R/tiles.R

In reinforcelearn: Reinforcement Learning

#' Tile Coding
#'
#' Implementation of Sutton's tile coding software version 3.
#'
#' @param iht \[`IHT`] \cr A hash table created with `iht`.
#' @param n.tilings \[`integer(1)`] \cr Number of tilings.
#' @param state \[`vector(2)`] \cr A two-dimensional state observation.
#'   Make sure to scale the observation to unit variance before.
#' @param action \[`integer(1)`] \cr Optional: If supplied the action space
#'   will also be tiled. All distinct actions will result in different tile numbers.
#'
#' @return `iht` creates a hash table, which can then be passed on to `tiles`.
#'   `tiles` returns an integer vector of size `n.tilings` with the active tile numbers.
#'
#' @md
#'
#' @details
#' Tile coding is a way of representing the values of a vector of continuous variables as a large
#' binary vector with few 1s and many 0s. The binary vector is not represented explicitly,
#' but as a list of the components that are 1s. The main step is to partition, or tile,
#' the continuous space multiple times and select one tile from each tiling, that corresponding
#' the the vector's value. Each tile is converted to an element in the big binary vector,
#' and the list of the tile (element) numbers is returned as the representation of the vector's value.
#' Tile coding is recommended as a way of applying online learning methods to domains with continuous
#' state or action variables. \[copied from manual]
#'
#' See detailed manual on the web.
#' In comparison to the Python implementation indices start with 1 instead of 0. The hash table is
#' implemented as an environment, which is an attribute of an R6 class.
#'
#' Make sure that the size of the hash table is large enough, else an error will be triggered,
#' when trying to assign a value to a full hash table.
#'
#' @references Sutton and Barto (Book draft 2017): Reinforcement Learning: An Introduction
#' @rdname tilecoding
#' @export
#' @examples
#' # Create hash table
#' hash = iht(1024)
#'
#' # Partition state space using 8 tilings
#' tiles(hash, n.tilings = 8, state = c(3.6, 7.21))
#' tiles(hash, n.tilings = 8, state = c(3.7, 7.21))
#' tiles(hash, n.tilings = 8, state = c(4, 7))
#' tiles(hash, n.tilings = 8, state = c(- 37.2, 7))
#'
tiles = function(iht, n.tilings, state, action = integer(0)) {
  checkmate::assertClass(iht, "IHT")
  checkmate::assertInt(n.tilings)
  checkmate::assertVector(state)
  checkmate::assertIntegerish(action, max.len = 1)

  qfloats = floor(state * n.tilings)
  active.tiles = rep(0, n.tilings)
  coords = rep(0, length(state) + 1)

  for (tiling in seq_len(n.tilings)) {
    tiling = tiling - 1
    tiling2 = tiling * 2
    coords[1] = tiling
    b = tiling
    for (q in seq_along(qfloats)) {
      coords[q + 1] = (qfloats[q] + b) %/% n.tilings
      b = b + tiling2
    }
    coords = append(coords, action)
    active.tiles[tiling + 1] = hashcoords(paste(coords, collapse = ""), iht)
  }

  return(active.tiles)
}

hashcoords = function(coords, iht) {
  iht$add2Env(coords)
  iht$checkFull()
  iht$getIndex(coords)
}

#' @rdname tilecoding
#' @param max.size \[`integer(1)`] \cr Maximal size of hash table.
#' @export
#' @md
iht = function(max.size) {
  checkmate::assertInt(max.size)
  IHTClass$new(max.size)
}

IHTClass = R6::R6Class("IHT",
  public = list(
    i = 0,
    max.size = NULL,
    e = NULL,

    initialize = function(max.size) {
      self$max.size = max.size
      self$e = new.env(size = max.size)
    },

    checkFull = function() {
      if (length(self$e) > self$max.size) {
        stop("Tile Coding failed because hash table IHT is full!")
      }
    },

    add2Env = function(coords) {
      if (!exists(coords, envir = self$e, inherits = FALSE)) {
        self$i = self$i + 1
        self$checkFull()
        self$e[[coords]] = self$i
      }
    },

    getIndex = function(coords) {
      return(self$e[[coords]])
    }
  )
)

#' Make n hot vector.
#'
#' @param x \[`integer`] \cr Which features are active?
#' @param len \[`integer(1)`] \cr Length of the feature vector.
#' @param out \[`character(1)`] \cr Format of the output. Can be a vector or a matrix.
#'
#' @return \[`matrix(1, len)`] A one-row matrix with `len` columns with every
#'   entry 0 except the columns specified by `x` which are 1.
#'
#' @md
#'
#' @export
#' @examples
#' nHot(c(1, 3), 5)
#' nHot(c(1, 3), 5, out = "vector")
nHot = function(x, len, out = "matrix") {
  checkmate::assertIntegerish(x, max.len = len)
  checkmate::assertInt(len)
  if (out == "matrix") {
    m = matrix(rep(0, len), nrow = 1)
    m[1, x] = 1
  } else {
    m = rep(0, len)
    m[x] = 1
  }
  m
}