-
Notifications
You must be signed in to change notification settings - Fork 0
/
stan.py
176 lines (151 loc) · 5.59 KB
/
stan.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
from datetime import datetime
from copy import copy,deepcopy
import sys
import numpy as np
from scipy.stats import bernoulli
import matplotlib.pyplot as plt
import pystan
from sklearn.grid_search import GridSearchCV
from sklearn.cross_validation import ShuffleSplit
from sklearn.base import BaseEstimator
from sklearn.metrics import roc_curve,auc
from trillion import load_components,load_odorants_tests_results
with open('trillion.stan') as f:
code = f.read()
class StanModel2(BaseEstimator):
def __init__(self,**kwargs):
for key,value in kwargs.items():
setattr(self,key,value)
def set_model(self,code):
self.model = pystan.StanModel(model_code=code)
#def __deepcopy__(self):
# return self
#self.model.__deepcopy__ = __deepcopy__
def set_data(self,data):
self.data = data
for key,value in data.items():
setattr(self,key,value)
'''
def get_model(self):
if type(self.model) is dict:
print self.model
print self.model.values()[1][0]
print self.model.values()[1][0].values()[1][0]
print self.model.values()[1][0].values()[1][0].values()[1][0]
for value in self.model.values():
if type(value) is list:
if value[0].__class__.__name__ == 'StanModel':
return value[0]
else:
return self.model
'''
def optimize(self,X,y):
for key in self.data.keys():
self.data[key] = getattr(self,key)
self.data.update({'n_obs':X.shape[0],
'subject_ids':X[:,0],
'test_ids':X[:,1],
'correct':y})
self.best = self.model.optimizing(data=self.data)
def get_params(self,deep=False):
return self.__dict__
def fit(self,X,y):
self.optimize(X,y)
for key,value in self.best.items():
setattr(self,key,value)
def transform(self,X,y=None,**fit_params):
return X
def predict(self,X):
"""X should be an Nx2 array, with subject ids in the first column
and test ids in the second column.
"""
subjects = X[:,0]
tests = X[:,1]
n_samples = X.shape[0]
prediction = np.zeros(n_samples)
for i in range(n_samples):
difficulty = self.r_ta[tests[i]-1]
discriminability = self.r_a[tests[i]-1]
skill = self.r_s[subjects[i]-1]
prediction[i] = 0.333 + 0.667/(1+np.exp(-discriminability*(skill-difficulty)));
return prediction
def score(self,X,y):
prediction = self.predict(X)
return np.log(bernoulli.pmf(y,prediction)).sum()
def hist(self):
p_correct_hit = self.p_correct[np.where(self.correct==1)]
p_correct_miss = self.p_correct[np.where(self.correct==0)]
plt.hist(p_correct_hit)
plt.hist(p_correct_miss)
plt.show()
def roc(self):
fpr, tpr, _ = roc_curve(self.correct,self.p_correct)
roc_auc = auc(fpr, tpr)
plt.plot(fpr,tpr)
plt.show()
return roc_auc
THEN = None
def tic():
global THEN
THEN = datetime.now()
def toc(activity='Something'):
now = datetime.now()
delta = now - THEN
seconds = delta.days*24*3600 + delta.seconds + delta.microseconds/1e6
print('%s took %.3g seconds' % (activity,seconds))
def get_posterior_means(fit):
means = {}
x = fit.extract()
for key,value in x.items()[:-1]:
means[key] = value.mean(axis=0)
return means
def make_data(search_data=None,test_size=0):
components = load_components()
odorants,tests,results = load_odorants_tests_results(components)
n_subjects = len(set([x.subject_id for x in results]))
n_tests = len(tests)
n_obs = int(len(results)*(1-test_size))
n_molecules = 128
mixtures1 = np.zeros((n_tests,n_molecules))
mixtures2 = np.zeros((n_tests,n_molecules))
for test_id,test in tests.items():
single = [components.index(molecule) for molecule in tests[test_id].single.components]
for molecule in single:
mixtures1[test_id-1,molecule] = 1.0/test.N
double = [components.index(molecule) for molecule in tests[test_id].double.components]
for molecule in double:
mixtures2[test_id-1,molecule] = 1.0/test.N
tests = np.array([result.test.id for result in results])
subjects = np.array([result.subject_id for result in results])
correct = np.array([int(result.correct) for result in results])
data = locals()
for to_del in ['components','odorants','results','single','double',
'test_id','test','molecule']:
del data[to_del]
if 'search_data' in data:
del data['search_data']
del data['test_size']
if search_data:
data.update({key:value[0] for key,value in search_data.items()})
return data
if __name__ == '__main__':
estimator = StanModel2()
estimator.set_model(code)
search_data = {'sigma_s':[0.3,1.0,3.0]}
test_size = 0.1
data = make_data(search_data=search_data,test_size=test_size)
estimator.set_data(data)
cv = ShuffleSplit(data['n_obs'],test_size=test_size)
grid = GridSearchCV(estimator,search_data,cv=cv)
y = data['correct']
X = np.vstack((data['subject_ids'],data['test_ids'])).transpose()
grid.fit(X,y)
'''
tic()
model = pystan.StanModel(model_code=code)
toc('Model compilation')
data = make_data()
best = model.optimizing(data=data)
fit = model.sampling(data=data)
means = fit
'''