptiveMH(c(0.6, 0.4, 10, 0.1))

Method clone()

The objects of this class are cloneable with this method.

Usage

specify_bvarGroupPriorPANEL$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

data(ilo_dynamic_panel)
spec = specify_bvarGroupPriorPANEL$new(
   data = ilo_dynamic_panel,
   p = 4,
   G = 2
)
#> Country groupings will be estimated.


## ------------------------------------------------
## Method `specify_bvarGroupPriorPANEL$get_data_matrices`
## ------------------------------------------------

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$get_data_matrices()
#> <DataMatricesBVARPANEL>
#>   Public:
#>     X: list
#>     Y: list
#>     clone: function (deep = FALSE) 
#>     get_data_matrices: function () 
#>     initialize: function (data, p = 1L, exogenous = NULL, type = rep("real", 
#>     type: real real real real


## ------------------------------------------------
## Method `specify_bvarGroupPriorPANEL$get_prior`
## ------------------------------------------------

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$get_prior()
#> <PriorBVARPANEL>
#>   Public:
#>     M: 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0  ...
#>     S_Sigma_inv: 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1
#>     S_inv: 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1
#>     W: 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0  ...
#>     a_w: 1
#>     clone: function (deep = FALSE) 
#>     eta: 5
#>     get_prior: function () 
#>     initialize: function (C, N, p, d = 0, stationary = rep(FALSE, N)) 
#>     lambda: 72
#>     mu_Sigma: 5
#>     mu_m: 1
#>     nu_s: 3
#>     s_s: 1
#>     s_w: 1
#>     sigma2_m: 1


## ------------------------------------------------
## Method `specify_bvarGroupPriorPANEL$get_starting_values`
## ------------------------------------------------

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$get_starting_values()
#> <StartingValuesBVARPANEL>
#>   Public:
#>     A: 0.998658549704844 -0.000719659043776375 0.00047824296095 ...
#>     A_c: -0.000646674632486275 0.00266780731315677 -0.00045436796 ...
#>     Sigma: 6.10134500198224 2.36586229815875 0.121378355577619 -1.2 ...
#>     Sigma_c: 7.36836732265444 -0.0500727537113589 -3.76010557288262 0 ...
#>     V: 12.2250943216527 -6.83704689315007 -2.05237337702113 -2. ...
#>     clone: function (deep = FALSE) 
#>     get_starting_values: function () 
#>     initialize: function (C, N, p, d = 0) 
#>     m: -0.00179692181464839
#>     nu: 5.1
#>     s: 0.307035920842831
#>     set_starting_values: function (last_draw) 
#>     w: 2.28017548314471


## ------------------------------------------------
## Method `specify_bvarGroupPriorPANEL$get_type`
## ------------------------------------------------

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$get_type()
#> [1] "wozniak"


## ------------------------------------------------
## Method `specify_bvarGroupPriorPANEL$set_global2pooled`
## ------------------------------------------------

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$set_global2pooled()


## ------------------------------------------------
## Method `specify_bvarGroupPriorPANEL$set_adaptiveMH`
## ------------------------------------------------

data(ilo_dynamic_panel)
spec = specify_bvarPANEL$new(
   data = ilo_dynamic_panel,
   p = 4
)
spec$set_adaptiveMH(c(0.6, 0.4, 10, 0.1))