Skip to content

Instantly share code, notes, and snippets.

@rioleo
Created November 8, 2011 02:13
Show Gist options
  • Save rioleo/1346816 to your computer and use it in GitHub Desktop.
Save rioleo/1346816 to your computer and use it in GitHub Desktop.
eulav-Iterations-Tnega
# valueIterationAgents.py
# -----------------------
# Licensing Information: Please do not distribute or publish solutions to this
# project. You are free to use and extend these projects for educational
# purposes. The Pacman AI projects were developed at UC Berkeley, primarily by
# John DeNero (denero@cs.berkeley.edu) and Dan Klein (klein@cs.berkeley.edu).
# For more info, see http://inst.eecs.berkeley.edu/~cs188/sp09/pacman.html
import mdp, util
from learningAgents import ValueEstimationAgent
class ValueIterationAgent(ValueEstimationAgent):
"""
* Please read learningAgents.py before reading this.*
A ValueIterationAgent takes a Markov decision process
(see mdp.py) on initialization and runs value iteration
for a given number of iterations using the supplied
discount factor.
"""
def __init__(self, mdp, discount = 0.9, iterations = 100):
"""
Your value iteration agent should take an mdp on
construction, run the indicated number of iterations
and then act according to the resulting policy.
Some useful mdp methods you will use:
mdp.getStates()
mdp.getPossibleActions(state)
mdp.getTransitionStatesAndProbs(state, action)
mdp.getReward(state, action, nextState)
"""
self.mdp = mdp
self.discount = discount
self.iterations = iterations
self.values = util.Counter() # A Counter is a dict with default 0
#
for iteration in range(iterations):
for state in mdp.getStates():
maxUtility = 0
maxReward = 0
for action in mdp.getPossibleActions(state):
tempUtility = 0
for i in mdp.getTransitionStatesAndProbs(state, action):
nextstate = i[0]
nextprob = i[1]
nextreward = mdp.getReward(state, action, nextstate)
tempUtility += nextprob*self.values[nextstate]
if tempUtility >= maxUtility:
maxUtility = tempUtility
maxReward = nextreward
self.values[state] = maxReward + maxUtility*discount
def getValue(self, state):
"""
Return the value of the state (computed in __init__).
"""
return self.values[state]
def getQValue(self, state, action):
"""
The q-value of the state action pair
(after the indicated number of value iteration
passes). Note that value iteration does not
necessarily create this quantity and you may have
to derive it on the fly.
"""
maxReward = 0
for stateProb in self.mdp.getTransitionStatesAndProbs(state, action):
nextstate = stateProb[0]
nextprob = stateProb[1]
tempReward = nextprob * self.values[nextstate]
if tempReward >= maxReward:
maxReward = tempReward
return maxReward
#*self.values(state)
def getPolicy(self, state):
"""
The policy is the best action in the given state
according to the values computed by value iteration.
You may break ties any way you see fit. Note that if
there are no legal actions, which is the case at the
terminal state, you should return None.
"""
if not self.mdp.getPossibleActions(state):
return None
maxi = 0
bestaction = ""
#print "States", self.mdp.getPossibleActions((0,0))
for a in self.mdp.getPossibleActions(state):
#print "x", a
r = self.getQValue(state, a)
if r > maxi:
maxi = r
bestaction = a
print "Best action", bestaction
return bestaction
#return None
#util.raiseNotDefined()
def getAction(self, state):
"Returns the policy at the state (no exploration)."
return self.getPolicy(state)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment