-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcompute_stats.py
271 lines (250 loc) · 9.56 KB
/
compute_stats.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
from __future__ import division
import sys,os.path,math,random
import scipy.stats as stats_
import numpy
import sys,os.path
import chess
import chess.pgn
def create_digest(test):
pgn_file=test+".pgn"
if not os.path.exists(pgn_file):
return False
digest_file=test+".digest"
if os.path.exists(digest_file):
return False
print("Creating %s" % digest_file)
pgn = open(pgn_file)
n=0
games={}
game=None
digest=open(digest_file,"w")
first_white=None
while True:
header=chess.pgn.read_headers(pgn)
if not header:
break
n+=1
if n%100==0:
sys.stdout.write("%d,%d \r" % (n,len(games.keys())))
sys.stdout.flush()
fen=header.get('FEN')
if not fen: # 8 book test
break
result=header['Result']
white=header['White']
if not first_white:
first_white=white
if fen in games.keys():
result_last,white_last=games[fen]
if result!="*" and result_last!="*" and white!=white_last:
if white==first_white:
digest.write(fen+" "+result+" "+result_last+"\n")
else:
digest.write(fen+" "+result_last+" "+result+"\n")
del games[fen]
else:
games[fen]=(result,white)
digest.close()
pgn.close()
return True
def to_score(result):
if result=="1-0":
return 2
elif result=="1/2-1/2":
return 1
elif result=="0-1":
return 0
def score_to_elo(s):
return -400*math.log(1/s-1)/math.log(10)
def elo_to_score(e):
return 1/(1+10**(-e/400))
def fe(a,b):
epsilon=1e-6
return abs(a-b)<epsilon
class compute:
def __init__(self,test,randomize=False):
self.calc=[]
self.stats={}
self.test=test
self.randomize=randomize
ret=self.assemble()
if not ret:
return
self.marginals()
self.chi2()
self.trinomial()
self.pentanomial()
self.ldw()
self.scores()
self.elos()
self.variances()
self.correlation()
self.variance_pentanomial()
def assemble(self):
self.calc.extend(['name','all'])
digest_file=test+".digest"
if not os.path.exists(digest_file) and not create_digest(test):
return
self.stats['name']=test
digest=open(digest_file,"r")
lines=digest.readlines()
collect1=[]
collect2=[]
for l in lines:
l=l.split()
result1=l[-2]
result2=l[-1]
collect1.append(result1)
collect2.append(result2)
if self.randomize:
random.shuffle(collect1)
random.shuffle(collect2)
self.stats['all']={}
for i in range(0,3):
for j in range(0,3):
self.stats['all'][(i,j)]=0
if len(collect1)==0:
return False
for u in range(0,len(collect1)):
result1=collect1[u]
result2=collect2[u]
score1=to_score(result1)
score2=to_score(result2)
self.stats['all'][(score1,2-score2)]+=1
return True
def marginals(self):
self.calc.extend(['white','black','N2'])
self.stats['white']=3*[0]
self.stats['black']=3*[0]
self.stats['N2']=0
for i in range(0,3):
for j in range(0,3):
s=self.stats['all'][(i,j)]
self.stats['white'][i]+=s
self.stats['black'][j]+=s
self.stats['N2']+=s
assert(sum(self.stats['white'])==self.stats['N2'])
assert(sum(self.stats['black'])==self.stats['N2'])
assert(sum(self.stats['all'].values())==self.stats['N2'])
def chi2(self):
self.calc.extend(['chi2','p_chi2'])
self.stats['chi2']=0
e_s=0
for i in range(0,3):
for j in range(0,3):
e=(self.stats['white'][i]*self.stats['black'][j])/self.stats['N2']
e_s+=e
o=self.stats['all'][(i,j)]
if e<5:
del self.stats['chi2']
return
self.stats['chi2']+=(e-o)**2/e
self.stats['p_chi2']=1-stats_.chi2.cdf(self.stats['chi2'],(3-1)*(3-1))
assert(fe(e_s,self.stats['N2']))
def trinomial(self):
self.calc.append('trinomial')
self.stats['trinomial']=3*[0]
for i in range(0,3):
self.stats['trinomial'][i]=self.stats['white'][i]+self.stats['black'][i]
assert(sum(self.stats['trinomial'])==2*self.stats['N2'])
def pentanomial(self):
self.calc.append('pentanomial')
self.stats['pentanomial']=5*[0]
for i in range(0,3):
for j in range(0,3):
self.stats['pentanomial'][i+j]+=self.stats['all'][(i,j)]
assert(sum(self.stats['pentanomial'])==self.stats['N2'])
def ldw(self):
self.calc.extend(['ldw','ldw_white','ldw_black','ldw_pentanomial'])
self.stats['ldw']=3*[0]
self.stats['ldw_white']=3*[0]
self.stats['ldw_black']=3*[0]
self.stats['ldw_pentanomial']=5*[0]
for i in range(0,3):
self.stats['ldw'][i]=self.stats['trinomial'][i]/(2*self.stats['N2'])
self.stats['ldw_white'][i]=self.stats['white'][i]/self.stats['N2']
self.stats['ldw_black'][i]=self.stats['black'][i]/self.stats['N2']
for i in range(0,5):
self.stats['ldw_pentanomial'][i]=self.stats['pentanomial'][i]/self.stats['N2']
assert(fe(sum(self.stats['ldw']),1))
assert(fe(sum(self.stats['ldw_white']),1))
assert(fe(sum(self.stats['ldw_black']),1))
assert(fe(sum(self.stats['ldw_pentanomial']),1))
def scores(self):
self.calc.extend(['score','score_white','score_black'])
self.stats['score']=self.stats['ldw'][2]+self.stats['ldw'][1]/2
self.stats['score_white']=self.stats['ldw_white'][2]+self.stats['ldw_white'][1]/2
self.stats['score_black']=self.stats['ldw_black'][2]+self.stats['ldw_black'][1]/2
assert(fe(self.stats['score'],(self.stats['score_white']+self.stats['score_black'])/2))
def elos(self):
self.calc.extend(['elo','elo_white','elo_black'])
self.stats['elo']=score_to_elo(self.stats['score'])
self.stats['elo_white']=score_to_elo(self.stats['score_white'])
self.stats['elo_black']=score_to_elo(self.stats['score_black'])
assert(fe(elo_to_score(self.stats['elo']),self.stats['score']))
assert(fe(elo_to_score(self.stats['elo_white']),self.stats['score_white']))
assert(fe(elo_to_score(self.stats['elo_black']),self.stats['score_black']))
def variances(self):
self.calc.extend(['variance','variance_white','variance_black'])
self.stats['variance']=(self.stats['ldw'][2]+self.stats['ldw'][1]/4-self.stats['score']**2)
self.stats['variance_white']=(self.stats['ldw_white'][2]+self.stats['ldw_white'][1]/4-self.stats['score_white']**2)
self.stats['variance_black']=(self.stats['ldw_black'][2]+self.stats['ldw_black'][1]/4-self.stats['score_black']**2)
variance_alt=(self.stats['variance_white']+self.stats['variance_black'])/2+(self.stats['score_white']-self.stats['score_black'])**2/4
assert(fe(self.stats['variance'],variance_alt))
def correlation(self):
self.calc.extend(['covariance','correlation','p_correlation'])
m2=0
for i in range(0,3):
for j in range(0,3):
m2+=(i/2)*(j/2)*self.stats['all'][(i,j)]/self.stats['N2']
self.stats['covariance']=m2-self.stats['score_white']*self.stats['score_black']
try:
self.stats['correlation']=self.stats['covariance']/(self.stats['variance_white']*self.stats['variance_black'])**.5
except ZeroDivisionError:
return
# t-transformation
if self.stats['N2']>=2:
c=self.stats['correlation']
N2=self.stats['N2']
t=c*((N2-2)/(1-c*c))**.5
self.stats['p_correlation']=2*(1-stats_.t.cdf(abs(t),N2-2))
# fisher transformation
if self.stats['N2']>3:
z=numpy.arctanh(self.stats['correlation'])*(self.stats['N2']-3)**.5
self.stats['p_correlation2']=2*(1-stats_.norm.cdf(abs(z)))
def variance_pentanomial(self):
self.calc.extend(['variance_pentanomial','variance_ratio'])
m2=0
m1=0
for i in range(0,5):
m2+=self.stats['ldw_pentanomial'][i]*(i/2)*(i/2)
m1+=self.stats['ldw_pentanomial'][i]*(i/2)
self.stats['variance_pentanomial']=(m2-(2*self.stats['score'])**2)/2
try:
self.stats['variance_ratio']=self.stats['variance_pentanomial']/self.stats['variance']
except ZeroDivisionError:
pass
if self.stats.get('covariance')!=None:
variance_alt=(self.stats['variance_white']+self.stats['variance_black']+2*self.stats['covariance'])/2
assert(fe(self.stats['variance_pentanomial'],variance_alt))
def __str__(self):
s=''
for k in self.calc:
v=self.stats.get(k)
if v!=None:
s+="%s=%s\n" % (k,repr(v))
return s
if __name__=='__main__':
work=[]
for l in sys.argv[1:]:
test=l.split(".")[0]
work.append(test)
work=list(set(work))
for test in work:
c=compute(test,randomize=False)
if os.path.exists(test+".pgn"):
stats_file=test+".stats"
f=open(stats_file,"w")
print("Writing %s" % stats_file)
f.write(str(c))
f.close()