R/AAA_pomdp-package.R

In pomdp: Infrastructure for Partially Observable Markov Decision Processes (POMDP)

#' @title `r packageDescription("pomdp")$Package`: `r packageDescription("pomdp")$Title`
#'
#' @description Provides the infrastructure to define and analyze the solutions of Partially Observable Markov Decision Process (POMDP) models. Interfaces for various exact and approximate solution algorithms are available including value iteration, Point-Based Value Iteration (PBVI) and Successive Approximations of the Reachable Space under Optimal Policies (SARSOP).
#'
#' @section Key functions:
#' - Problem specification: [POMDP], [MDP]
#' - Solvers: [solve_POMDP()], [solve_MDP()], [solve_SARSOP()]
#'
#' @author Michael Hahsler
#' @docType package
#' @name pomdp-package
#' 
#' @import Rcpp
#' @importFrom Matrix spMatrix crossprod coerce Math Math2 cBind rBind nnzero
#' @importFrom methods as
#' @importFrom utils head tail read.table type.convert
#' @importFrom foreach foreach times %dopar% getDoParWorkers
#' @useDynLib pomdp, .registration=TRUE
NULL

.onAttach <-function(libname, pkgname) {
  ### silence warning for no backend
  if (!foreach::getDoParRegistered())
    foreach::registerDoSEQ()
}