Skip to content
Branch: master
Find file Copy path
Find file Copy path
mkapnick Initial commit 99d613b Nov 20, 2013
0 contributors

Users who have contributed to this file

131 lines (106 sloc) 4.53 KB
# -----------------------
# Licensing Information: You are free to use or extend these projects for
# educational purposes provided that (1) you do not distribute or publish
# solutions, (2) you retain this notice, and (3) you provide clear
# attribution to UC Berkeley, including a link to
# Attribution Information: The Pacman AI projects were developed at UC Berkeley.
# The core projects and autograders were primarily created by John DeNero
# ( and Dan Klein (
# Student side autograding was added by Brad Miller, Nick Hay, and
# Pieter Abbeel (
import mdp, util
from learningAgents import ValueEstimationAgent
class ValueIterationAgent(ValueEstimationAgent):
* Please read before reading this.*
A ValueIterationAgent takes a Markov decision process
(see 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.getTransitionStatesAndProbs(state, action)
mdp.getReward(state, action, nextState)
self.mdp = mdp = discount
self.iterations = iterations
self.values = util.Counter() # A Counter is a dict with default 0
# Write value iteration code here
#print self.values
state = self.mdp.getStates()[2]
#print mdp.getPossibleActions(state)
nextState = mdp.getTransitionStatesAndProbs(state, mdp.getPossibleActions(state)[0])
#print nextState
#print "printed next state"
#print mdp.getReward(state, mdp.getPossibleActions(state)[0] ,nextState)
states = self.mdp.getStates()
#print self.mdp.getStartState()
for i in range(iterations):
valuesCopy = self.values.copy()
for state in states:
finalValue = None
for action in self.mdp.getPossibleActions(state):
currentValue = self.computeQValueFromValues(state,action)
if finalValue == None or finalValue < currentValue:
finalValue = currentValue
if finalValue == None:
finalValue = 0
valuesCopy[state] = finalValue
self.values = valuesCopy
def getValue(self, state):
Return the value of the state (computed in __init__).
return self.values[state]
def computeQValueFromValues(self, state, action):
Compute the Q-value of action in state from the
value function stored in self.values.
"*** YOUR CODE HERE ***"
value = 0
transitionFunction = self.mdp.getTransitionStatesAndProbs(state,action)
for nextState, probability in transitionFunction:
value += probability * (self.mdp.getReward(state, action, nextState)
+ ( * self.values[nextState]))
return value
def computeActionFromValues(self, state):
The policy is the best action in the given state
according to the values currently stored in self.values.
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.
"*** YOUR CODE HERE ***"
possibleActions = self.mdp.getPossibleActions(state)
if len(possibleActions) == 0:
return None
value = None
result = None
for action in possibleActions:
temp = self.computeQValueFromValues(state, action)
if value == None or temp > value:
value = temp
result = action
return result
def getPolicy(self, state):
return self.computeActionFromValues(state)
def getAction(self, state):
"Returns the policy at the state (no exploration)."
return self.computeActionFromValues(state)
def getQValue(self, state, action):
return self.computeQValueFromValues(state, action)
You can’t perform that action at this time.