-
Notifications
You must be signed in to change notification settings - Fork 10
Expand file tree
/
Copy pathmulti_recon_classif.py
More file actions
152 lines (93 loc) · 6.17 KB
/
multi_recon_classif.py
File metadata and controls
152 lines (93 loc) · 6.17 KB
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
import argparse
from util_functions.recon_classif_settings import get_multi_recon_classif_manager
"""
This script makes it possible to design and train multiple Deep Learning
models, each taking a subset of ECG leads as input and generating a full 12-lead
ECG as output. The reconstruction loss is given by the probability of
detecting specific clinical labels from the reconstructed signals.
"""
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('-device', '--device', type=str, default=None)
parser.add_argument('-input', '--input_leads', type=str, default=None)
parser.add_argument('-output', '--output_leads', type=str, default=None)
parser.add_argument('-dataset', '--dataset', type=str, default=None)
parser.add_argument('-detectset', '--detectset', type=str, default=None)
parser.add_argument('-data_size', '--data_size', type=str, default=None)
parser.add_argument('-use_residual', '--use_residual', type=str, default=None)
parser.add_argument('-epoch', '--epoch_num', type=int, default=None)
parser.add_argument('-batch', '--batch_size', type=int, default=None)
parser.add_argument('-ppercent', '--prioritize_percent', type=float, default=None)
parser.add_argument('-psize', '--prioritize_size', type=int, default=None)
parser.add_argument('-optim', '--optimizer', type=str, default=None)
parser.add_argument('-lr', '--learning_rate', type=float, default=None)
parser.add_argument('-mom', '--momentum', type=float, default=None)
parser.add_argument('-decay', '--weight_decay', type=float, default=None)
parser.add_argument('-nest', '--nesterov', type=str, default=None)
parser.add_argument('-alpha', '--alpha', type=float, default=None)
parser.add_argument('-parallel', '--parallel', type=str, default=None)
parser.add_argument('-format', '--plot_format', type=str, default='png')
parser.add_argument('-name', '--output_name', type=str, default=None)
parser.add_argument('-plot', '--plot', action='store_const', const=True, default=False)
parser.add_argument('-test', '--test', action='store_const', const=True, default=False)
parser.add_argument('-train', '--train', action='store_const', const=True, default=False)
parser.add_argument('-eval', '--eval', action='store_const', const=True, default=False)
parser.add_argument('-plot_train', '--plot_train', action='store_const', const=True, default=False)
parser.add_argument('-plot_test', '--plot_test', action='store_const', const=True, default=False)
parser.add_argument('-plot_random', '--plot_random', action='store_const', const=True, default=False)
parser.add_argument('-plot_error', '--plot_error', action='store_const', const=True, default=False)
parser.add_argument('-plot_sub_class', '--plot_sub_class', action='store_const', const=True, default=False)
parser.add_argument('-clinical_test', '--clinical_test', action='store_const', const=True, default=False)
args = vars(parser.parse_args())
train_ticks = [x * .05 / 10 for x in range(0, 11)]
valid_ticks = [x * .05 / 10 for x in range(0, 11)]
multi_manager, sub_classes = get_multi_recon_classif_manager(args)
sub_classes = []
if args['plot']:
print('Plot...')
multi_manager.load_train_stats()
multi_manager.load_valid_stats()
multi_manager.plot_train_stats(loss_ticks = train_ticks, plot_format = args['plot_format'])
multi_manager.plot_valid_stats(loss_ticks = valid_ticks, plot_format = args['plot_format'])
multi_manager.load_test_stats()
multi_manager.plot_test_stats(plot_sub_classes = sub_classes, plot_format = args['plot_format'])
multi_manager.load_model()
multi_manager.plot_error_example(plot_format = args['plot_format'])
multi_manager.plot_random_example(plot_format = args['plot_format'])
multi_manager.plot_sub_class_example(plot_sub_classes = sub_classes, plot_format = args['plot_format'])
else:
if args['train']:
print('Train...')
multi_manager.train()
if args['test']:
print('Test...')
multi_manager.test()
if args['eval']:
print('Evaluate...')
multi_manager.eval()
if args['plot_train']:
print('Plot train performance...')
multi_manager.load_train_stats()
multi_manager.load_valid_stats()
multi_manager.plot_train_stats(loss_ticks = train_ticks, plot_format = args['plot_format'])
multi_manager.plot_valid_stats(loss_ticks = valid_ticks, plot_format = args['plot_format'])
if args['plot_test']:
print('Plot test performance...')
multi_manager.load_test_stats()
multi_manager.plot_test_stats(plot_sub_classes = sub_classes, plot_format = args['plot_format'])
if args['plot_random']:
print('Plot random example...')
multi_manager.load_model()
multi_manager.plot_random_example(plot_format = args['plot_format'])
if args['plot_error']:
print('Plot error example...')
multi_manager.load_model()
multi_manager.plot_error_example(plot_format = args['plot_format'])
if args['plot_sub_class']:
print('Plot error examples...')
multi_manager.load_model()
multi_manager.plot_sub_class_example(plot_sub_classes = sub_classes, plot_format = args['plot_format'])
if args['clinical_test']:
print('Clinical test...')
multi_manager.load_model()
multi_manager.emulate_clinical_test(size=6000, thresholds = [0.123, 0.319, 0.509], sub_classes=['st_elevation_or_acute_infarct', 'prior_infarct', 'non_st_elevation_or_infarct'], sub_class_sizes=[0.5, 0.25, 0.25])