forecast method for PosteriorBVARPANEL #5

R/forecast.R

@@ -0,0 +1,67 @@
+#' @title Forecasting using Hierarchical Pannel Vector Autoregressions
+#'
+#' @description Samples from the joint predictive density of the dependent 
+#' variables for all countries at forecast horizons 
+#' from 1 to \code{horizon} specified as an argument of the function.
+#' 
+#' @method forecast PosteriorBVARPANEL
+#' 
+#' @param posterior posterior estimation outcome - an object of class 
+#' PosteriorBVARPANEL obtained by running the \code{estimate} function.
+#' @param horizon a positive integer, specifying the forecasting horizon.
+#' @param exogenous_forecast not used here ATM; included for compatibility with 
+#' generic \code{forecast}.
+#' 
+#' @return A list of class \code{PanelForecasts} containing the
+#' draws from the predictive density and data. The output list includes element:
+#' 
+#' \describe{
+#'  \item{forecasts}{an \code{horizonxNxCxS} array with the draws from predictive density}
+#'  \item{forecasts_cpp}{an unspecified object passed for computations in **cpp**}
+#'  \item{Y}{a \code{C}-element list with \code{T_cxN} matrices with the country-specific data}
+#' }
+#'
+#' @author Tomasz Woźniak \email{wozniak.tom@pm.me}
+#' 
+#' @examples
+#' data(ilo_cubic_panel)                                   # load the data
+#' set.seed(123)
+#' specification = specify_bvarPANEL$new(ilo_cubic_panel)  # specify the model
+#' burn_in       = estimate(specification, 10)             # run the burn-in
+#' posterior     = estimate(burn_in, 10)                   # estimate the model
+#' predictive    = forecast(posterior, 2)                  # forecast 2 years ahead
+#' 
+#' # workflow with the pipe |>
+#' ############################################################
+#' set.seed(123)
+#' ilo_cubic_panel |>
+#'   specify_bvarPANEL$new() |>
+#'   estimate(S = 10) |> 
+#'   estimate(S = 20) |> 
+#'   forecast(horizon = 2) -> predictive
+#' 
+#' @export
+forecast.PosteriorBVARPANEL = function(posterior, horizon = 1, exogenous_forecast = NULL) {
+
+
+  posterior_A_c_cpp     = posterior$posterior$A_c_cpp
+  posterior_Sigma_c_cpp = posterior$posterior$Sigma_c_cpp
+  X_c             = posterior$last_draw$data_matrices$X
+  Y_c             = posterior$last_draw$data_matrices$Y
+
+  fore            = .Call(`_bvarPANELs_forecast_bvarPANEL`, posterior_A_c_cpp, posterior_Sigma_c_cpp, X_c, horizon)
+
+  N               = dim(Y_c[[1]])[2]
+  S               = dim(posterior_A_c_cpp)[1]
+  C               = length(Y_c)
+  forecasts      = array(NA, c(horizon, N, S, C))
+  for (c in 1:C) {
+    forecasts[,,,c] = fore$forecasts_cpp[c,1][[1]]
+  }
+  forecasts       = aperm(forecasts, c(1,2,4,3))
+  fore$forecasts  = forecasts
+  fore$Y          = Y_c
+  class(fore)     = "PanelForecasts"
+
+  return(fore)
+}