Function standard_grid

rl/grid_world.py:99–123

()

def standard_grid():
  # define a grid that describes the reward for arriving at each state
  # and possible actions at each state
  # the grid looks like this
  # x means you can't go there
  # s means start position
  # number means reward at that state
  # .  .  .  1
  # .  x  . -1
  # s  .  .  .
  g = Grid(3, 4, (2, 0))
  rewards = {(0, 3): 1, (1, 3): -1}
  actions = {
    (0, 0): ('D', 'R'),
    (0, 1): ('L', 'R'),
    (0, 2): ('L', 'D', 'R'),
    (1, 0): ('U', 'D'),
    (1, 2): ('U', 'D', 'R'),
    (2, 0): ('U', 'R'),
    (2, 1): ('L', 'R'),
    (2, 2): ('L', 'R', 'U'),
    (2, 3): ('L', 'U'),
  }
  g.set(rewards, actions)
  return g


def negative_grid(step_cost=-0.1):