-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathplot_auc_simdata.py
251 lines (229 loc) · 13.2 KB
/
plot_auc_simdata.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
import numpy as np
import matplotlib.pyplot as plt
import scipy
import pandas as pd
import seaborn as sns
from utilities import *
from BNN_model import *
# Initialize random number generator
RANDOM_SEED = 42
np.random.seed(RANDOM_SEED)
rng = np.random.default_rng(RANDOM_SEED)
print(f"Running on PyMC v{pm.__version__}")
#SAVEDIR = "/data/evenmm/plots/"
SAVEDIR = "./plots/Bayesian_estimates_simdata_BNN/"
script_index = 1
# Settings
if int(script_index % 3) == 0:
true_sigma_obs = 0
elif int(script_index % 3) == 1:
true_sigma_obs = 1
elif int(script_index % 3) == 2:
true_sigma_obs = 2.5
if script_index >= 3:
RANDOM_EFFECTS = True
else:
RANDOM_EFFECTS = False
RANDOM_EFFECTS_TEST = False
model_name = "BNN"
N_patients = 150
psi_prior="lognormal"
WEIGHT_PRIOR = "Student_out" #"Horseshoe" # "Student_out" #"symmetry_fix" #"iso_normal" "Student_out"
N_samples = 10_000
N_tuning = 10_000
ADADELTA = True
target_accept = 0.99
CI_with_obs_noise = True
PLOT_RESISTANT = True
FUNNEL_REPARAMETRIZATION = False
MODEL_RANDOM_EFFECTS = True
N_HIDDEN = 2
P = 5 # Number of covariates
P0 = int(P / 2) # A guess of the true number of nonzero parameters is needed for defining the global shrinkage parameter
true_omega = np.array([0.10,0.05,0.20])
M_number_of_measurements =7
y_resolution = 80 # Number of timepoints to evaluate the posterior of y in
true_omega_for_psi = 0.1
max_time = 1200 #3000 #1500
days_between_measurements = int(max_time/M_number_of_measurements)
measurement_times = days_between_measurements * np.linspace(0, M_number_of_measurements, M_number_of_measurements)
evaluation_time = measurement_times[:4][-1] + 1
print(evaluation_time)
treatment_history = np.array([Treatment(start=0, end=measurement_times[-1], id=1)])
name = "simdata_"+model_name+"_"+str(script_index)+"_M_"+str(M_number_of_measurements)+"_P_"+str(P)+"_N_pax_"+str(N_patients)+"_N_sampl_"+str(N_samples)+"_N_tune_"+str(N_tuning)+"_FUNNEL_"+str(FUNNEL_REPARAMETRIZATION)+"_RNDM_EFFECTS_"+str(RANDOM_EFFECTS)+"_WT_PRIOR_"+str(WEIGHT_PRIOR+"_N_HIDDN_"+str(N_HIDDEN))
N_patients_test = 50
recur_or_not_BNN = [
[1.,0.,0.,1.,0.,0.,0.,0.,1.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,1.,0.
,0.,1.,0.,0.,1.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,1.,0.,0.,0.,0.,1.,0.,0.,0.
,1.,0.],
[1.,0.,0.,1.,0.,0.,1.,0.,1.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,1.,0.
,0.,1.,0.,0.,1.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,1.,0.,0.,0.,0.,1.,0.,0.,0.
,1.,0.],
[1.,0.,0.,1.,0.,0.,1.,0.,1.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,1.,0.
,0.,1.,0.,0.,1.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,1.,0.,0.,0.,1.,1.,0.,0.,0.
,1.,0.],
[1.,0.,0.,1.,0.,0.,0.,0.,1.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,1.,0.
,0.,1.,0.,0.,1.,0.,0.,0.,0.,0.,0.,0.,0.,0.,0.,1.,0.,0.,0.,0.,1.,0.,0.,0.
,1.,0.],
[1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0,
1, 0],
[1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0,
1, 0,]
]
p_recs_BNN = np.array([[1.00000e+00,0.00000e+00,5.19250e-02,1.00000e+00,5.00000e-05,0.00000e+00
,4.05575e-01,2.00000e-04,7.45575e-01,0.00000e+00,2.47500e-03,0.00000e+00
,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00
,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,1.00000e+00,9.59275e-01
,0.00000e+00,9.74175e-01,0.00000e+00,0.00000e+00,9.99825e-01,0.00000e+00
,6.75000e-04,0.00000e+00,2.50000e-05,0.00000e+00,0.00000e+00,6.13250e-02
,0.00000e+00,0.00000e+00,0.00000e+00,1.00000e+00,0.00000e+00,0.00000e+00
,0.00000e+00,1.50000e-04,9.42375e-01,0.00000e+00,0.00000e+00,0.00000e+00
,1.00000e+00,0.00000e+00],
[1.00000e+00,0.00000e+00,1.77750e-02,1.00000e+00,0.00000e+00,0.00000e+00
,2.96350e-01,2.50000e-05,6.81375e-01,0.00000e+00,2.25000e-04,0.00000e+00
,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00
,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,1.00000e+00,9.39000e-01
,0.00000e+00,9.65450e-01,0.00000e+00,0.00000e+00,9.99925e-01,0.00000e+00
,5.00000e-05,0.00000e+00,2.50000e-05,0.00000e+00,0.00000e+00,2.48000e-02
,0.00000e+00,0.00000e+00,0.00000e+00,1.00000e+00,0.00000e+00,0.00000e+00
,0.00000e+00,0.00000e+00,9.55525e-01,0.00000e+00,0.00000e+00,0.00000e+00
,1.00000e+00,0.00000e+00],
[1.00000e+00,0.00000e+00,3.31500e-02,1.00000e+00,2.75000e-04,0.00000e+00
,2.98825e-01,8.25000e-04,5.79400e-01,0.00000e+00,2.82500e-03,0.00000e+00
,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00
,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,1.00000e+00,8.51800e-01
,0.00000e+00,8.87375e-01,0.00000e+00,0.00000e+00,9.98425e-01,0.00000e+00
,1.20000e-03,0.00000e+00,6.75000e-04,0.00000e+00,0.00000e+00,5.97250e-02
,0.00000e+00,0.00000e+00,0.00000e+00,1.00000e+00,2.50000e-05,0.00000e+00
,0.00000e+00,1.32500e-03,8.75850e-01,0.00000e+00,0.00000e+00,0.00000e+00
,9.99875e-01,0.00000e+00],
[9.99875e-01,0.00000e+00,2.30250e-01,9.99225e-01,1.36500e-02,0.00000e+00
,4.30125e-01,2.60750e-02,6.46450e-01,0.00000e+00,7.52250e-02,0.00000e+00
,0.00000e+00,0.00000e+00,4.75000e-03,0.00000e+00,0.00000e+00,0.00000e+00
,0.00000e+00,0.00000e+00,0.00000e+00,0.00000e+00,1.00000e+00,8.53925e-01
,0.00000e+00,8.64100e-01,0.00000e+00,0.00000e+00,9.83975e-01,0.00000e+00
,4.88250e-02,0.00000e+00,4.55000e-03,0.00000e+00,0.00000e+00,2.28150e-01
,2.00000e-04,0.00000e+00,0.00000e+00,9.99925e-01,1.00000e-04,0.00000e+00
,0.00000e+00,2.06250e-02,7.88850e-01,0.00000e+00,0.00000e+00,0.00000e+00
,9.83175e-01,0.00000e+00],
[9.99650e-01, 0.00000e+00, 1.89525e-01, 9.98775e-01, 1.33500e-02, 0.00000e+00,
3.88825e-01, 2.35000e-02, 5.92750e-01, 0.00000e+00, 5.76000e-02, 0.00000e+00,
0.00000e+00, 0.00000e+00, 3.47500e-03, 0.00000e+00, 0.00000e+00, 0.00000e+00,
0.00000e+00, 0.00000e+00, 7.50000e-05, 0.00000e+00, 1.00000e+00, 8.00275e-01,
0.00000e+00, 8.17250e-01, 0.00000e+00, 0.00000e+00, 9.73450e-01, 0.00000e+00,
3.60750e-02, 0.00000e+00, 5.22500e-03, 0.00000e+00, 0.00000e+00, 1.95325e-01,
1.50000e-04, 0.00000e+00, 0.00000e+00, 9.99650e-01, 9.25000e-04, 0.00000e+00,
0.00000e+00, 1.84750e-02, 7.66100e-01, 0.00000e+00, 0.00000e+00, 0.00000e+00,
9.77350e-01, 0.00000e+00],
[9.99600e-01, 0.00000e+00, 1.96975e-01, 9.97275e-01, 2.17500e-02, 0.00000e+00,
3.97275e-01, 2.98500e-02, 6.12750e-01, 0.00000e+00, 3.67500e-02, 0.00000e+00,
0.00000e+00, 0.00000e+00, 5.95000e-03, 0.00000e+00, 0.00000e+00, 5.00000e-05,
0.00000e+00, 0.00000e+00, 2.00000e-04, 0.00000e+00, 1.00000e+00, 7.83650e-01,
0.00000e+00, 7.90425e-01, 0.00000e+00, 0.00000e+00, 9.58925e-01, 0.00000e+00,
2.50000e-02, 0.00000e+00, 1.67000e-02, 0.00000e+00, 0.00000e+00, 1.40300e-01,
2.00000e-04, 0.00000e+00, 2.50000e-05, 9.99200e-01, 6.75000e-04, 0.00000e+00,
0.00000e+00, 3.85750e-02, 7.53050e-01, 0.00000e+00, 0.00000e+00, 0.00000e+00,
9.70200e-01, 0.00000e+00,]
])
recur_or_not_LIN = [
[1., 0., 0., 1., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0.,
0., 1., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 1., 1., 0., 0., 0.,
1., 0.],
[1., 0., 0., 1., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0.,
0., 1., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0.,
1., 0.],
[1., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0.,
0., 1., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0.,
1., 0.],
[1., 0., 0., 1., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0.,
0., 1., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 1., 1., 0., 0., 0.,
1., 0.],
[1., 0., 0., 1., 0., 0., 1., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0.,
0., 1., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0.,
1., 0.],
[1., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0.,
0., 1., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0.,
1., 0.]
]
p_recs_LIN = np.array([
[9.99175e-01, 0.00000e+00, 5.81800e-01, 9.39900e-01, 9.50000e-03, 0.00000e+00,
1.87775e-01, 1.79375e-01, 6.75025e-01, 0.00000e+00, 8.99500e-02, 1.25000e-04,
0.00000e+00, 0.00000e+00, 5.87750e-02, 0.00000e+00, 0.00000e+00, 1.00000e-04,
0.00000e+00, 0.00000e+00, 1.12500e-03, 4.75000e-04, 1.00000e+00, 9.34475e-01,
2.50000e-05, 8.79500e-01, 0.00000e+00, 0.00000e+00, 9.82225e-01, 0.00000e+00,
8.20000e-02, 0.00000e+00, 1.67000e-02, 0.00000e+00, 2.50000e-05, 2.66025e-01,
3.00000e-04, 0.00000e+00, 1.25000e-04, 9.94900e-01, 6.97500e-03, 0.00000e+00,
0.00000e+00, 7.74250e-02, 5.17300e-01, 0.00000e+00, 0.00000e+00, 0.00000e+00,
9.17650e-01, 0.00000e+00,],
[9.99700e-01, 0.00000e+00, 6.19600e-01, 9.59500e-01, 5.37500e-03, 0.00000e+00,
1.84250e-01, 1.59575e-01, 7.24525e-01, 0.00000e+00, 8.30750e-02, 5.00000e-05,
0.00000e+00, 0.00000e+00, 5.20250e-02, 0.00000e+00, 0.00000e+00, 0.00000e+00,
0.00000e+00, 0.00000e+00, 4.00000e-04, 2.50000e-05, 1.00000e+00, 9.58650e-01,
0.00000e+00, 9.15225e-01, 0.00000e+00, 0.00000e+00, 9.92675e-01, 0.00000e+00,
7.89250e-02, 0.00000e+00, 8.57500e-03, 0.00000e+00, 0.00000e+00, 2.69350e-01,
1.00000e-04, 0.00000e+00, 5.00000e-05, 9.97825e-01, 2.25000e-03, 0.00000e+00,
0.00000e+00, 6.46500e-02, 5.55475e-01, 0.00000e+00, 0.00000e+00, 0.00000e+00,
9.37575e-01, 0.00000e+00,],
[9.99750e-01, 0.00000e+00, 6.52225e-01, 9.66275e-01, 7.00000e-03, 0.00000e+00,
1.91450e-01, 1.79550e-01, 7.28450e-01, 0.00000e+00, 7.64750e-02, 2.50000e-05,
0.00000e+00, 0.00000e+00, 4.55750e-02, 0.00000e+00, 0.00000e+00, 0.00000e+00,
0.00000e+00, 0.00000e+00, 4.00000e-04, 2.50000e-05, 1.00000e+00, 9.63475e-01,
0.00000e+00, 9.21650e-01, 0.00000e+00, 0.00000e+00, 9.95000e-01, 0.00000e+00,
7.29000e-02, 0.00000e+00, 8.00000e-03, 0.00000e+00, 0.00000e+00, 2.56025e-01,
1.25000e-04, 0.00000e+00, 1.00000e-04, 9.98675e-01, 3.55000e-03, 0.00000e+00,
0.00000e+00, 7.01000e-02, 6.09500e-01, 0.00000e+00, 0.00000e+00, 0.00000e+00,
9.43975e-01, 0.00000e+00,],
[1.00000e+00, 0.00000e+00, 5.81725e-01, 9.89125e-01, 1.35000e-03, 0.00000e+00,
1.12675e-01, 1.16075e-01, 7.28225e-01, 0.00000e+00, 2.96000e-02, 2.50000e-05,
0.00000e+00, 0.00000e+00, 2.08000e-02, 0.00000e+00, 0.00000e+00, 0.00000e+00,
0.00000e+00, 0.00000e+00, 7.50000e-05, 2.50000e-05, 1.00000e+00, 9.74750e-01,
0.00000e+00, 9.46425e-01, 0.00000e+00, 0.00000e+00, 9.97725e-01, 0.00000e+00,
2.30500e-02, 0.00000e+00, 3.47500e-03, 0.00000e+00, 0.00000e+00, 1.92150e-01,
0.00000e+00, 0.00000e+00, 0.00000e+00, 9.99825e-01, 1.42500e-03, 0.00000e+00,
0.00000e+00, 3.19000e-02, 5.33925e-01, 0.00000e+00, 0.00000e+00, 0.00000e+00,
9.78950e-01, 0.00000e+00,],
[1.00000e+00, 0.00000e+00, 6.28575e-01, 9.96000e-01, 4.50000e-04, 0.00000e+00,
1.01475e-01, 9.34500e-02, 7.89375e-01, 0.00000e+00, 2.32000e-02, 0.00000e+00,
0.00000e+00, 0.00000e+00, 1.48000e-02, 0.00000e+00, 0.00000e+00, 0.00000e+00,
0.00000e+00, 0.00000e+00, 0.00000e+00, 0.00000e+00, 1.00000e+00, 9.88100e-01,
0.00000e+00, 9.74425e-01, 0.00000e+00, 0.00000e+00, 9.99425e-01, 0.00000e+00,
1.77250e-02, 0.00000e+00, 1.60000e-03, 0.00000e+00, 0.00000e+00, 2.04900e-01,
0.00000e+00, 0.00000e+00, 0.00000e+00, 9.99950e-01, 3.00000e-04, 0.00000e+00,
0.00000e+00, 2.22750e-02, 5.74200e-01, 0.00000e+00, 0.00000e+00, 0.00000e+00,
9.89875e-01, 0.00000e+00,],
[1.00000e+00, 0.00000e+00, 7.00900e-01, 9.92800e-01, 5.00000e-04, 0.00000e+00,
8.87750e-02, 1.08375e-01, 8.12125e-01, 0.00000e+00, 3.29500e-02, 0.00000e+00,
0.00000e+00, 0.00000e+00, 2.14500e-02, 0.00000e+00, 0.00000e+00, 0.00000e+00,
0.00000e+00, 0.00000e+00, 0.00000e+00, 0.00000e+00, 1.00000e+00, 9.92275e-01,
0.00000e+00, 9.81775e-01, 0.00000e+00, 0.00000e+00, 9.99600e-01, 0.00000e+00,
2.54500e-02, 0.00000e+00, 1.20000e-03, 0.00000e+00, 0.00000e+00, 2.55350e-01,
0.00000e+00, 0.00000e+00, 0.00000e+00, 9.99925e-01, 2.50000e-04, 0.00000e+00,
0.00000e+00, 2.75000e-02, 5.86725e-01, 0.00000e+00, 0.00000e+00, 0.00000e+00,
9.89675e-01, 0.00000e+00,]
])
import sklearn.metrics as metrics
for ii in range(len(recur_or_not_BNN)):
fpr_BNN, tpr_BNN, threshold = metrics.roc_curve(recur_or_not_BNN[ii], p_recs_BNN[ii]) #(y_test, preds)
roc_auc_BNN = metrics.auc(fpr_BNN, tpr_BNN)
fpr_LIN, tpr_LIN, threshold = metrics.roc_curve(recur_or_not_LIN[ii], p_recs_LIN[ii]) #(y_test, preds)
roc_auc_LIN = metrics.auc(fpr_LIN, tpr_LIN)
print("threshold:\n", threshold)
print("fpr_BNN:\n", fpr_BNN)
print("tpr_BNN:\n", tpr_BNN)
print("roc_auc_BNN:\n", roc_auc_BNN)
print("\nfpr_BNN:\n", fpr_LIN)
print("tpr_BNN:\n", tpr_LIN)
print("roc_auc_BNN:\n", roc_auc_LIN)
#plt.title('Receiver Operating Characteristic')
plt.plot(fpr_LIN, tpr_LIN, color=plt.cm.viridis(0.3), label = 'Linear reg. (AUC = %0.2f)' % roc_auc_LIN)
plt.plot(fpr_BNN, tpr_BNN, color=plt.cm.viridis(0.7), label = 'BNN (AUC = %0.2f)' % roc_auc_BNN)
plt.legend(loc = 'lower right')
plt.plot([0,1], [0,1], color='grey', linestyle='--')
plt.xlim([0,1])
plt.ylim([0,1])
plt.ylabel('True Positive Rate')
plt.xlabel('False Positive Rate')
plt.savefig(SAVEDIR+"AUC_"+str(ii)+"_"+str(N_patients_test)+"_test_patients_"+name+".pdf")
plt.show()