import random
import sys
import mdp
import environment
import util
import optparse
class Gridworld(mdp.MarkovDecisionProcess):
  """
    Gridworld
  """
  def __init__(self, grid):
    # layout
    if type(grid) == type([]): grid = makeGrid(grid)
    self.grid = grid
    
    # parameters
    self.livingReward = 0.0
    self.noise = 0.2
        
  def setLivingReward(self, reward):
    """
    The (negative) reward for exiting "normal" states.
    
    Note that in the R+N text, this reward is on entering
    a state and therefore is not clearly part of the state's
    future rewards.
    """
    self.livingReward = reward
        
  def setNoise(self, noise):
    """
    The probability of moving in an unintended direction.
    """
    self.noise = noise

[code python][/code]