import random
from typing import Tuple, Optional
from prt_sim.jhu.base import BaseEnvironment
[docs]
class CardDeck:
"""For shuffling and dealing cards"""
def __init__(self):
self.cards = [1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5,
6, 6, 6, 6, 7, 7, 7, 7, 8, 8, 8, 8, 9, 9, 9, 9, 10, 10, 10, 10,
10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10]
self.deal_seq = []
def shuffle_cards(self):
self.deal_seq = random.sample(self.cards, 24)
def deal_card(self):
return self.deal_seq.pop(0)
[docs]
class Blackjack(BaseEnvironment):
r"""
Blackjack simulation class
"""
def __init__(self):
self.deck = CardDeck()
self.agent_total = 0
self.usable_ace = 0
self.dealer_card = 0
self.dealer_total = 0
self.dealer_ace = 0
self.current_state = 0
self.num_states = 203
self.num_actions = 2
[docs]
def get_number_of_states(self) -> int:
return self.num_states
[docs]
def get_number_of_actions(self) -> int:
return self.num_actions
[docs]
def reset(self,
seed: Optional[int] = None,
randomize_start: Optional[bool] = False
) -> int:
assert not randomize_start, "Randomizing the start is not supported"
if seed is not None:
random.seed(seed)
self.deck.shuffle_cards()
self.agent_total = 0
self.usable_ace = 0
self.dealer_card = 0
self.dealer_total = 0
self.dealer_ace = 0
self.current_state = 0
# deal a face up card and a second card to the dealer
self.dealer_card = self.deck.deal_card()
d_card_2 = self.deck.deal_card()
self.dealer_total = self.dealer_card + d_card_2
if self.dealer_card == 1 or d_card_2 == 1:
self.dealer_ace = 1
self.dealer_total += 10
# print("Dealer has", self.dealer_card, "and", d_card_2)
# print("Dealer has", self.dealer_total, "points.")
# deal two cards to the agent
card_1 = self.deck.deal_card()
card_2 = self.deck.deal_card()
self.agent_total = card_1 + card_2
if card_1 == 1 or card_2 == 1:
self.usable_ace = 1
self.agent_total += 10
# print("Agent has", card_1, "and", card_2)
# print("Agent has", self.agent_total, "points.")
# check to see if the agent has a natural (ace + face card)
if self.agent_total == 21:
if self.dealer_total == 21:
self.current_state = 202 # tie game
else:
self.current_state = 203 # agent wins
# otherwise, deal enough cards to the agent so that the total is >11
else:
while self.agent_total < 12:
new_card = self.deck.deal_card()
self.agent_total += new_card
if new_card == 1 and self.usable_ace == 0 and self.agent_total < 12:
self.usable_ace = 1
self.agent_total += 10
print("Agent drew a", new_card, "and now has", self.agent_total, "points.")
# now determine the initial state
self.current_state = self.get_state_index()
# reset complete; return the initial state
return self.current_state
# Use the agent's action to determine the next state and reward
[docs]
def execute_action(self,
action: int
) -> Tuple[int, float, bool]:
# action is 'stick'
if action == 0:
# dealer's turn
while self.dealer_total < 17:
new_card = self.deck.deal_card()
self.dealer_total += new_card
if new_card == 1 and self.dealer_ace == 0 and self.dealer_total < 12:
self.dealer_ace = 1
self.agent_total += 10
if self.dealer_total > 21 and self.dealer_ace == 1:
self.dealer_ace = 0
self.agent_total -= 10
if self.dealer_total > 21:
# dealer busted; agent wins
new_state = 203
reward = 1
game_end = True
else:
if self.dealer_total > self.agent_total:
# dealer wins
new_state = 201
reward = -1
game_end = True
elif self.dealer_total < self.agent_total:
# agent wins
new_state = 203
reward = 1
game_end = True
else:
# tie
new_state = 202
reward = 0
game_end = True
# action is 'hit'
elif action == 1:
new_state = self.get_next_state()
if new_state == 201:
reward = -1
game_end = True
else:
reward = 0
game_end = False
# print("new_state =", new_state, "reward = ", reward, "game_end =", game_end)
self.current_state = new_state
return new_state, float(reward), game_end
def get_state(self):
return self.current_state
def get_state_index(self):
a_idx = self.agent_total - 12
d_idx = 10 * (self.dealer_card - 1)
u_idx = 100 * self.usable_ace
return a_idx + d_idx + u_idx
def get_next_state(self):
new_card = self.deck.deal_card()
self.agent_total += new_card
if self.agent_total > 21 and self.usable_ace == 1:
self.usable_ace = 0
self.agent_total -= 10
# print("Agent drew a", new_card, "and now has", self.agent_total, "points.")
if self.agent_total > 21:
new_state = 201 # 201 is the losing state
else:
new_state = self.get_state_index()
return new_state
