-
Notifications
You must be signed in to change notification settings - Fork 1
/
MrZeroTree.py
executable file
·536 lines (477 loc) · 22.6 KB
/
MrZeroTree.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
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
#!/usr/bin/env python3
# -*- coding: UTF-8 -*-
from Util import log,calc_score
from Util import ORDER_DICT,ORDER_DICT2,ORDER_DICT5,SCORE_DICT,INIT_CARDS
from MrRandom import MrRandom
from MrGreed import MrGreed
from MrZeroTreeSimple import GameState,MrZeroTreeSimple,MCTS_EXPL,BETA
from ScenarioGenerator.ScenarioGen import ScenarioGen
from ScenarioGenerator.ImpScenarioGen import ImpScenarioGen
from OfflineInterface import OfflineInterface
from MCTS.mcts import mcts
import torch
import torch.nn.functional as F
import copy,math,time,random,numpy
print_level=0
BETA_POST_RECT=0.015
log("BETA_POST_RECT: %.3f, BETA: %.2f"%(BETA_POST_RECT,BETA))
class MrZeroTree(MrZeroTreeSimple):
def __init__(self,room=0,place=0,name="default",pv_net=None,device=None,train_mode=False,
sample_b=-3,sample_k=-3,mcts_b=50,mcts_k=2):
MrRandom.__init__(self,room,place,name)
if device==None:
devnum=torch.cuda.device_count()
self.device=torch.device("cuda:%d"%(random.randint(0,devnum-1)))
elif isinstance(device,str):
self.device=torch.device(device)
else:
self.device=device
if pv_net==None:
self.load_pv_net(net_para_loc="Zero-29th-25-11416629-720.pt")
elif isinstance(pv_net,str):
self.load_pv_net(net_para_loc=pv_net)
else:
self.pv_net=pv_net
self.sample_b=sample_b
self.sample_k=sample_k
self.mcts_b=mcts_b
self.mcts_k=mcts_k
self.train_mode=train_mode
if self.train_mode:
self.train_datas=[]
#self.int_method_printed_flag=False
"""def wasserstein(l):
return (l[0]-l[1]).abs().sum()+(l[1]-l[2]).abs().sum()+(l[0]-l[2]).abs().sum()
def select_interact_cards(self,legal_choice,level=2):
input("useless")
oh_score=MrZeroTreeSimple.score_lists_oh(self.scores,self.place)
oh_table=MrZeroTreeSimple.four_cards_oh(self.cards_on_table,self.place)
cards_remain=ScenarioGen.gen_cards_remain(self.history,self.cards_on_table,self.cards_list)
#cards_remain.sort()
oh_card=torch.zeros(52*4)
for c in self.cards_list:
oh_card[ORDER_DICT[c]]=1
for c in cards_remain:
oh_card[52*1+ORDER_DICT[c]]=1/3
oh_card[52*2+ORDER_DICT[c]]=1/3
oh_card[52*3+ORDER_DICT[c]]=1/3
l_re=[]
for c in cards_remain:
p_temp=[]
for i in range(1,4):
oh_card_cp=oh_card.clone()
for j in range(1,4):
if j==i:
oh_card_cp[52*j+ORDER_DICT[c]]=1
else:
oh_card_cp[52*j+ORDER_DICT[c]]=0
netin=torch.cat([oh_card_cp,oh_score,oh_table])
with torch.no_grad():
p,_=self.pv_net(netin.to(self.device))
p_legal=torch.tensor([p[ORDER_DICT[c]] for c in legal_choice])
p_legal-=p_legal.max()
p_temp.append(p_legal)
l_re.append((c,MrZeroTree.wasserstein(p_temp).item()))
l_re.sort(key=lambda x:x[1],reverse=True)
l_re=l_re[0:min(len(l_re),level)]
return [c for c,v in l_re]"""
def cards_lists_oh_post_rect(cards_lists,place):
"""
return a 208-length one hot, in raletive order
the order is [me,me+1,me+2,me+3]
"""
oh=torch.zeros(52*4)
for c in cards_lists[place]:
oh[ORDER_DICT[c]]=1
for i in range(1,4):
for c in cards_lists[(place+i)%4]:
oh[52*1+ORDER_DICT[c]]=1/3
oh[52*2+ORDER_DICT[c]]=1/3
oh[52*3+ORDER_DICT[c]]=1/3
return oh
def prepare_ohs_post_rect(cards_lists,cards_on_table,score_lists,place):
oh_card=MrZeroTree.cards_lists_oh_post_rect(cards_lists,place)
oh_score=MrZeroTreeSimple.score_lists_oh(score_lists,place)
oh_table=MrZeroTreeSimple.four_cards_oh(cards_on_table,place)
return torch.cat([oh_card,oh_score,oh_table])
def possi_rectify_pvnet(self,cards_lists,scores,cards_on_table,pnext,legal_choice,choice):
netin=MrZeroTree.prepare_ohs_post_rect(cards_lists,cards_on_table,scores,pnext)
with torch.no_grad():
p,_=self.pv_net(netin.to(self.device))
#p_legal=[(c,p[ORDER_DICT[c]]) for c in legal_choice if c[0]!="C"] #G on Feb 9th
p_legal=[(c,p[ORDER_DICT[c]]) for c in legal_choice if c[0]==choice[0]] #Important!
#p_legal=[(c,p[ORDER_DICT[c]]) for c in legal_choice] #Before Jan 19th
v_max=max((v for c,v in p_legal))
p_line=[(c,1+BETA_POST_RECT*(v-v_max)/BETA) for c,v in p_legal]
possi_line=max((v for c,v in p_line if c==choice).__next__(),0.1)
"""p_exp=[(c,math.exp(BETA_POST_RECT/BETA*(v-v_max))) for c,v in p_legal]
v_sum=sum((v for c,v in p_exp))
p_exp=[(c,v/v_sum) for c,v in p_exp]
possi_exp=(v for c,v in p_exp if c==choice).__next__()"""
#log(["%s: %.4f, %.4f"%(p_legal[i][0],p_line[i][1],p_exp[i][1]) for i in range(len(p_legal))])
return possi_line
def decide_rect_necessity(self,thisuit,choice):
# C
if thisuit==choice[0] and choice[1] not in "234567":
return 3
# D
if thisuit=="A" and choice[1] not in "234567":
return 4
return -1
def possi_rectify(self,cards_lists,thisuit):
"""
posterior probability rectify
cards_lists is in absolute order
"""
cards_lists=copy.deepcopy(cards_lists)
scores=copy.deepcopy(self.scores)
result=1.0
for history in [self.cards_on_table,]+self.history[::-1]:
if len(history)==5:
for c in history[1:]:
if c in SCORE_DICT:
scores[last_winner].remove(c)
last_winner=history[0]
cards_on_table=copy.copy(history)
pnext=(cards_on_table[0]+len(history)-1)%4
for i in range(len(cards_on_table)-1):
pnext=(pnext-1)%4
choice=cards_on_table.pop()
cards_lists[pnext].append(choice)
#不用修正我自己
if pnext==self.place:
continue
#决定是否需要修正
nece=self.decide_rect_necessity(thisuit,choice)
if nece<0:
continue
suit=cards_on_table[1][0] if len(cards_on_table)>1 else "A"
cards_dict=MrGreed.gen_cards_dict(cards_lists[pnext])
legal_choice=MrGreed.gen_legal_choice(suit,cards_dict,cards_lists[pnext])
possi_pvnet=self.possi_rectify_pvnet(cards_lists,scores,cards_on_table,pnext,legal_choice,choice)
if print_level>=4:
log("rectify %s(%d): %.4e"%(choice,nece,possi_pvnet),end="");input()
result*=possi_pvnet
else:
assert len(scores[0])==len(scores[1])==len(scores[2])==len(scores[3])==0, "scores left not zero: %s"%(scores,)
assert len(cards_lists[0])==len(cards_lists[1])==len(cards_lists[2])==len(cards_lists[3])==13, "cards_lists not equal 4x13: %s"%(cards_lists,)
if print_level>=3:
log("final cards possi: %.4e"%(result))
return result
"""def cards_lists_oh_post_rect(cards_lists,void_info,place):
oh=torch.zeros(52*4)
for c in cards_lists[place]:
oh[ORDER_DICT[c]]=1
n_suit={s:sum([not void_info[(place+i)%4][s] for i in range(1,4)]) for s in "SHDC"}
for c in cards_lists[(place+1)%4]+cards_lists[(place+2)%4]+cards_lists[(place+3)%4]:
for i in range(1,4):
if not void_info[(place+i)%4][c[0]]:
oh[52*i+ORDER_DICT[c]]=1/n_suit[c[0]]
num_cards=sum([len(cards_lists[i]) for i in range(4)])
assert abs(oh.sum()-num_cards)<1e-4, "%f != %f"%(oh.sum(),num_cards)
return oh
def prepare_ohs_post_rect(cards_lists,void_info,cards_on_table,score_lists,place):
oh_card=MrZeroTree.cards_lists_oh_post_rect(cards_lists,void_info,place)
oh_score=MrZeroTreeSimple.score_lists_oh(score_lists,place)
oh_table=MrZeroTreeSimple.four_cards_oh(cards_on_table,place)
return torch.cat([oh_card,oh_score,oh_table])
def possi_rectify_pvnet(self,cards_lists,void_info,scores,cards_on_table,pnext,legal_choice,choice):
netin=MrZeroTree.prepare_ohs_post_rect(cards_lists,void_info,cards_on_table,scores,pnext)
with torch.no_grad():
p,_=self.pv_net(netin.to(self.device))
p_legal=[(c,p[ORDER_DICT[c]]) for c in legal_choice if c[0]==choice[0]]
v_max=max((v for c,v in p_legal))
#p_line=[(c,1+BETA_POST_RECT*(v-v_max)/BETA) for c,v in p_legal]
#possi_line=max((v for c,v in p_line if c==choice).__next__(),0.1)
p_exp=[(c,math.exp(BETA_POST_RECT/BETA*(v-v_max))) for c,v in p_legal]
possi_exp=(v for c,v in p_exp if c==choice).__next__()
return possi_exp
def possi_rectify(self,cards_lists_ori,thisuit,cards_played,scores_stage,void_info_stage):
cards_lists=copy.deepcopy(cards_lists_ori)
for i in range(4):
cards_lists[i]+=cards_played[i]
else:
assert len(cards_lists[0])==len(cards_lists[1])==len(cards_lists[2])==len(cards_lists[3])==13
stage=0
result=1.0
for history in self.history+[self.cards_on_table]:
assert stage+sum([len(cards_lists[i]) for i in range(4)])==52
pnext=history[0]
cards_on_table=[pnext,]
for c_num,c in enumerate(history[1:]):
rect_flag=True
if pnext==self.place:
rect_flag=False
else:
nece=self.decide_rect_necessity(thisuit,c)
if nece<0:
rect_flag=False
if rect_flag:
suit=history[1][0] if c_num>0 else "A"
cards_dict=MrGreed.gen_cards_dict(cards_lists[pnext])
legal_choice=MrGreed.gen_legal_choice(suit,cards_dict,cards_lists[pnext])
possi_pvnet=self.possi_rectify_pvnet(cards_lists,void_info_stage[stage],scores_stage[stage],
cards_on_table,pnext,legal_choice,c)
result*=possi_pvnet
if print_level>=4:
log("rectify %s(%d): %.4e"%(c,nece,possi_pvnet),end="");input()
cards_lists[pnext].remove(c)
cards_on_table.append(c)
pnext=(pnext+1)%4
stage+=1
assert cards_lists==cards_lists_ori
if print_level>=3:
log("final result: %.4e"%(result),end="");input()
return result"""
def public_info(self):
"""
collect public information for possi_rectfy, including:
cards_played,
scores at different stage,
break suits at different stage
"""
input("not using")
cards_played=[[],[],[],[]] #absolute order
scores=[[],[],[],[]]
void_info=[{'S':False,'H':False,'D':False,'C':False},{'S':False,'H':False,'D':False,'C':False},
{'S':False,'H':False,'D':False,'C':False},{'S':False,'H':False,'D':False,'C':False},]
scores_stage=[]
void_info_stage=[]
for r_num,history in enumerate(self.history+[self.cards_on_table]):
pnext=history[0]
for c_num,c in enumerate(history[1:]):
void_info_stage.append(copy.deepcopy(void_info))
scores_stage.append(copy.deepcopy(scores))
cards_played[pnext].append(c)
if pnext!=history[0] and c[0]!=history[1][0]:
void_info[pnext][history[1][0]]=True
pnext=(pnext+1)%4
if r_num<len(self.history)-1:
winner=self.history[r_num+1][0]
elif r_num==len(self.history)-1:
winner=self.cards_on_table[0]
else:
continue
for c in history[1:]:
if c in SCORE_DICT:
scores[winner].append(c)
num_cards_played=sum([len(i) for i in cards_played])
assert 4*len(self.history)+len(self.cards_on_table)-1==num_cards_played
assert scores==self.scores
return cards_played,scores_stage,void_info_stage
def pick_a_card(self):
#确认桌上牌的数量和自己坐的位置相符
assert (self.cards_on_table[0]+len(self.cards_on_table)-1)%4==self.place
#utility datas
suit=self.decide_suit() #inherited from MrRandom
cards_dict=MrGreed.gen_cards_dict(self.cards_list)
#如果别无选择
if cards_dict.get(suit)!=None and len(cards_dict[suit])==1:
choice=cards_dict[suit][0]
if print_level>=1:
log("I have no choice but %s"%(choice))
return choice
if len(self.cards_list)==1:
return self.cards_list[0]
if print_level>=1:
log("my turn: %s, %s, %s"%(self.cards_on_table,self.cards_list,self.scores))
legal_choice=MrGreed.gen_legal_choice(suit,cards_dict,self.cards_list)
#imp_cards=self.select_interact_cards(legal_choice)
if self.sample_k>=0:
sce_num=self.sample_b+int(self.sample_k*len(self.cards_list))
assert self.sample_b>=0 and sce_num>0
sce_gen=ScenarioGen(self.place,self.history,self.cards_on_table,self.cards_list,number=sce_num)
scenarios=[i for i in sce_gen]
else:
assert self.sample_k<0 and self.sample_b<0
sce_gen=ImpScenarioGen(self.place,self.history,self.cards_on_table,self.cards_list,suit,
level=-1*self.sample_k,num_per_imp=-1*self.sample_b)
#imp_cards=imp_cards,num_per_imp=-1*self.sample_b)
scenarios=sce_gen.get_scenarios()
#cards_played,scores_stage,void_info_stage=self.public_info()
scenarios_weight=[]
cards_lists_list=[]
for cll in scenarios:
if print_level>=3:
log("analyzing: %s"%(cll))
cards_lists=[None,None,None,None]
cards_lists[self.place]=copy.copy(self.cards_list)
for i in range(3):
cards_lists[(self.place+i+1)%4]=cll[i]
#scenarios_weight.append(1.0)
#scenarios_weight.append(self.possi_rectify(cards_lists,suit,cards_played,scores_stage,void_info_stage))
scenarios_weight.append(self.possi_rectify(cards_lists,suit))
#scenarios_weight[-1]*=self.int_equ_class(cards_lists,suit)
#scenarios_weight[-1]*=self.int_equ_class_li(cards_lists,suit)
cards_lists_list.append(cards_lists)
else:
del scenarios
if print_level>=2:
log("scenarios_weight: %s"%(["%.4f"%(i) for i in scenarios_weight],))
weight_sum=sum(scenarios_weight)
scenarios_weight=[i/weight_sum for i in scenarios_weight]
assert (sum(scenarios_weight)-1)<1e-6, "scenario weight is %.8f: %s"%(sum(scenarios_weight),scenarios_weight,)
#legal_choice=MrGreed.gen_legal_choice(suit,cards_dict,self.cards_list)
d_legal={c:0 for c in legal_choice}
searchnum=self.mcts_b+self.mcts_k*len(legal_choice)
for i,cards_lists in enumerate(cards_lists_list):
#initialize gamestate
gamestate=GameState(cards_lists,self.scores,self.cards_on_table,self.place)
#mcts
if self.mcts_k>=0:
searcher=mcts(iterationLimit=searchnum,rolloutPolicy=self.pv_policy,
explorationConstant=MCTS_EXPL)
searcher.search(initialState=gamestate)
for action,node in searcher.root.children.items():
d_legal[action]+=scenarios_weight[i]*node.totalReward/node.numVisits
elif self.mcts_k==-1:
input("not using")
netin=MrZeroTree.prepare_ohs(cards_lists,self.cards_on_table,self.scores,self.place)
with torch.no_grad():
p,_=self.pv_net(netin.to(self.device))
p_legal=[(c,p[ORDER_DICT[c]]) for c in legal_choice]
p_legal.sort(key=lambda x:x[1],reverse=True)
d_legal[p_legal[0][0]]+=1
elif self.mcts_k==-2:
input("not using")
assert self.sample_b==1 and self.sample_k==0 and self.mcts_b==0, "This is raw-policy mode"
netin=MrZeroTree.prepare_ohs_post_rect(cards_lists,self.cards_on_table,self.scores,self.place)
with torch.no_grad():
p,_=self.pv_net(netin.to(self.device))
p_legal=[(c,p[ORDER_DICT[c]]) for c in legal_choice]
p_legal.sort(key=lambda x:x[1],reverse=True)
return p_legal[0][0]
else:
raise Exception("reserved")
if print_level>=2:
log("d_legal: %s"%({k:float("%.1f"%(v)) for k,v in d_legal.items()}))
#time.sleep(5+10*random.random())
best_choice=MrGreed.pick_best_from_dlegal(d_legal)
"""
if len(legal_choice)>1:
g=self.g_aux[self.place]
g.cards_on_table=copy.copy(self.cards_on_table)
g.history=copy.deepcopy(self.history)
g.scores=copy.deepcopy(self.scores)
g.cards_list=copy.deepcopy(self.cards_list)
gc=g.pick_a_card()
netin=MrZeroTree.prepare_ohs(cards_lists,self.cards_on_table,self.scores,self.place)
with torch.no_grad():
p,_=self.pv_net(netin.to(self.device))
p_legal=[(c,p[ORDER_DICT[c]].item()) for c in legal_choice if c[0]==gc[0]]
v_max=max((v for c,v in p_legal))
p_legal=[(c,1+BETA_POST_RECT*(v-v_max)/BETA) for c,v in p_legal]
p_legal.sort(key=lambda x:x[1],reverse=True)
p_choice=(v for c,v in p_legal if c==gc).__next__()
possi=max(p_choice,0.2)
log("greed, %s, %s, %s, %.4f"%(gc,suit,gc==p_legal[0][0],possi),logfile="stat_sim.txt",fileonly=True)
p_legal=[(c,p[ORDER_DICT[c]].item()) for c in legal_choice if c[0]==best_choice[0]]
v_max=max((v for c,v in p_legal))
p_legal=[(c,1+BETA_POST_RECT*(v-v_max)/BETA) for c,v in p_legal]
p_legal.sort(key=lambda x:x[1],reverse=True)
p_choice=(v for c,v in p_legal if c==best_choice).__next__()
possi=max(p_choice,0.2)
log("zerotree, %s, %s, %s, %.4f"%(best_choice,suit,best_choice==p_legal[0][0],possi),logfile="stat_sim.txt",fileonly=True)"""
return best_choice
@staticmethod
def family_name():
return 'Mr.ZeroTree'
def example_DJ():
zt3=MrZeroTree(room=255,place=3,name='zerotree3',mcts_b=10,mcts_k=2,sample_b=-1,sample_k=-2)
zt3.cards_list=["HQ","HJ","H8","SA","S5","S4","S3","CQ","CJ","C4"]
zt3.cards_on_table=[1,"DJ","D8"]
zt3.history=[[0,"H3","H5","H4","H7"],[3,"S6","SJ","HK","S10"],[0,"DQ","DA","D9","D3"]]
zt3.scores=[["HK"],[],[],["H3","H5","H4","H7"]]
"""cards_dict=MrGreed.gen_cards_dict(zt3.cards_list)
legal_choice=MrGreed.gen_legal_choice("D",cards_dict,zt3.cards_list)
zt3.select_interact_cards(legal_choice)"""
"""cards_played,scores_stage,void_info_stage=zt3.public_info()
log(cards_played)
for i in range(14):
log("after stage %d: %s\n%s"%(i,scores_stage[i],void_info_stage[i]),end="")
input()"""
#log(zt3.pick_a_card())
#return
l=[zt3.pick_a_card() for i in range(50)]
log("%d %d %d"%(len([i[0] for i in l if i[0]=="H"]),len([i[0] for i in l if i[0]=="C"]),len([i[0] for i in l if i[0]=="S"])))
def example_SQ():
zt3=MrZeroTree(room=255,place=3,name='zerotree3',mcts_b=10,mcts_k=2,sample_b=-1,sample_k=-2)
zt3.cards_list=["HQ","HJ","H8","H7","SA","S6","S5","S4","S3","CQ","CJ","D3"]
zt3.cards_on_table=[1,"S7","SJ"]
zt3.history=[[1,"C9","C7","C4","H9"],]
zt3.scores=[[],["H9"],[],[]]
"""cards_dict=MrGreed.gen_cards_dict(zt3.cards_list)
legal_choice=MrGreed.gen_legal_choice("S",cards_dict,zt3.cards_list)
zt3.select_interact_cards(legal_choice)"""
cards_played,scores_stage,void_info_stage=zt3.public_info()
log(cards_played)
for i in range(6):
log("after stage %d: %s\n%s"%(i,scores_stage[i],void_info_stage[i]),end="")
input()
return
log(zt3.pick_a_card())
def example_SQ2():
zt3=MrZeroTree(room=255,place=3,name='zerotree3',mcts_b=10,mcts_k=2,sample_b=-1,sample_k=-2)
zt3.cards_list=["HQ","HJ","H8","H7","SA","S3","CJ","C4","CQ","D3","D10"]
zt3.cards_on_table=[2,"S6"]
zt3.history=[[0,"S7","SK","S10","S8"],[1,"C2","C9","C8","C7"]]
zt3.scores=[[],[],[],[]]
log(zt3.pick_a_card())
if __name__=="__main__":
example_DJ()
#example_SQ()
#example_SQ2()
#burdens()
#irrelevant_cards()
"""
# C4="2345": 65.7(4.8)
# C2="23456": 67.5(4.8)
# C ="234567": 69.4(4.5)
# C3="2345678": 61.5(4.6)
# C5="23456789": 65.8(4.7)
# C6="34567": 63.37(4.65)
# C ="234567": 69.4(4.5)
# D2="2345": 66.5(4.5)
# D4="23456": 72.0(4.5)
# D ="234567": 68.0(4.6)
# D5="2345678": 67.3(4.5)
# D3="23456789": 61.8(4.7)
# D8="": 63.9(4.5)
# D7="3456": 69.24 4.62
# D6="34567": 68.46 4.55
# D ="234567": 68.0(4.6)
# H=C+F+following: 68.8(4.6)
#if thisuit=="A" and suit=="A":
# return True
# N=DxH: 61.3(4.4)
#if thisuit=="A":
# if suit=="A":
# return True
# elif choice[1] not in "234567":
# return True
# H=if True: 64.8(4.7)
# G=C+D+F: 59.2(4.5) 60.2(4.6) WHY?
# K=C+F: 66.4(4.6)
# L=C+D: 75.6(4.8)
# L2=C+D4: 66.6(4.7)
# M=D+F: 64.4(4.6)
# 修正贴牌的想法不错，但是和其他修正相容性不好。
# 但相容性不好可能是retrict_flag导致的，这还有待研究
# F: 68.4(4.7) 67.2(4.7)
#if suit!="A" and choice[0]!=suit:
# F2: 65.6(4.4)
#if thisuit!="A" and suit!="A" and choice[0]!=suit:
# F3: 65.3(4.8)
#if thisuit=="A" and suit!="A" and choice[0]!=suit:
# J: 63.0(4.7)
#if thisuit=="A" and choice[1] not in "234567":
# if suit=="A":
# return True
# elif choice[0]!=suit:
# return True
# J2: 62.7(4.8) 65.6(4.7)
#if thisuit=="A":
# if suit=="A":
# return True
# elif choice[0]!=suit:
# return True"""