-
Notifications
You must be signed in to change notification settings - Fork 0
/
chess_ai.py
376 lines (359 loc) · 18 KB
/
chess_ai.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
# -*- coding: utf-8 -*-
from subprocess import call
from time import sleep
from game import Game
from test_helpers import heuristic_gen, get_successors
from node import Node
import heuristics
import random
import time
import json
# open JSON file to read cached oves
with open("./moves_cache.json", "r") as f:
try:
cache_moves = json.load(f)
# if the file is empty the ValueError will be thrown
except ValueError:
cache_moves = {'even': {}, 'odd': {}}
even_moves = cache_moves['even']
odd_moves = cache_moves['odd']
# Magenta = '\033[95m'
# Blue = '\033[94m'
# Green = '\033[92m'
# Yellow = '\033[93m'
# Red = '\033[91m'
# Clear = '\033[0m'
# Bold = '\033[1m'
# Underline = '\033[4m'
class Game_Engine():
def __init__(self, board_state):
self.game = Game(board_state)
self.computer = AI(self.game, 5)
def prompt_user(self):
print("\033[94m\033[1m===================================================================")
print ("\033[93m ______________ \n"
" __ ____/__ /_____________________\n"
" _ / __ __ \ _ \_ ___/_ ___/\n"
" / /___ _ / / / __/(__ )_(__ ) \n"
" \____/ /_/ /_/\___//____/ /____/ \n"
" ")
print("\033[94m===================================================================\033[0m\033[22m")
print("\nWelcome! To play, enter a command, e.g. '\033[95me2e4\033[0m'. To quit, type '\033[91mff\033[0m'.")
self.computer.print_board(str(self.game))
try:
while self.game.status < 2:
user_move = raw_input("\nMake a move: \033[95m")
print("\033[0m")
while user_move not in self.game.get_moves() and user_move != "ff":
user_move = raw_input("Please enter a valid move: ")
if user_move == "ff":
print("You surrendered.")
break;
self.game.apply_move(user_move)
captured = self.captured_pieces(str(self.game))
start_time = time.time()
self.computer.print_board(str(self.game), captured)
print("\nCalculating...\n")
if self.game.status < 2:
current_state = str(self.game)
computer_move = self.computer.ab_make_move(current_state)
PIECE_NAME = {'p': 'pawn', 'b': 'bishop', 'n': 'knight', 'r': 'rook', 'q': 'queen', 'k': 'king'}
start = computer_move[:2]
end = computer_move[2:4]
piece = PIECE_NAME[self.game.board.get_piece(self.game.xy2i(computer_move[:2]))]
captured_piece = self.game.board.get_piece(self.game.xy2i(computer_move[2:4]))
if captured_piece != " ":
captured_piece = PIECE_NAME[captured_piece.lower()]
print("---------------------------------")
print("Computer's \033[92m{piece}\033[0m at \033[92m{start}\033[0m captured \033[91m{captured_piece}\033[0m at \033[91m{end}\033[0m.").format(piece = piece, start = start, captured_piece = captured_piece, end = end)
print("---------------------------------")
else:
print("---------------------------------")
print("Computer moved \033[92m{piece}\033[0m at \033[92m{start}\033[0m to \033[92m{end}\033[0m.".format(piece = piece, start = start, end = end))
print("---------------------------------")
print("\033[1mNodes visited:\033[0m \033[93m{}\033[0m".format(self.computer.node_count))
print("\033[1mNodes cached:\033[0m \033[93m{}\033[0m".format(len(self.computer.cache)))
print("\033[1mNodes found in cache:\033[0m \033[93m{}\033[0m".format(self.computer.found_in_cache))
print("\033[1mElapsed time in sec:\033[0m \033[93m{time}\033[0m".format(time=time.time() - start_time))
self.game.apply_move(computer_move)
captured = self.captured_pieces(str(self.game))
self.computer.print_board(str(self.game), captured)
user_move = raw_input("Game over. Play again? y/n: ")
if user_move.lower() == "y":
self.game = Game()
self.computer.game = self.game
self.prompt_user()
# cache moves into JSON file
with open("./moves_cache.json", "w") as f:
if self.computer.max_depth % 2 == 0:
for key in self.computer.cache:
cache_moves["even"][key] = self.computer.cache[key]
json.dump(cache_moves, f)
else:
for key in self.computer.cache:
cache_moves["odd"][key] = self.computer.cache[key]
json.dump(cache_moves, f)
except KeyboardInterrupt:
with open("./moves_cache.json", "w") as f:
if self.computer.max_depth % 2 == 0:
for key in self.computer.cache:
cache_moves["even"][key] = self.computer.cache[key]
json.dump(cache_moves, f)
else:
for key in self.computer.cache:
cache_moves["odd"][key] = self.computer.cache[key]
json.dump(cache_moves, f)
print("\nYou quitter!")
# def write_to_cache(self):
def captured_pieces(self, board_state):
piece_tracker = {'P': 8, 'B': 2, 'N': 2, 'R': 2, 'Q': 1, 'K': 1, 'p': 8, 'b': 2, 'n': 2, 'r': 2, 'q': 1, 'k': 1}
captured = {
"w": [],
"b": []
}
for char in board_state.split()[0]:
if char in piece_tracker:
piece_tracker[char] -= 1
for piece in piece_tracker:
if piece_tracker[piece] > 0:
if piece.isupper():
captured['w'] += piece_tracker[piece] * piece
else:
captured['b'] += piece_tracker[piece] * piece
piece_tracker[piece] = 0
return captured
class AI():
def __init__(self, game, max_depth=4, leaf_nodes=[], node_count=0):
self.max_depth = max_depth
self.leaf_nodes = heuristic_gen(leaf_nodes)
self.game = game
self.node_count = node_count
if self.max_depth % 2 == 0:
self.cache = cache_moves['even']
else:
self.cache = cache_moves['odd']
self.found_in_cache = 0
def print_board(self, board_state, captured={"w": [], "b": []}):
PIECE_SYMBOLS = {'P': '♟',
'B': '♝',
'N': '♞',
'R': '♜',
'Q': '♛',
'K': '♚',
'p': '\033[36m\033[1m♙\033[0m',
'b': '\033[36m\033[1m♗\033[0m',
'n': '\033[36m\033[1m♘\033[0m',
'r': '\033[36m\033[1m♖\033[0m',
'q': '\033[36m\033[1m♕\033[0m',
'k': '\033[36m\033[1m♔\033[0m'}
board_state = board_state.split()[0].split("/")
board_state_str = "\n"
white_captured = " ".join(PIECE_SYMBOLS[piece] for piece in captured['w'])
black_captured = " ".join(PIECE_SYMBOLS[piece] for piece in captured['b'])
for i, row in enumerate(board_state):
board_state_str += str(8-i)
for char in row:
if char.isdigit():
board_state_str += " ♢" * int(char)
else:
board_state_str += " " + PIECE_SYMBOLS[char]
if i == 0:
board_state_str += " Captured:" if len(white_captured) > 0 else ""
if i == 1:
board_state_str += " " + white_captured
if i == 6:
board_state_str += " Captured:" if len(black_captured) > 0 else ""
if i == 7:
board_state_str += " " + black_captured
board_state_str += "\n"
board_state_str += " A B C D E F G H"
self.found_in_cache = 0
self.node_count = 0
print(board_state_str)
def get_moves(self, board_state=None):
if board_state == None:
board_state = str(self.game)
possible_moves = []
for move in Game(board_state).get_moves():
if (len(move) < 5 or move[4] == "q"):
clone = Game(board_state)
clone.apply_move(move)
node = Node(str(clone))
node.algebraic_move = move
possible_moves.append(node)
return possible_moves
def get_heuristic(self, board_state=None):
cache_parse = board_state.split(" ")[0] + " " + board_state.split(" ")[1]
if board_state == None:
board_state = str(self.game)
if cache_parse in self.cache:
self.found_in_cache += 1
return self.cache[cache_parse]
clone = Game(board_state)
total_points = 0
# total piece count
total_points += heuristics.material(board_state, 100)
total_points += heuristics.piece_moves(clone, 50)
total_points += heuristics.in_check(clone, 1)
total_points += heuristics.pawn_structure(board_state, 1)
self.cache[cache_parse] = total_points
return total_points
def minimax(self, node, current_depth=0):
current_depth += 1
if current_depth == self.max_depth:
# get heuristic of each node
node.value = self.get_heuristic(node.board_state)
return node.value
if current_depth % 2 == 0:
# min player's turn
self.is_turn = False
return min([self.minimax(child_node, current_depth) for child_node in self.get_moves(node.board_state, self.is_turn)])
else:
# max player's turn
self.is_turn = True
return max([self.minimax(child_node, current_depth) for child_node in self.get_moves(node.board_state, self.is_turn)])
def make_move(self, node):
self.is_turn = True
possible_moves = self.get_moves(node.board_state, self.is_turn)
for move in possible_moves:
move.value = self.minimax(move, 1)
best_move = possible_moves[0]
for move in possible_moves:
if move.value > best_move.value:
best_move = move
# best_move at this point stores the move with the highest heuristic
return best_move
def ab_make_move(self, board_state):
possible_moves = self.get_moves(board_state)
alpha = float("-inf")
beta = float("inf")
best_move = possible_moves[0]
for move in possible_moves:
board_value = self.ab_minimax(move, alpha, beta, 1)
if alpha < board_value:
alpha = board_value
best_move = move
best_move.value = alpha
# best_move at this point stores the move with the highest heuristic
return best_move.algebraic_move
def ab_minimax(self, node, alpha, beta, current_depth=0):
current_depth += 1
if current_depth == self.max_depth:
board_value = self.get_heuristic(node.board_state)
if current_depth % 2 == 0:
# pick largest number, where root is black and even depth
if (alpha < board_value):
alpha = board_value
self.node_count += 1
return alpha
else:
# pick smallest number, where root is black and odd depth
if (beta > board_value):
beta = board_value
self.node_count += 1
return beta
if current_depth % 2 == 0:
# min player's turn
for child_node in self.get_moves(node.board_state):
if alpha < beta:
board_value = self.ab_minimax(child_node,alpha, beta, current_depth)
if beta > board_value:
beta = board_value
return beta
else:
# max player's turn
for child_node in self.get_moves(node.board_state):
if alpha < beta:
board_value = self.ab_minimax(child_node,alpha, beta, current_depth)
if alpha < board_value:
alpha = board_value
return alpha
# if __name__ == "__main__":
# import unittest
# class Test_AI(unittest.TestCase):
# # def test_minimax(self):
# # data_set_1 = [8, 12, -13, 4, 1, 1, 20, 17, -5,
# # -1, -15, -12, -11, -1, 1, 17, -3, 12,
# # -7, 14, 9, 18, 4, -15, 8, 0, -6]
# # first_test_AI = AI(4, 3, data_set_1)
# # self.assertEqual(first_test_AI.minimax(Node()), 8, "Should return correct minimax when given b = 3 and d = 3")
# # data_set_2 = [-4, -17, 6, 10, -6, -1, 16, 12,
# # -12, 16, -18, -18, -20, -15, -18, -8,
# # 8, 0, 11, -14, 11, -20, 8, -2,
# # -17, -18, -11, 10, -8, -14, 7, -17]
# # second_test_AI = AI(6, 2, data_set_2)
# # self.assertEqual(second_test_AI.minimax(Node()), -8, "Should return correct minimax when given b = 2 and d = 5")
# # data_set_3 = [-7, 14, -11, -16, -3, -19, 17, 0, 15,
# # 5, -12, 18, -12, 17, 11, 12, 5, -4,
# # 13, -12, 9, 0, 12, 12, -10, 1, -19,
# # 20, 6, 13, 9, 14, 7, -3, 4, 11,
# # -14, -10, -13, -18, 17, -6, 0, -8, -1,
# # 3, 14, 6, -1, -7, 3, 8, 2, 10,
# # 6, -19, 15, -4, -10, -1, -19, -2, 6,
# # -4, 14, -3, -9, -20, 11, -18, 15, -1,
# # -9, -10, 15, 0, 8, -4, -12, 4, -17]
# # third_test_AI = AI(5, 3, data_set_3)
# # self.assertEqual(third_test_AI.minimax(Node()), -4, "Should return correct minimax when given b = 3 and d = 4")
# #
# # def test_make_move(self):
# # data_set_1 = [-4, -17, 6, 10, -6, -1, 16, 12,
# # -12, 16, -18, -18, -20, -15, -18, -8,
# # 8, 0, 11, -14, 11, -20, 8, -2,
# # -17, -18, -11, 10, -8, -14, 7, -17]
# # first_test_AI = AI(6, 2, data_set_1)
# # self.assertEqual(first_test_AI.make_move(Node()).value, -8, "Should return best move given node w/ current board state")
# # data_set_2 = [-7, 14, -11, -16, -3, -19, 17, 0, 15,
# # 5, -12, 18, -12, 17, 11, 12, 5, -4,
# # 13, -12, 9, 0, 12, 12, -10, 1, -19,
# # 20, 6, 13, 9, 14, 7, -3, 4, 11,
# # -14, -10, -13, -18, 17, -6, 0, -8, -1,
# # 3, 14, 6, -1, -7, 3, 8, 2, 10,
# # 6, -19, 15, -4, -10, -1, -19, -2, 6,
# # -4, 14, -3, -9, -20, 11, -18, 15, -1,
# # -9, -10, 15, 0, 8, -4, -12, 4, -17]
# # second_test_AI = AI(5, 3, data_set_2)
# # self.assertEqual(second_test_AI.make_move(Node()).value, -4, "Should return best move when many moves are possible")
# #
# # def test_ab(self):
# # data_set_1_prune = [8, 12, -13, 4, 1, 1, 20,
# # -1, -15, -12,
# # -7, 14, 9, 18, 8, 0, -6]
# # data_set_1_unprune = [-4, -17, 6, 10, -6, -1, 16, 12,
# # -12, 16, -18, -18, -20, -15, -18, -8,
# # 8, 0, 11, -14, 11, -20, 8, -2,
# # -17, -18, -11, 10, -8, -14, 7, -17]
# # first_prune_test_ab_AI = AI(4, 3, data_set_1_prune)
# # first_unprune_test_ab_AI = AI(4, 3, data_set_1_unprune)
# # self.assertEqual(first_prune_test_ab_AI.ab_make_move(Node()).value, 8, "Should return correct number with pruning when given b = 3 and d = 3")
# # self.assertEqual(first_unprune_test_ab_AI.ab_make_move(Node()).value == 8, False, "Should fail for unpruned dataset")
#
# # def test_get_moves(self):
# # new_game = Game()
# # first_test_AI = AI(new_game, 4, 0)
# # # White move
# # self.assertEqual(len(first_test_AI.get_moves()), 20, "Should get all initial moves for white")
# # current_turn = str(new_game).split(" ")[1]
# # self.assertEqual(current_turn, "w", "Should start as white's turn")
# # new_game.apply_move("a2a4")
# # # Black move
# # current_turn = str(new_game).split(" ")[1]
# # self.assertEqual(current_turn, "b", "Should switch to black's turn")
# # self.assertEqual(len(first_test_AI.get_moves()), 20, "Should get all initial moves for black")
# # new_game.apply_move("b8a6")
# # # White move
# # current_turn = str(new_game).split(" ")[1]
# # self.assertEqual(current_turn, "w", "Should start as white's turn")
# # self.assertEqual(len(first_test_AI.get_moves()), 21, "Should get all moves for white 3rd turn")
# def test_make_move(self):
# new_game = Game()
# first_test_AI = AI(new_game, 2, 0)
# first_test_AI.print_board(str(new_game))
# new_game.apply_move("a2a3")
# first_test_AI.print_board(str(new_game))
# new_game.apply_move(first_test_AI.ab_make_move(str(new_game)))
# first_test_AI.print_board(str(new_game))
#
# unittest.main()
if __name__ == '__main__':
new_test = Game_Engine('rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1')
new_test.prompt_user()