def betterEvaluationFunction(currentGameState):
"""
Your extreme ghost-hunting, pellet-nabbing, food-gobbling, unstoppable
evaluation function.
DESCRIPTION: <write something here so we know what you did>
"""
from searchAgents import manhattan_distance
import sys
newPos = currentGameState.getPacmanPosition()
newFood = currentGameState.getFood()
newGhostStates = currentGameState.getGhostStates()
newScaredTimes = [ghostState.scaredTimer for ghostState in newGhostStates]
# Find the nearest ghost.
ghost_distance = sys.maxint
for ghost_state in newGhostStates:
dist = manhattan_distance(ghost_state.getPosition(), newPos)
ghost_distance = min(ghost_distance, dist)
# Find the nearest food.
food_distance = sys.maxint
for x in range(newFood.width):
for y in range(newFood.height):
if newFood[x][y]:
dist = manhattan_distance(newPos, (x,y))
food_distance = min(food_distance, dist)
# Make sure we can handle situations without any ghosts or food.
if ghost_distance == sys.maxint: ghost_distance = 0
if food_distance == sys.maxint: food_distance = 0
# Start with the 'default' state score.
score = currentGameState.getScore()
# The ghost is nearby - penalize Pacman by slapping him across the face
# with negative points. Since our evaluation need not be continuous (ie.
# it's ok to have abrupt changes missing derivatives), using an if-
# statement here is a cheap way to make this evaluation function pretty
# much unbeatable. One could also imagine some kind of polynomial
# function mimicking the if-statement without breaking continuity.
if ghost_distance < 6:
score -= 100.0 / (ghost_distance+1)
# Help Pacman detect nearby food slightly.
score += 1.0 / (food_distance+1)
return score
def evaluationFunction(self, currentGameState, action):
"""
Design a better evaluation function here.
The evaluation function takes in the current and proposed successor
GameStates (pacman.py) and returns a number, where higher numbers are better.
The code below extracts some useful information from the state, like the
remaining food (newFood) and Pacman position after moving (newPos).
newScaredTimes holds the number of moves that each ghost will remain
scared because of Pacman having eaten a power pellet.
Print out these variables to see what you're getting, then combine them
to create a masterful evaluation function.
"""
# Useful information you can extract from a GameState (pacman.py)
successorGameState = currentGameState.generatePacmanSuccessor(action)
newPos = successorGameState.getPacmanPosition()
newFood = successorGameState.getFood()
newGhostStates = successorGameState.getGhostStates()
newScaredTimes = [ghostState.scaredTimer for ghostState in newGhostStates]
"""
här är skitkoden som funkar dåligt.
from searchAgents import manhattan_distance
score = successorGameState.getScore()
# gd = närmaste spöke
gd = sys.maxint
for ghost_state in newGhostStates:
gd = min(gd, manhattan_distance(ghost_state.getPosition(), newPos))
# nf = närmaste mat
nf = sys.maxint
for x in range(newFood.width):
for y in range(newFood.height):
if newFood[x][y]: nf = min(nf, manhattan_distance(newPos, (x,y)))
if gd == sys.maxint: gd = 0
if nf == sys.maxint: nf = 0
score += 10.0/(nf+1)**0.25
score += 10*gd**0.5*0.07
"""
from searchAgents import manhattan_distance
import sys
# Find the nearest ghost.
ghost_distance = sys.maxint
for ghost_state in newGhostStates:
dist = manhattan_distance(ghost_state.getPosition(), newPos)
ghost_distance = min(ghost_distance, dist)
# Find the nearest food.
food_distance = sys.maxint
for x in range(newFood.width):
for y in range(newFood.height):
if newFood[x][y]:
dist = manhattan_distance(newPos, (x,y))
food_distance = min(food_distance, dist)
# Make sure we can handle situations without any ghosts or food.
if ghost_distance == sys.maxint: ghost_distance = 0
if food_distance == sys.maxint: food_distance = 0
# Start with the 'default' state score.
score = successorGameState.getScore()
# The ghost is nearby - penalize Pacman by slapping him across the face
# with negative points. Since our evaluation need not be continuous (ie.
# it's ok to have abrupt changes missing derivatives), using an if-
# statement here is a cheap way to make this evaluation function pretty
# much unbeatable. One could also imagine some kind of polynomial
# function mimicking the if-statement without breaking continuity.
if ghost_distance < 6:
score -= 100.0 / (ghost_distance+1)
# Help Pacman detect nearby food slightly.
score += 1.0 / (food_distance+1)
return score
