-
Notifications
You must be signed in to change notification settings - Fork 0
/
alpha_dqn_engine.py
336 lines (260 loc) · 9.77 KB
/
alpha_dqn_engine.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
# -*- coding:utf-8 -*-
import numpy as np
import copy
import platform,multiprocessing
from multiprocessing import Process,Queue
import gameui
# 游戏棋盘 4*4大小 左上角坐标是(0,0)，X轴向右+1，Y轴向下+1，如下
# (0,1) (0,1) (0,2) (0,3)
# (1,1) (1,1) (1,2) (1,3)
# (2,1) (2,1) (2,2) (2,3)
# (3,1) (3,1) (3,2) (3,3)
class GameEnv(object):
VALID_CACHE = None
def __init__(self,render = False):
#np.random.seed(1)
self.invalied_num = 0
self.render = render
if(self.render):
self.render_process = SubProcess()
self.render_process.start()
self.initGame()
def initGame(self):
self.map_score = np.zeros((4, 4), dtype=np.int)
# init_score_array = np.random.randint(0,4,size=(2,2))
#
# if((init_score_array[0] == init_score_array[1]).all()):
# # 如果random出来的值相等，就重新random一次
# init_score_array = np.random.randint(0, 4, size=(2, 2))
#
# self.map_score[init_score_array[0][0]][init_score_array[0][1]] = 2
# self.map_score[init_score_array[1][0]][init_score_array[1][1]] = 2
self.invalied_num = 0
self.create_new_element()
self.create_new_element()
if(self.render):
self.render_process.put(self.map_score.tolist())
@classmethod
def init_cache(cls):
GameEnv.VALID_CACHE = np.zeros((65536,2))
print("start init_cache")
MAX_NUM = 16
temp_engine = GameEnv()
for a in range(MAX_NUM):
for b in range(MAX_NUM):
for c in range(MAX_NUM):
for d in range(MAX_NUM):
key = (a<<12) + (b<<8) + (c<<4) + d
key_ret = [False,False]
v_d = np.array([a, b, c, d])
merged_array,_ = temp_engine.handle_array(copy.deepcopy(v_d))
if (not (merged_array == v_d).all()):
key_ret[0] = True
merged_array,_ = temp_engine.handle_array(copy.deepcopy(v_d),reverse=True)
if (not (merged_array == v_d).all()):
key_ret[1] = True
GameEnv.VALID_CACHE[key] = key_ret
print("finish init_cache")
def create_new_element(self):
filter_array = []
for i in range(4):
for j in range(4):
if(self.map_score[i][j] == 0):
filter_array.append((i,j))
if(len(filter_array) == 0):
return
init_score_array = np.random.randint(0, len(filter_array))
x = filter_array[init_score_array][0]
y = filter_array[init_score_array][1]
if(np.random.uniform() >= 0.9):
new_num = 4
else:
new_num = 2
#new_num = 2
self.map_score[x][y] = new_num
return (x*4 + y)*10 + new_num
def game_over(self):
if(self.invalied_num >=5):
return True
for i in range(4):
array = self.map_score[i]
merged_array,_ = self.merge_array(array)
if (len(merged_array) < 4):
return False
array = self.map_score[:,i]
merged_array,_ = self.merge_array(array)
if (len(merged_array) < 4):
return False
return True
def valid_actioins2(self):
ret = np.zeros(4,dtype=np.bool)
for i in range(4):
array = copy.deepcopy(self.map_score[i])
merged_array,_ = self.handle_array(copy.deepcopy(array))
if(not (merged_array == array).all()):
ret[0] = True
merged_array, _ = self.handle_array(copy.deepcopy(array),reverse=True)
if (not (merged_array == array).all()):
ret[1] = True
array = copy.deepcopy(self.map_score[:, i])
merged_array, _ = self.handle_array(copy.deepcopy(array))
if (not (merged_array == array).all()):
ret[2] = True
merged_array, _ = self.handle_array(copy.deepcopy(array), reverse=True)
if (not (merged_array == array).all()):
ret[3] = True
if(ret.all()):
return ret
return ret
def valid_actioins(self):
ret = np.zeros(4,dtype=np.bool)
for i in range(4):
array = copy.deepcopy(self.map_score[i])
array = array + (array == 0)
array = np.log2(array).astype(int).tolist()
key = (array[0]<<12) + (array[1]<<8) + (array[2]<<4) + array[3]
ret[0:2] += GameEnv.VALID_CACHE[key].astype(bool)
array = copy.deepcopy(self.map_score[:, i])
array = array + (array == 0)
array = np.log2(array).astype(int).tolist()
key = (array[0]<<12) + (array[1]<<8) + (array[2]<<4) + array[3]
ret[2:4] += GameEnv.VALID_CACHE[key].astype(bool)
if(ret.all()):
return ret
return ret
def set_state(self,state):
self.map_score = np.array(state)
if (self.render):
self.render_process.put(self.map_score.tolist())
def get_value(self):
value = 0
for i in range(4):
for j in range(4):
value += int(self.map_score[i][j] / 16)
return value
def merge_array(self,array,reverse = False):
# 第一步 去除中间的0
# 第二步，合并相邻的相等的
reward = 0
filter_array = []
if(reverse):
array = array[::-1]
just_merged = False
for element in array:
if (element > 0):
if (len(filter_array) == 0 or filter_array[len(filter_array) - 1] != element or just_merged):
filter_array.append(element)
just_merged = False
else:
reward += filter_array[len(filter_array) - 1]
filter_array[len(filter_array) - 1] *= 2
just_merged = True
filter_array = filter_array if not reverse else filter_array[::-1]
# if(reward == 0):
# return filter_array,reward
# else:
# filter_array,rt = self.merge_array(filter_array,reverse=reverse)
# reward += rt
return filter_array,reward
def handle_array(self,array, reverse = False):
filter_array,reward = self.merge_array(array,reverse=reverse)
if(not reverse):
array[0:len(filter_array)] = filter_array
array[len(filter_array):len(array)] = np.zeros(len(array) - len(filter_array))
else:
array[len(array) - len(filter_array):len(array)] = filter_array
array[0:len(array) - len(filter_array)] = np.zeros(len(array) - len(filter_array))
return array,reward
#action: 0-左 1-右 2-上 3-下
def move(self,action,create_new_element=True):
pre_map = np.array(self.map_score)
reward = 0
for i in range(4):
if (action in [0, 2]):
reverse = False
else:
reverse = True
if(action in [0,1]):
array = self.map_score[i]
else:
array = self.map_score[:,i]
array,rt = self.handle_array(array,reverse)
if(action in [0,1]):
self.map_score[i] = array
else:
self.map_score[:,i] = array
reward += rt
if(self.game_over()):
reward = reward
if(not (pre_map == self.map_score).all()):
if(create_new_element):
new_num = self.create_new_element()
else:
new_num = None
self.invalied_num = 0
else:
new_num = None
reward = 0
self.invalied_num += 1
if(self.render):
self.render_process.put(self.map_score.tolist())
return reward,new_num
def print_map(self):
for i in range(4):
str_tmp = ""
for j in range(4):
str_tmp += (" " + str(self.map_score[i][j].item())).ljust(5)
print(str_tmp)
# return:(地图所有分值,最大块分值)
def get_score(self):
all_score = 0
max_score = 0
not_zero_num = 0
for i in range(4):
for j in range(4):
tmp_score = self.map_score[i][j]
not_zero_num = not_zero_num + 1 if tmp_score > 0 else not_zero_num
all_score += tmp_score
if(tmp_score > max_score):
max_score = tmp_score
return all_score,max_score,not_zero_num
def destroy(self):
pass
class SubProcess(object):
def __init__(self):
self.q = Queue()
self.p = Process(target=self.f)
def f(self):
gameui.start_render(self.q)
def put(self,item):
self.q.put(item)
def start(self):
self.p.start()
if __name__ == "__main__":
if platform.system() == "Darwin":
multiprocessing.set_start_method('spawn')
GameEnv.init_cache()
g = GameEnv(render=True)
g.map_score[0] = np.array([2, 16, 8, 32])
g.map_score[1] = np.array([16, 4, 4, 8])
g.map_score[2] = np.array([4, 32, 64, 128])
g.map_score[3] = np.array([8, 4, 8, 256])
g.print_map()
print(g.valid_actioins())
# import time
# t1 = time.time()
# v1 = g.valid_actioins()
# t2 = time.time()
# v2 = g.valid_actioins2()
# t3 = time.time()
# print(v1,t2 - t1)
# print(v2,t3 - t2)
while True:
getc = input("请输入：")
reward = g.move(int(getc))
print("reward",reward)
if(g.game_over()):
print("游戏结束")
g.print_map()
break
g.print_map()