# baselineTeam.py
# —————
# 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 http://ai.berkeley.edu.
#
# Attribution Information: The Pacman AI projects were developed at UC Berkeley.
# The core projects and autograders were primarily created by John DeNero
# (denero@cs.berkeley.edu) and Dan Klein (klein@cs.berkeley.edu).
# Student side autograding was added by Brad Miller, Nick Hay, and
# Pieter Abbeel (pabbeel@cs.berkeley.edu).
# negin 😀
# baselineTeam.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
from captureAgents import CaptureAgent
import random, time, util, sys
from game import Directions
from util import nearestPoint
#################
# Team creation #
#################
def createTeam(firstIndex, secondIndex, isRed,
first=’TimidAgent’, second=’DefensiveReflexAgent’):
“””
This function should return a list of two agents that will form the
team, initialized using firstIndex and secondIndex as their agent
index numbers. isRed is True if the red team is being created, and
will be False if the blue team is being created.
As a potentially helpful development aid, this function can take
additional string-valued keyword arguments (“first” and “second” are
such arguments in the case of this function), which will come from
the –redOpts and –blueOpts command-line arguments to capture.py.
For the nightly contest, however, your team will be created without
any extra arguments, so you should make sure that the default
behavior is what you want for the nightly contest.
“””
return [eval(first)(firstIndex), eval(second)(secondIndex)]
# return [eval(first)(firstIndex)]
##########
# Agents #
##########
class ReflexCaptureAgent(CaptureAgent):
“””
A base class for reflex agents that chooses score-maximizing actions
“””
def registerInitialState(self, gameState):
CaptureAgent.registerInitialState(self, gameState)
def chooseAction(self, gameState):
“””
Picks among the actions with the highest Q(s,a).
“””
actions = gameState.getLegalActions(self.index)
# print gameState
# You can profile your evaluation time by uncommenting these lines
# start = time.time()
values = [self.evaluate(gameState, a) for a in actions]
# print ‘eval time for agent %d: %.4f’ % (self.index, time.time() – start)
maxValue = max(values)
bestActions = [a for a, v in zip(actions, values) if v == maxValue]
foodLeft = len(self.getFood(gameState).asList())
if foodLeft <= 2:
bestDist = 9999
for action in actions:
successor = self.getSuccessor(gameState, action)
pos2 = successor.getAgentPosition(self.index)
dist = self.getMazeDistance(self.start, pos2)
if dist < bestDist:
bestAction = action
bestDist = dist
return bestAction
return random.choice(bestActions)
def getSuccessor(self, gameState, action):
"""
Finds the next successor which is a grid position (location tuple).
"""
successor = gameState.generateSuccessor(self.index, action)
pos = successor.getAgentState(self.index).getPosition()
if pos != nearestPoint(pos):
# Only half a grid position was covered
return successor.generateSuccessor(self.index, action)
else:
return successor
def evaluate(self, gameState, action):
"""
Computes a linear combination of features and feature weights
"""
features = self.getFeatures(gameState, action)
weights = self.getWeights(gameState, action)
return features * weights
def getFeatures(self, gameState, action):
"""
Returns a counter of features for the state
"""
features = util.Counter()
successor = self.getSuccessor(gameState, action)
features['successorScore'] = self.getScore(successor)
return features
def getWeights(self, gameState, action):
"""
Normally, weights do not depend on the gamestate. They can be either
a counter or a dictionary.
"""
return {'successorScore': 1.0}
def getMostDenseArea(self, gameState):
ourFood = self.getFoodYouAreDefending(gameState).asList()
distance = [self.getMazeDistance(gameState.getAgentPosition(self.index), a) for a in ourFood]
nearestFood = ourFood[0]
nearestDstance = distance[0]
for i in range(len(distance)):
if distance[i] < nearestDstance:
nearestFood = ourFood[i]
nearestDstance = distance[i]
return nearestFood
class DefensiveReflexAgent(ReflexCaptureAgent):
lastSuccess = 0
flag = 1
flag2 = 0
currentFoods = []
s = []
"""
A reflex agent that keeps its side Pacman-free. Again,
this is to give you an idea of what a defensive agent
could be like. It is not the best or only way to make
such an agent.
"""
def getFeatures(self, gameState, action):
self.start = self.getMostDenseArea(gameState)
features = util.Counter()
successor = self.getSuccessor(gameState, action)
myState = successor.getAgentState(self.index)
myPos = myState.getPosition()
self.s = (18, 7)
# print self.s
# Computes whether we're on defense (1) or offense (0)
features['onDefense'] = 1
if myState.isPacman: features['onDefense'] = 0
features['Boundries'] = self.getMazeDistance(myPos, self.s)
if (self.flag2 == 0):
self.flag2 = 1
self.currentFoods = self.getFoodYouAreDefending(gameState).asList()
# Computes distance to invaders we can see
enemies = [successor.getAgentState(i) for i in self.getOpponents(successor)]
invaders = [a for a in enemies if a.isPacman and a.getPosition() != None]
features['numInvaders'] = len(invaders)
if len(invaders) > 0:
dists = [self.getMazeDistance(myPos, a.getPosition()) for a in invaders]
pos = [a.getPosition() for a in invaders]
nearestPos = pos[0]
nearestDst = dists[0]
for i in range(len(dists)):
if dists[i] < nearestDst:
nearestPos = pos[i]
nearestDst = dists[i]
features['invaderPDistance'] = nearestDst
# print len(self.currentFoods), len(self.getFoodYouAreDefending(gameState).asList())
if (features['invaderDistance'] == 1 or features['invaderPDistance'] == 1 or features[
'invaderLDistance'] == 1):
# print "here1"
self.flag = 0
self.lastSuccess = nearestPos
features['invaderLDistance'] = self.getMazeDistance(myPos, self.lastSuccess)
self.currentFoods = self.getFoodYouAreDefending(gameState).asList()
# print "Got Him", self.lastSuccess , self.flag
if (len(self.currentFoods) > len(self.getFoodYouAreDefending(gameState).asList())):
# print “here2”
nextFoods = self.getFoodYouAreDefending(gameState).asList()
# print “Found Him”
for i in range(len(self.currentFoods)):
# print self.currentFoods[i][0], self.currentFoods[i][1], nextFoods[i][0], nextFoods[i][1]
if (len(self.currentFoods) > 0 and len(nextFoods) > i):
# print “i: “,i,len(nextFoods),self.currentFoods[i][0],nextFoods[i][0],self.currentFoods[i][1],nextFoods[i][1]
if (self.currentFoods[i][0] != nextFoods[i][0] or self.currentFoods[i][1] != nextFoods[i][1]):
features[‘invaderPDistance’] = self.getMazeDistance(myPos, self.currentFoods[i])
# print “MYYYY”, self.currentFoods[i]
self.lastSuccess = self.currentFoods[i]
self.currentFoods = nextFoods
break
# elif(self.flag==0):
# print “here4”
# features[‘invaderDistance’]=self.getMazeDistance(myPos, self.lastSuccess)
# elif(self.flag==1):
# print “here3”
# if(self.lastSuccess==0):
# print “hii”
# self.lastSuccess=self.getMostDenseArea(gameState)
# features[‘invaderDistance’]=self.getMazeDistance(myPos, self.lastSuccess)
if action == Directions.STOP: features[‘stop’] = 1
rev = Directions.REVERSE[gameState.getAgentState(self.index).configuration.direction]
if action == rev: features[‘reverse’] = 1
return features
def getWeights(self, gameState, action):
return {‘numInvaders’: -1000, ‘onDefense’: 100, ‘invaderDistance’: -10, ‘invaderPDistance’: -20,
‘invaderLDistance’: -5, ‘Boundries’: -10, ‘stop’: -100, ‘reverse’: -2}
class TimidAgent(CaptureAgent):
“””
A Dummy agent to serve as an example of the necessary agent structure.
You should look at baselineTeam.py for more details about how to
create an agent as this is the bare minimum.
“””
def __init__(self, index, timeForComputing=.1):
CaptureAgent.__init__(self, index, timeForComputing=.1)
self.escapepath = []
self.eaten = 0
self.height = 0
self.width = 0
self.plan = [[], []]
def registerInitialState(self, gameState):
“””
This method handles the initial setup of the
agent to populate useful fields (such as what team
we’re on).
A distanceCalculator instance caches the maze distances
between each pair of positions, so your agents can use:
self.distancer.getDistance(p1, p2)
IMPORTANT: This method may run for at most 15 seconds.
“””
self.eaten = 0
self.height = len(gameState.getWalls()[0])
for w in gameState.getWalls().asList():
if w[1] == 0:
self.width += 1
print(self.height)
print(self.width)
”’
Make sure you do not delete the following line. If you would like to
use Manhattan distances instead of maze distances in order to save
on initialization time, please take a look at
CaptureAgent.registerInitialState in captureAgents.py.
”’
CaptureAgent.registerInitialState(self, gameState)
”’
Your initialization code goes here, if you need any.
”’
def chooseAction(self, gameState):
“””
Picks among actions randomly.
“””
start = time.time()
mypos = gameState.getAgentPosition(self.index)
if self.getPreviousObservation() is not None:
if self.getPreviousObservation().hasFood(mypos[0], mypos[1]):
self.eaten += 1
nearestFood = self.nearestFood(gameState)
nearestEnemy = self.getNearestEnemy(gameState)
if not gameState.getAgentState(self.index).isPacman:
self.eaten = 0
while len(self.plan[0]) == 0:
y = random.choice(range(0, self.height, 1))
if not gameState.hasWall(int(self.width / 2), y):
self.plan = [[int(self.width / 2) + 1, y],
self.bfs(gameState, self.width, self.height, nearestEnemy,
[int(self.width / 2), y])]
if len(self.plan[1]) == 0:
if not len(self.plan[0]) == 0:
self.plan[1] = self.bfs(gameState, self.width, self.height, nearestEnemy, self.plan[0])
self.escapepath = self.plan[1]
else:
self.plan = [[], []]
if len(nearestEnemy) > 0:
if nearestEnemy[1] < 4 : #and len(self.escapepath) == 0:
self.escapepath = self.bfs(gameState, self.width, self.height, nearestEnemy, [self.width / 2 - 4])
print ("RUN!!!")
else:
self.escapepath = []
if self.eaten == 5:
self.escapepath = self.bfs(gameState, 36, 17, nearestEnemy, [self.width / 2 - 1])
# self.debugDraw(self.escapepath, [1.0, 1.0, 1.0], True)
# print self.escapepath
actions = gameState.getLegalActions(self.index)
values = [self.evaluate(gameState, nearestFood, nearestEnemy, self.escapepath, a) for a in actions]
maxValue = max(values)
bestActions = [a for a, v in zip(actions, values) if v == maxValue]
action = random.choice(bestActions)
return action
def escapePath(self, game_state, width, height, enemy):
stack = util.Stack()
myState = game_state.getAgentState(self.index)
myPos = myState.getPosition()
visited = []
if len(enemy) > 0:
visited = [[enemy[0][0], enemy[0][1]]]
stack.push([myPos[0], myPos[1]])
path = []
while not stack.isEmpty():
myPos = [int(myPos[0]), int(myPos[1] + 0.5)]
psize = len(visited)
loop = []
right = [myPos[0] – 1, myPos[1]]
if right[0] >= 0 and right not in visited and not game_state.hasWall(right[0], right[1]):
stack.push(right)
visited.append(right)
if right in visited: loop.append(right)
up = [myPos[0], myPos[1] + 1]
if up[1] < height and up not in visited and not game_state.hasWall(up[0], up[1]):
stack.push(up)
visited.append(up)
if up in visited: loop.append(up)
down = [myPos[0], myPos[1] - 1]
if down[1] >= 0 and down not in visited and not game_state.hasWall(down[0], down[1]):
stack.push(down)
visited.append(down)
if down in visited: loop.append(down)
left = [myPos[0] + 1, myPos[1]]
if left[0] < width and left not in visited and not game_state.hasWall(left[0], left[1]):
stack.push(left)
visited.append(left)
if left in visited: loop.append(left)
if len(loop) > 0:
for i in reversed(path):
if abs(i[0] – myPos[0]) + abs(i[1] – myPos[1]) > 1:
path.remove(i)
else:
break
if myPos[0] == 1 and myPos[1] == 2:
# self.debugDraw(path, [1.0, 1.0, 1.0], True)
# print path
return path
myPos = stack.pop()
for i in reversed(path):
if abs(i[0] – myPos[0]) + abs(i[1] – myPos[1]) > 1:
path.remove(i)
else:
break
path.append(myPos)
return path
def nearestFood(self, gameState):
food = self.getFood(gameState).asList()
distance = [self.getMazeDistance(gameState.getAgentPosition(self.index), a) for a in food]
if len(food) < 3:
previous = self.getFoodYouAreDefending(gameState).asList()[0]
return [previous, self.getMazeDistance(gameState.getAgentPosition(self.index), previous)]
nearestFood = food[0]
nearestDstance = distance[0]
for i in range(len(distance)):
if distance[i] < nearestDstance:
nearestFood = food[i]
nearestDstance = distance[i]
return [nearestFood, nearestDstance]
def getNearestEnemy(self, gameState):
myState = gameState.getAgentState(self.index)
myPos = myState.getPosition()
# Computes distance to invaders we can see
enemies = [gameState.getAgentState(i) for i in self.getOpponents(gameState)]
invaders = [a for a in enemies if not a.isPacman and a.getPosition() != None]
dists = [self.getMazeDistance(myPos, a.getPosition()) for a in invaders]
scare = 0
if len(invaders) == 0:
return []
else:
nearestEnemy = invaders[0].getPosition()
isPacman = invaders[0].isPacman
nearestDstance = dists[0]
for i in range(len(dists)):
if dists[i] < nearestDstance:
nearestEnemy = invaders[i].getPosition()
nearestDstance = dists[i]
scare = invaders[i].scaredTimer
# print scare
# self.debugDraw(nearestEnemy, [1.0, 0.5, 0.5], True)
return [nearestEnemy, nearestDstance, scare, isPacman]
def evaluate(self, gameState, nearestFood, nearestEnemy, escapepath, action):
score = 0
scorelist = [0, 0, 0, 0, 0, 0, 0, 0, 0]
next = gameState.generateSuccessor(self.index, action)
nextpos = next.getAgentPosition(self.index)
nextscore = next.getScore()
if nextscore > gameState.getScore():
score += 2
scorelist[0] = 2
# if len(nearestEnemy) > 0 > nearestEnemy[2] and nearestEnemy[1] < 3:
# score -= 2 * (self.getMazeDistance(next.getAgentPosition(self.index), nearestEnemy[0]) - nearestEnemy[1])
# scorelist[1] -= 2 * (
# self.getMazeDistance(next.getAgentPosition(self.index), nearestEnemy[0]) - nearestEnemy[1])
# if len(nearestEnemy) == 0 or nearestEnemy[1:
if self.getMazeDistance(next.getAgentPosition(self.index), nearestFood[0]) < nearestFood[1]:
score += 1
scorelist[2] = 1
pre = self.getPreviousObservation()
if pre != None:
if self.getPreviousObservation().getAgentPosition(self.index) == nextpos:
score -= 5
scorelist[3] = -5
if len(nearestEnemy) > 0 and nearestEnemy[1] < 4:
if next.getAgentState(self.index).isPacman:
score += (self.getMazeDistance(next.getAgentPosition(self.index), nearestEnemy[0]) - nearestEnemy[1])
scorelist[4] = (
self.getMazeDistance(next.getAgentPosition(self.index), nearestEnemy[0]) - nearestEnemy[1])
nextActions = next.getLegalActions(self.index)
if len(nextActions) == 2:
score -= 100
scorelist[5] = -100
else:
score += 2
scorelist[6] = 2
if len(escapepath) > 0:
if [nextpos[0], nextpos[1]] in escapepath:
if not (len(nearestEnemy) > 0 > nearestEnemy[2] and nearestEnemy[1] < 3):
score += 10
scorelist[7] = 10
if action == Directions.STOP:
score = -10
scorelist[8] = -10
return score
def dfs(self, game_state, width, height, enemy):
stack = util.Stack()
myState = game_state.getAgentState(self.index)
myPos = myState.getPosition()
visited = []
if len(enemy) > 0:
visited = [[enemy[0][0], enemy[0][1]]]
stack.push([myPos[0], myPos[1]])
path = []
while not stack.isEmpty():
myPos = [int(myPos[0]), int(myPos[1] + 0.5)]
psize = len(visited)
loop = []
right = [myPos[0] – 1, myPos[1]]
if right[0] >= 0 and right not in visited and not game_state.hasWall(right[0], right[1]):
stack.push(right)
visited.append(right)
if right in visited: loop.append(right)
up = [myPos[0], myPos[1] + 1]
if up[1] < height and up not in visited and not game_state.hasWall(up[0], up[1]):
stack.push(up)
visited.append(up)
if up in visited: loop.append(up)
down = [myPos[0], myPos[1] - 1]
if down[1] >= 0 and down not in visited and not game_state.hasWall(down[0], down[1]):
stack.push(down)
visited.append(down)
if down in visited: loop.append(down)
left = [myPos[0] + 1, myPos[1]]
if left[0] < width and left not in visited and not game_state.hasWall(left[0], left[1]):
stack.push(left)
visited.append(left)
if left in visited: loop.append(left)
if len(loop) > 0:
for i in reversed(path):
if abs(i[0] – myPos[0]) + abs(i[1] – myPos[1]) > 1:
path.remove(i)
else:
break
if myPos[0] == 1 and myPos[1] == 2:
# self.debugDraw(path, [1.0, 1.0, 1.0], True)
# print path
return path
myPos = stack.pop()
for i in reversed(path):
if abs(i[0] – myPos[0]) + abs(i[1] – myPos[1]) > 1:
path.remove(i)
else:
break
path.append(myPos)
return path
def bfs(self, game_state, width, height, enemy, point):
queue = util.Queue()
myState = game_state.getAgentState(self.index)
myPos = myState.getPosition()
visited = []
if len(enemy) > 0:
visited = [[enemy[0][0], enemy[0][1]]]
queue.push([myPos[0], myPos[1]])
path = []
while not queue.isEmpty():
myPos = [int(myPos[0]), int(myPos[1] + 0.5)]
psize = len(visited)
loop = []
right = [myPos[0] – 1, myPos[1]]
if right[0] >= 0 and right not in visited and not game_state.hasWall(right[0], right[1]):
queue.push(right)
visited.append(right)
if right in visited: loop.append(right)
up = [myPos[0], myPos[1] + 1]
if up[1] < height and up not in visited and not game_state.hasWall(up[0], up[1]):
queue.push(up)
visited.append(up)
if up in visited: loop.append(up)
down = [myPos[0], myPos[1] - 1]
if down[1] >= 0 and down not in visited and not game_state.hasWall(down[0], down[1]):
queue.push(down)
visited.append(down)
if down in visited: loop.append(down)
left = [myPos[0] + 1, myPos[1]]
if left[0] < width and left not in visited and not game_state.hasWall(left[0], left[1]):
queue.push(left)
visited.append(left)
if left in visited: loop.append(left)
# if len(loop) > 0:
# for i in reversed(path):
# if abs(i[0] – myPos[0]) + abs(i[1] – myPos[1]) > 1:
# path.remove(i)
# else:
# break
myPos = queue.pop()
path.append(myPos)
# print path
if len(point) == 1:
if myPos[0] == point[0]:
a = myPos
f = []
for i in reversed(path):
if abs(a[0] – i[0]) + abs(a[1] – i[1]) <= 1:
f.append(i)
a = i
# self.debugDraw(f, [1.0, 1.0, 1.0], True)
return f
else:
if myPos[0] == point[0] and myPos[1] == point[1]:
a = myPos
f = []
for i in reversed(path):
if abs(a[0] - i[0]) + abs(a[1] - i[1]) <= 1:
f.append(i)
a = i
# self.debugDraw(f, [1.0, 1.0, 1.0], True)
return f
return []