/
earthquake_train_2_hidden.py
124 lines (114 loc) · 3.58 KB
/
earthquake_train_2_hidden.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
import modules.inference as inference
import modules.models_factory as models_factory
import time
import os
from collections import deque
num_of_layers = 5
num_of_types = 2
time_scale = 1e-3
time_log = time.time()
folder_name = (
"/your/path/to/experiments/" # replace this path with your path to store results
+ str(time_log)
+ "/"
)
os.mkdir(folder_name)
model_factory = models_factory.ModlesFactory(
model_name="OEOT",
data_path="/your/path/to/data_earthquake/", # replace this path with your path to the data folder
time_scale=time_scale,
)
model_factory.load_evidences("train", data_ratio=1, num_of_types=num_of_types)
model_factory.load_evidences("dev", data_ratio=1, num_of_types=num_of_types)
kernel_params_dq = deque(
[
0.6786821005630584,
0.22879353021881102,
0.01826475745966238,
2.6613670203623108,
0.7377663659680163,
1.6773569754135238,
15.260062664394468,
0.3586926388156386,
0.002648899490738625,
1.6628332918788664,
0.2861241794163845,
0.002226073892788341,
]
)
virtual_kernel_params_dq = deque(
[
0.16381098299846472,
0.2074614343203382,
27.14575410333683,
0.31021543478306324,
0.6438698904968448,
3.2479694005722077,
0.07731751275895048,
0.8377089962488387,
38.197144500976464,
0.723986915654927,
0.33995758482280797,
0.0736900266980569,
]
)
kernel_type = "GammaKernel"
var_ids = [1, 2, 3]
virtual_var_ids = [0, 1, 2, 3]
prior_inference_DPPLayers = models_factory.dpp_layers_factory(
model_name="OEOT",
evidences=model_factory.evidences,
valid_evidences=model_factory.valid_evidences,
num_of_types=num_of_types,
num_of_layers=num_of_layers,
kernel_type=kernel_type,
kernel_params_dq=kernel_params_dq,
virtual_kernel_params_dq=virtual_kernel_params_dq,
var_ids=var_ids,
virtual_var_ids=virtual_var_ids,
base_rate_init=0.08,
valid_base_rate_init=0.08,
virtual_base_rate_init=0.08,
)
dpp_layers_events_factory = model_factory.build_model_factory(
data_type="train", dpp_layers=prior_inference_DPPLayers,
)
valid_dpp_layers_events_factory = model_factory.build_model_factory(
data_type="dev", dpp_layers=prior_inference_DPPLayers,
)
print(f" num of events in evidence = {sum(model_factory.num_of_events_times_evidence)}")
print(
f" num of events in valid evidence = {sum(model_factory.valid_num_of_events_times_evidence)}"
)
inference_EM = inference.EM(
dpp_layers=prior_inference_DPPLayers,
num_of_epochs=4000,
sample_size=64,
sample_intervals=10,
valid_sample_size=10,
valid_sample_intervals=10,
log_folder_name=folder_name,
initial_burn_in_steps=0,
burn_in_steps=0,
valid_burn_in_steps=1,
evidences=model_factory.evidences,
valid_evidences=model_factory.valid_evidences,
end_time_tuple=model_factory.end_time_tuple,
valid_end_time_tuple=model_factory.valid_end_time_tuple,
batch_size=-1,
valid_batch_size=-1,
dpp_layers_events_factory=dpp_layers_events_factory,
valid_dpp_layers_events_factory=valid_dpp_layers_events_factory,
keep_last_samples=True,
opt=True,
num_of_events_evidence_each_example=model_factory.num_of_events_times_evidence,
valid_num_of_events_evidence_each_example=model_factory.valid_num_of_events_times_evidence,
time_scale=time_scale,
alpha=0.01,
virtual_alpha=0.1,
optimization_method="adam",
track_period=2,
track_validation=False,
fix_kernel_params=False,
)
inference_EM.iterations_withoutparents()