R/goose_race.R

In mtnman38/lab: Useful functions for developing ML and AI processes

#' Build an experiment of a goose moving around a grid.
#'
#' @param x integer width of the grid
#' @param y integer height of the grid
#' @param n integer number of elements blocking path
#' @param p vector of probability of finding an obstacle and not
#' @param starting_point vector of length two, e.g., \code{c(x, y)} for where to start the goose
#' @param ending_point vector of length two, e.g., \code{c(x, y)} for where to end the race
#' @return An object of type goose with locations (a tibble), starting_point (numeric), current_location (nuemric), on_obstacle (logical)
#'
#' @export
goose_race <- function(x = 10,
                       y = 10,
                       n = 30,
                       p = c(0.1, 0.99),
                       starting_point = c(1, 1),
                       ending_point = c(10, 10)) {
  basic_grid <- tidyr::expand_grid(x = 1:x, y = 1:y)
  n_grid <- nrow(basic_grid)
  basic_grid$obstacle_present <- sample(x = c(1, 0),
                                        size = n_grid,
                                        replace = TRUE,
                                        prob = p)
  if(length(starting_point) != 2) stop("Vector length for starting point must be two")
  if(length(ending_point) != 2) stop("Vector length for ending point must be two")
  rowx <- which(basic_grid$x == starting_point[1] & basic_grid$y == starting_point[2])
  object <- structure(list(locations = basic_grid,
                           starting_point = starting_point,
                           ending_point = ending_point,
                           current_location = starting_point,
                           on_obstacle = basic_grid$obstacle_present[rowx] == 1),
                      class = "goose_race")
  return(object)
}