def __call__(self, individual_0):
"""
Function call operator. Starts a Tron game and sets the
strategies of the two players according to the Individual
function arguments. The fitness of the winner increases by 1.
:param individual_0: Individual solution
:type individual_0: Individual
"""
print(self.__class__.__name__, '.__call__', individual_0.fitness)
strategy_0 = self.decode_individual(individual_0)
for game in range(self.game_repetitions):
#Create Tron game instance
#Pass in the AI player
tron_game = TronAdversarial(rows=self.rows,
bike_width=self.bike_width,
draw_board=self.draw_board,
strategy=strategy_0)
tron_game.run()
winner = tron_game.winner
if winner[0] == 1:
individual_0.fitness += 1
print(self.__class__.__name__, ".__call__ winner", winner,
individual_0.fitness)
def __call__(self, individual_0):
"""
Function call operator. Starts a Tron game and sets the
strategies of the two players according to the Individual
function arguments. The fitness of the winner increases by 1.
:param individual_0: Individual solution
:type individual_0: Individual
"""
print(self.__class__.__name__, '.__call__', individual_0.fitness)
strategy_0 = self.decode_individual(individual_0)
for game in range(self.game_repetitions):
#Create Tron game instance
#Pass in the AI player
tron_game = TronAdversarial(rows=self.rows,
bike_width=self.bike_width,
draw_board=self.draw_board,
strategy=strategy_0)
tron_game.run()
winner = tron_game.winner
if winner[0] == 1:
individual_0.fitness += 1
print(self.__class__.__name__, ".__call__ winner", winner,
individual_0.fitness)
def __call__(self, individual_0):
"""
Function call operator. Starts a Tron game and sets the
strategy of the player according to the Individual
function arguments. The fitness is the length of the tail.
:param individual_0: Individual solution
:type individual_0: Individual
"""
# The strategy is the executable GP-tree
strategy = individual_0.genome
# Play a number of games
for game in range(self.game_repetitions):
# Create Tron game instance
tron_game = TronAdversarial(rows=self.rows,
bike_width=self.bike_width,
draw_board=self.draw_board,
strategy=strategy)
# Create an AI GP player
tron_game.players[1] = \
PlayerAIGP(x=tron_game.players[1].x,
y=tron_game.players[1].y,
direction=tron_game.players[1].direction,
color=tron_game.players[1].color,
alive=tron_game.players[1].alive,
id_=tron_game.players[1].ID,
canvas=tron_game.players[1].canvas,
board=tron_game.players[1].board,
strategy=tron_game.players[1].strategy,
# Set the strategy
# evaluation function in
# the tron player
evaluator=self.evaluate_strategy)
# Run the tron game
tron_game.run()
# Get the winner of the game
winner = tron_game.winner
# Get the ID of the winner
if winner[0] == 1:
# Increase the fitness of the AI if it is the winner
individual_0.fitness += 1
print(self.__class__.__name__, ".__call__ winner", winner,
individual_0.fitness)
def __call__(self, individual_0):
"""
Function call operator. Starts a Tron game and sets the
strategy of the player according to the Individual
function arguments. The fitness is the length of the tail.
:param individual_0: Individual solution
:type individual_0: Individual
"""
# The strategy is the executable GP-tree
strategy = individual_0.genome
# Play a number of games
for game in range(self.game_repetitions):
# Create Tron game instance
tron_game = TronAdversarial(rows=self.rows,
bike_width=self.bike_width,
draw_board=self.draw_board,
strategy=strategy)
# Create an AI GP player
tron_game.players[1] = \
PlayerAIGP(x=tron_game.players[1].x,
y=tron_game.players[1].y,
direction=tron_game.players[1].direction,
color=tron_game.players[1].color,
alive=tron_game.players[1].alive,
id_=tron_game.players[1].ID,
canvas=tron_game.players[1].canvas,
board=tron_game.players[1].board,
strategy=tron_game.players[1].strategy,
# Set the strategy
# evaluation function in
# the tron player
evaluator=self.evaluate_strategy)
# Run the tron game
tron_game.run()
# Get the winner of the game
winner = tron_game.winner
# Get the ID of the winner
if winner[0] == 1:
# Increase the fitness of the AI if it is the winner
individual_0.fitness += 1
print(self.__class__.__name__, ".__call__ winner", winner,
individual_0.fitness)
def __call__(self, individual_0, individual_1):
"""
Function call operator. Starts a Tron game and sets the
strategies of the two players according to the Individual
function arguments. The fitness of the winner increases by 1.
:param individual_0: Individual solution
:type individual_0: Individual
:param individual_1: Individual solution
:type individual_1: Individual
"""
# Decode the strategies
print(self.__class__.__name__, '.__call__', individual_0.fitness,
individual_1.fitness)
strategy_0 = self.decode_individual(individual_0)
strategy_1 = self.decode_individual(individual_1)
# TODO for 2 ai players the outcome for each repetion will be the same
# since there is no stochastic element
for game in range(self.game_repetitions):
#Create Tron game instance
tron_game = TronAdversarial(rows=self.rows,
bike_width=self.bike_width,
draw_board=self.draw_board,
strategy=None)
#Set players in Tron
tron_game.players[0] = PlayerAI(x=self.rows / 2,
y=self.rows / 2,
direction=(0, 1),
color="Blue",
alive=True,
id_=0,
canvas=tron_game.canvas,
board=tron_game.board,
strategy=strategy_0)
tron_game.players[1] = PlayerAI(x=self.rows / 2 + 1,
y=self.rows / 2 + 1,
direction=(0, 1),
color="Green",
alive=True,
id_=1,
canvas=tron_game.canvas,
board=tron_game.board,
strategy=strategy_1)
# Run the Tron game
tron_game.run()
# Get the winner
winner = tron_game.winner
# Increase the fitness of the winner
if winner[0] == 0:
individual_0.fitness += 1
elif winner[0] is not None:
individual_1.fitness += 1
print(self.__class__.__name__, ".__call__ winner", winner,
individual_0.fitness, individual_1.fitness)
def __call__(self, individual_0, individual_1):
"""
Function call operator. Starts a Tron game and sets the
strategy of the player according to the Individual
function arguments. The fitness is the length of the tail.
:param individual_0: Individual solution
:type individual_0: Individual
:param individual_1: Individual solution
:type individual_1: Individual
"""
# Set the strategies as the GP tree of the individuals
self.strategies.append(individual_0.genome)
self.strategies.append(individual_1.genome)
# Play a number of games
for game in range(self.game_repetitions):
# Create Tron game instance
tron_game = TronAdversarial(rows=self.rows,
bike_width=self.bike_width,
draw_board=self.draw_board,
strategy=self.strategies[0])
# Create a list of players
players = []
# Loop over the strategies and the players in he Tron game
for player, strategy in zip(tron_game.players, self.strategies):
# Append the players to the player list
players.append(
PlayerAIGP(
x=player.x,
y=player.y,
direction=player.direction,
color=player.color,
alive=player.alive,
id_=player.ID,
canvas=player.canvas,
board=player.board,
strategy=strategy,
# Set the strategy evaluation function in the tron player
evaluator=self.evaluate_strategy))
# Set the players in the tron game
tron_game.players = players
# Run the tron game
tron_game.run()
# Get the winner
winner = tron_game.winner
# Check ID of the winner
if winner[0] == 0:
# Assign fitness
individual_0.wins += 1
elif winner[0] is not None:
# Assign fitness
individual_1.wins += 1
else:
# Punish behavior for drawing
individual_0.wins -= 2
individual_1.wins -= 2
# Increase the number of games played
individual_0.games += 1
individual_1.games += 1
print(self.__class__.__name__, ".__call__ winner", winner,
individual_0.fitness)
def __call__(self, individual_0, individual_1):
"""
Function call operator. Starts a Tron game and sets the
strategy of the player according to the Individual
function arguments. The fitness is the length of the tail.
:param individual_0: Individual solution
:type individual_0: Individual
:param individual_1: Individual solution
:type individual_1: Individual
"""
# Set the strategies as the GP tree of the individuals
self.strategies.append(individual_0.genome)
self.strategies.append(individual_1.genome)
# Play a number of games
for game in range(self.game_repetitions):
# Create Tron game instance
tron_game = TronAdversarial(rows=self.rows,
bike_width=self.bike_width,
draw_board=self.draw_board,
strategy=self.strategies[0])
# Create a list of players
players = []
# Loop over the strategies and the players in he Tron game
for player, strategy in zip(tron_game.players, self.strategies):
# Append the players to the player list
players.append(PlayerAIGP(x=player.x,
y=player.y,
direction=player.direction,
color=player.color,
alive=player.alive,
id_=player.ID,
canvas=player.canvas,
board=player.board,
strategy=strategy,
# Set the strategy evaluation function
# in the tron player
evaluator=self.evaluate_strategy))
# Set the players in the tron game
tron_game.players = players
# Run the tron game
tron_game.run()
# Get the winner
winner = tron_game.winner
# Check ID of the winner
if winner[0] == 0:
# Assign fitness
individual_0.wins += 1
elif winner[0] is not None:
# Assign fitness
individual_1.wins += 1
else:
# Punish behavior for drawing
individual_0.wins -= 2
individual_1.wins -= 2
# Increase the number of games played
individual_0.games += 1
individual_1.games += 1
print(self.__class__.__name__, ".__call__ winner", winner,
individual_0.wins, individual_1.wins)
