-
Notifications
You must be signed in to change notification settings - Fork 0
/
gameAIButBetter.py
438 lines (312 loc) · 11.2 KB
/
gameAIButBetter.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
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
'''enhanced AI using Q-Learning, cleaned up a lot of the previous code'''
import random
import shelve
import datetime
import numpy as np
y, x = 6, 7
gameboard = [[0 for j in range(x)] for i in range(y)]
COIN = 7
MOVES = []
TOPPED_OUT = []
NOT_TOPPED_OUT = list(range(x))
print('loading shelve....')
q_table = shelve.open('q_table_shelf.db', writeback=True)
def print_gameboard():
'''prints gameboard'''
for i in gameboard:
print(i)
def check_connect_list(list_to_check):
'''checks if sent list has a 4 of the same coins'''
if len(list_to_check) != 4 or 0 in list_to_check:
return False
if sum(list_to_check) in [20, 28]:
return True
return False
def check_diagonals(x_pos, y_pos):
'''checks for 4 connect4 on x, y'''
index_error_count = 0
for j in range(0, 4):
list_diagonal_top = []
list_diagonal_bottom = []
for i in range(1 + j, 5 + j):
if y_pos - 4 + i >= 0 and x_pos - 4 + i >= 0:
try:
list_diagonal_top.append(
gameboard[y_pos - 4 + i][x_pos - 4 + i])
except IndexError:
index_error_count = index_error_count + 1
if y_pos + 4 - i >= 0 and x_pos - 4 + i >= 0:
try:
list_diagonal_bottom.append(
gameboard[y_pos + 4 - i][x_pos - 4 + i])
except IndexError:
index_error_count = index_error_count + 1
if check_connect_list(list_diagonal_top) is True:
return '4connected'
if check_connect_list(list_diagonal_bottom) is True:
return '4connected'
return False
def check_connect_4(x_pos, y_pos):
'''checks if there is a 4 sequence from the recived x and y coord'''
index_error_count = 0
for i in range(1, 5):
try:
if check_connect_list(gameboard[y_pos][x_pos - 4 + i: x_pos + i]) is True:
return '4connected'
except IndexError:
index_error_count = index_error_count + 1
try:
if check_connect_list([gameboard[i][x_pos] for i in range(y_pos - 0, y_pos + 4)]) is True:
return '4connected'
except IndexError:
index_error_count = index_error_count + 1
return check_diagonals(x_pos, y_pos)
def probable_move(x_pos, y_pos):
'''checks if coin placed at this position connects4'''
for i in [7, 5]:
gameboard[y_pos][x_pos] = i
if check_connect_4(x_pos, y_pos) == '4connected':
gameboard[y_pos][x_pos] = 0
return '4connected'
gameboard[y_pos][x_pos] = 0
return False
def keywithmaxval(dic):
'''retruns the key of the highest value in the dictionary'''
k = list(dic.keys())
val = np.array(list(dic.values()))
return k[int(random.choice(np.argwhere(val == np.amax(val))))]
EPSILION = 1
DISCOUNT_FACTOR = 0.9
LEARNING_RATE = 0.9
def is_creating_4_above(x_pos):
'''checks if there is a connect 4 above current pos'''
if top[x_pos] == 0:
return False
y_pos = top[x_pos]
for i in [7, 5]:
gameboard[y_pos - 1][x_pos] = i
if check_connect_4(x_pos, y_pos - 1) == '4connected':
gameboard[y_pos - 1][x_pos] = 0
return True
gameboard[y_pos - 1][x_pos] = 0
if check_connect_4(x_pos, top[x_pos] - 1) == '4connected':
print('checking above xpos')
return True
return False
def stop_4_connecting():
'''check_connect_4(takes x_pos and then y_pos) to highest y_pos and check for every x_pos'''
if TURN == 0:
random.choice([2, 3, 4])
if TURN == 1:
if top[3] == 4:
return random.choice([2, 4])
return 3
length = len(top)
for i in range(length):
if top[i] == -1:
continue
if probable_move(i, top[i]) == '4connected':
if check_connect_4(i, top[i]) == '4connected':
return i
return i
move = random.choice(NOT_TOPPED_OUT)
gameboard_as_key = ''.join(str(item)
for innerlist in gameboard for item in innerlist)
if np.random.random() < EPSILION:
move = keywithmaxval(q_table[gameboard_as_key])
count = 0
while is_creating_4_above(move) is True:
if count == 5:
move = random.choice(NOT_TOPPED_OUT)
break
move = random.choice(NOT_TOPPED_OUT)
count = count + 1
return move
def valid_move(x_pos, y_pos):
'''checks if position is a valid move or not'''
if x_pos in TOPPED_OUT:
return False
#print(NOT_TOPPED_OUT)
if y_pos == 0:
TOPPED_OUT.append(x_pos)
NOT_TOPPED_OUT.remove(x_pos)
return True
def valid_move_x(x_pos):
'''sanitizes input and checks if given position is valid or not'''
try:
x_pos = int(x_pos) - 1
except ValueError:
return False
if x_pos in NOT_TOPPED_OUT:
return True
return False
def add_coin(x_pos):
'''adds a coin at given position'''
#MOVES.append(str(x_pos))
for j in range(y - 1, -1, -1):
if gameboard[j][x_pos] == 0 and valid_move(x_pos, j) is True:
gameboard[j][x_pos] = COIN
y_pos = j
top[x_pos] = top[x_pos] - 1
return check_connect_4(x_pos, y_pos)
def add_coin_key_maker(x_pos):
'''adds a coin at given position tempraraitoly'''
#MOVES.append(str(x_pos))
y_pos = top[x_pos]
gameboard[y_pos][x_pos] = COIN
key = ''.join(str(item) for innerlist in gameboard for item in innerlist)
gameboard[y_pos][x_pos] = 0
actions_possible = NOT_TOPPED_OUT.copy()
if y_pos == 0:
actions_possible.remove(x_pos)
return key, actions_possible
def remove_coin(x_pos):
'''removes coin at given position'''
y_pos = top[x_pos]
gameboard[y_pos + 1][x_pos] = 0
top[x_pos] = top[x_pos] + 1
def get_move_reward(x_pos):
'''calculates the reward for that move at the global board state'''
y_pos = top[x_pos]
for i in [7, 5]:
gameboard[y_pos][x_pos] = i
if check_connect_4(x_pos, y_pos) == '4connected':
gameboard[y_pos][x_pos] = 0
if i == 7:
#print('aha saved a connect4 at ' + str(x_pos + 1))
return -500
#print('i win in this simple move at ' + str(x_pos + 1))
return 10000
gameboard[y_pos][x_pos] = 0
return -1
def generate_next_move(board_state):
'''next move is generated either through random or highest q value, EPSILION mediates this'''
gameboard_as_key = ''.join(str(item)
for innerlist in board_state for item in innerlist)
if gameboard_as_key not in q_table:
q_table[gameboard_as_key] = {el: 0 for el in NOT_TOPPED_OUT}
action = stop_4_connecting()
reward = get_move_reward(action)
old_q_value = q_table[gameboard_as_key][action]
#add_coin(action)
#add_coin_key_maker(action)
gameboard_as_key_next, next_moves = add_coin_key_maker(action)
#print('TURN: ' + str(TURN) + ' -> ', end=' ')
#print(next_moves)
#print_gameboard()
if next_moves == []:
next_moves = NOT_TOPPED_OUT
if gameboard_as_key_next not in q_table:
q_table[gameboard_as_key_next] = {el: 0 for el in next_moves}
#if TURN <= 41:
#remove_coin(action)
#print_gameboard()
temporal_difference = reward + \
(DISCOUNT_FACTOR * q_table[gameboard_as_key]
[keywithmaxval(q_table[gameboard_as_key_next].copy())]) - old_q_value
interim_dict = q_table[gameboard_as_key]
new_q_value = old_q_value + (LEARNING_RATE * temporal_difference)
interim_dict[action] = new_q_value
q_table[gameboard_as_key] = interim_dict
print(q_table[gameboard_as_key])
return action
def swap_gameboard():
'''to change all the 5s to 7s and vice versa, used to calculatate board
for AI2 but feed to the main db'''
for i in range(x):
for j in range(5, top[i], -1):
if gameboard[j][i] == 7:
gameboard[j][i] = 5
else:
gameboard[j][i] = 7
def take_input(player):
'''takes input if users chance, otherwise generates ai_move'''
if player == 7: # change to 7 for
COIN = 5
#print_gameboard()
swap_gameboard()
#print('AI2 is calculating its next move.....')
ai_2_move = generate_next_move(gameboard)
#print('Playing at: ' + str(ai_2_move + 1))
COIN = 7
swap_gameboard()
return ai_2_move
# if player == 9:
# return random.choice(NOT_TOPPED_OUT)
#
# if player == 8:
# move = input('drop COIN at: ').rstrip()
# while valid_move_x(move) is False:
# move = input('last input invalid, go again: ').rstrip()
# return int(move) - 1
#print('AI is calculating its next move.....')
#ai_move = stop_4_connecting()
ai_move = generate_next_move(gameboard)
#ai_move = random.randint(0, 6)
#print('Playing at: ' + str(ai_move + 1))
return ai_move
def stamp_time():
'''now time - start time is printed in hours'''
later_time = datetime.datetime.now()
difference = later_time - first_time
print(str(divmod((difference.days * 24 * 60 * 60) +
difference.seconds, 60)[0] / 60) + ' Hours gone by')
EPISODES = 3
WINS_7 = 0
WINS_5 = 0
DRAWS = 0
DRAW = False
first_time = datetime.datetime.now()
#print_gameboard()
for e in range(EPISODES):
print('---------------------------------------new game----------------------------------------')
gameboard = [[0 for j in range(x)] for i in range(y)]
TURN = 0
COIN = 7
top = [5 for i in range(x)]
while add_coin(take_input(COIN)) != '4connected':
if TURN % 2 == 0:
COIN = 5
else:
COIN = 7
print_gameboard()
TURN = TURN + 1
if TURN == 42:
print('omg a draw!!1! at game: ' +
str(e + 1) + '/' + str(EPISODES))
DRAW = True
break
print_gameboard()
#with open('pastGames.txt', 'a') as f:
# f.write('\n')
# f.write(''.join(MOVES))
# f.close()
MOVES = []
TOPPED_OUT = []
NOT_TOPPED_OUT = list(range(x))
if DRAW is False:
if COIN == 7:
WINS_7 = WINS_7 + 1
else:
WINS_5 = WINS_5 + 1
print('wow ' + str(COIN) + ' wins game ' +
str(e + 1) + '/' + str(EPISODES))
else:
DRAWS = DRAWS + 1
DRAW = False
print('5 wins: ' + str(WINS_5) + ' | 7 wins: ' +
str(WINS_7) + ' | draws: ' + str(DRAWS))
print('5 win%: ' + str(WINS_5 * 100 / (e + 1)) + ' | 7 win%: ' +
str(WINS_7 * 100 / (e + 1)) + ' | draw%: ' + str(DRAWS * 100 / (e + 1)))
print(str(((e + 1) / EPISODES) * 100) + '%' + ' completion')
if (e + 1) % (EPISODES/100) == 0:
print('quicksaving')
q_table.sync()
later_time = datetime.datetime.now()
difference = later_time - first_time
print(str(divmod((difference.days * 24 * 60 * 60) +
difference.seconds, 60)[0] / 60) + ' Hours gone by')
q_table.close()
later_time = datetime.datetime.now()
difference = later_time - first_time
print(divmod((difference.days * 24 * 60 * 60) + difference.seconds, 60))