def train(self, num_of_train, opponent, debug=0, frenzy=0, recover_rate=0):
#recover_rate=1 means recovers from frenzy immediately
#recover_rate=0 means it never recovers
rate = 1 - recover_rate
frenzy_degree = frenzy
self.frenzy = frenzy
self.sbb.frenzy = frenzy
self.bbb.frenzy = frenzy
game = holdem.Holdem(2, 4, 4, debug)
for i in range(num_of_train):
if np.random.rand() < frenzy_degree:
self.frenzy = frenzy
self.sbb.frenzy = frenzy
self.bbb.frenzy = frenzy
else:
self.frenzy = 0
self.sbb.frenzy = frenzy
self.bbb.frenzy = frenzy
frenzy_degree *= rate
if i % 2 == 1:
result = self.sbb.sim_one_hand(opponent, game, 1, debug=debug)
else:
result = self.bbb.sim_one_hand(opponent, game, 0, debug=debug)
game.endRound()
# print result[1]
self.learn_one(result[0], result[1], self.status.dealer)
self.frenzy = 0
self.sbb.frenzy = 0
self.bbb.frenzy = 0
return
def compete(self, opponent, num_of_games=100, debug=1):
self.frenzy=0
start_cash=0
game=holdem.Holdem(2, 4, 4, debug)
for i in range(num_of_games):
result=self.sim_one_hand(opponent, game, dealer=i%2, debug=debug)
game.endRound()
if debug:
print "End of one hand. The winning is", result[1], "\n"
start_cash= start_cash+ result[1]
return start_cash
def train(self, num_of_train, opponent, debug=0, frenzy=0):
self.frenzy = frenzy
game = holdem.Holdem(2, 4, 4, debug)
for i in range(num_of_train):
result = self.sim_one_hand(opponent,
game,
dealer=i % 2,
debug=debug)
game.endRound()
# print result[1]
self.learn_one(result[0], result[1])
self.status = StatStatus()
opponent.status = StatStatus()
self.frenzy = 0
def compete(self, opponent, num_of_games=100, debug=0):
start_cash = 0
game = holdem.Holdem(2, 4, 4, debug)
for i in range(num_of_games):
if i % 2 == 1:
result = self.sbb.sim_one_hand(opponent,
game,
dealer=1,
debug=debug)
else:
result = self.bbb.sim_one_hand(opponent,
game,
dealer=0,
debug=debug)
game.endRound()
if debug:
print "End of one hand. The winning is", result[1], "\n"
start_cash += result[1]
return start_cash
def train(self,num_of_train, opponent, debug=0, frenzy=0,
recover_rate=0):
#recover_rate=1 means recovers from frenzy immediately
#recover_rate=0 means it never recovers
rate=1-recover_rate
frenzy_degree=frenzy
self.frenzy= frenzy
game= holdem.Holdem(2, 4, 4, debug);
for i in range(num_of_train):
if np.random.rand() < frenzy_degree:
self.frenzy=frenzy
else:
self.frenzy=0
frenzy_degree *=rate
result=self.sim_one_hand(opponent, game, dealer=i%2, debug=debug)
game.endRound()
# print result[1]
self.learn_one(result[0], result[1])
self.frenzy= 0