-
Notifications
You must be signed in to change notification settings - Fork 0
/
normal.py
157 lines (138 loc) · 5.52 KB
/
normal.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
# -*- coding: utf-8 -*-
"""
Created on Fri May 10 12:13:01 2019
@author: FaiHuntrakool
"""
import pandas as pd
import numpy as np
n_cov=2
sample=1000
n_bin=10
mat=np.random.normal(1,2,(sample,n_cov))
solver='SA'
def create_bins_array(initial_mat,no_bins,n_cov):
bmat=np.zeros((len(initial_mat),len(initial_mat[0])))
for b in range(n_cov):
data=initial_mat[:,b]
bins=np.linspace(data.min(),data.max(),no_bins)
digitized=np.digitize(data,bins)
bmat[:,b]=digitized
return bmat
bins=create_bins_array(mat,n_bin,n_cov)
import matplotlib.pyplot as plt
for i in range(n_cov):
cov=pd.DataFrame(bins[:,i]).rename({0:'bins'},axis=1)
plt.title('Covariate '+str(i+1))
plt.hist(list(cov['bins']),density=False,histtype='bar',color='black')
plt.xticks(range(0,n_bin))
plt.show()
###########model formulation ################################
"""check duayyy"""
from pyqubo import Array
s=Array.create('s',shape=sample,vartype='SPIN')
def objective_function(s,bins,n_bin):
H=0
for col in range(len(mat[0])):
for b in range(1,n_bin+1):
M=0
for val,ss in zip(bins[:,col],s):
if val==b:
M=M+ss
H=H+(M-0)**2
return H-0
H=objective_function(s,bins,n_bin)
model=H.compile()
qubo,offset=model.to_qubo()
bqm=model.to_dimod_bqm()###check
################# show qubo matrix ##########################
import re
qmat=np.zeros((sample,sample))
for key,val in qubo.items():
x=int(int(re.search(r'\d+', key[0]).group()))
y=int(int(re.search(r'\d+', key[1]).group()))
qmat[min(x,y)][max(x,y)]=val
import pandas as pd
q=pd.DataFrame(qmat)
#############################################################
from dwave.system.samplers import DWaveSampler
from dwave.system.composites import EmbeddingComposite
import neal
import timeit
start = timeit.default_timer()
if solver=='Exact':
import dimod
sampler=dimod.ExactSolver()
response=sampler.sample(bqm)
else:
#sampler=EmbeddingComposite(DWaveSampler())
sampler=neal.SimulatedAnnealingSampler()
response=sampler.sample(bqm,num_reads=100)
stop = timeit.default_timer()
print('time:= '+str(stop-start)+'s')
###########Embedding visualization #########################
import matplotlib.pyplot as plt
for i in range(n_cov):
cov=pd.DataFrame(bins[:,i]).rename({0:'bins'},axis=1)
plt.title('Covariate '+str(i+1))
plt.hist(list(cov['bins']),density=False,histtype='bar',color='black')
plt.xticks(range(0,n_bin))
plt.show()
"""evaluation"""
###########result####################################
from collections import defaultdict
result=defaultdict(list)
for r in response.data(['sample', 'energy']):
result[r.energy].append(r.sample)
res=pd.DataFrame()
min_energy=min(result.keys())
x=result[min_energy]
res=pd.DataFrame.from_dict(x[0],'index').transpose()
###########visualization####################################
group=pd.DataFrame(res.iloc[0,:]).reset_index().drop('index',axis=1).rename({0:'group'},axis=1)
colors=['red','blue']
for i in range(n_cov):
cov=pd.DataFrame(bins[:,i]).rename({0:'bins'},axis=1)
group=pd.DataFrame(res.iloc[0,:]).reset_index().drop('index',axis=1).rename({0:'group'},axis=1)
df=pd.concat([cov, group], axis=1)
plt.title('Covariate '+str(i+1))
plt.hist([list(df[df['group']==0]['bins']),list(df[df['group']==1]['bins'])],density=False,histtype='bar',color=colors,bins=n_bin)
plt.show()
##########Variance, mean (Compare with randomization technique)##################################
dcov=dict()
for i in range(n_cov):
cov=pd.DataFrame(mat[:,i]).rename({0:'bins'},axis=1)
df=pd.concat([cov, group], axis=1)
dcov[i]={'mean':df['bins'].mean(),
'variance':df['bins'].var(),
'mean_control':df[df['group']==0]['bins'].mean(),
'variance_control':df[df['group']==0]['bins'].var(),
'mean_treatment':df[df['group']==1]['bins'].mean(),
'variance_treatment':df[df['group']==1]['bins'].var()}
#randomization
group2=pd.DataFrame(np.random.randint(0,2,(sample,1))).rename({0:'group'},axis=1)
colors=['red','blue']
for i in range(n_cov):
cov=pd.DataFrame(bins[:,i]).rename({0:'bins'},axis=1)
df=pd.concat([cov, group2], axis=1)
plt.title('Covariate '+str(i+1))
plt.hist([list(df[df['group']==0]['bins']),list(df[df['group']==1]['bins'])],density=False,histtype='bar',color=colors,bins=n_bin)
plt.show()
dcovrand=dict()
for i in range(n_cov):
cov=pd.DataFrame(mat[:,i]).rename({0:'bins'},axis=1)
df=pd.concat([cov, group2], axis=1)
dcovrand[i]={'mean':df['bins'].mean(),
'variance':df['bins'].var(),
'mean_control':df[df['group']==0]['bins'].mean(),
'variance_control':df[df['group']==0]['bins'].var(),
'mean_treatment':df[df['group']==1]['bins'].mean(),
'variance_treatment':df[df['group']==1]['bins'].var()}
#print(pd.DataFrame.from_dict(dcovrand))
df=pd.DataFrame.from_dict(dcov).transpose()
df['different of means-optimization']=(df['mean_treatment']-df['mean_control']).abs()
df_rand=pd.DataFrame.from_dict(dcovrand).transpose()
df_rand['different of means-before optimization']=(df_rand['mean_treatment']-df_rand['mean_control']).abs()
result2 = pd.concat([df['different of means-optimization'], df_rand['different of means-before optimization']], axis=1, sort=False)
print(result2.mean())
print('solution')
print(min(result.keys()))