DeepECG-Reconstructor/process_dataclass.py at master · jingweili87/DeepECG-Reconstructor · GitHub

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

import argparse
import pathlib
from attr import dataclass
import numpy as np
from clean_dataclass import clean_dataclass
from analyze_dataclass import analyze_dataclass
from plot_functions.continuos_distribution import plot_violin_distribution
from util_functions.general import get_parent_folder
from util_functions.load_data_ids import *

def process_dataclass(data_class: str, clean: bool, analyze: bool, reset: bool):

    """
    This function makes it possible to (i) discern the elements of a given data_class
    between corrupted and cleaned elements and (ii) compute the conditional probability
    for the elements of the data_class to be associated with the different
    clinical labels used to describe the dataset

    The IDs of the corrupted/cleaned elements are stored in the folder ./../Data/Feature_map/Dataclass/data_class/'

    The statistical results are stored in the folder ./../Data/Analysis/Dataclass/data_class/Data/

    The figures describing the results are stored in the folder ./../Data/Analysis/Dataclass/data_class/Plot/

    """

    parent_folder = get_parent_folder()

    stats_folder = parent_folder + 'Analysis/Dataclass/' + data_class + '/Data/'
    plot_folder = parent_folder + 'Analysis/Dataclass/' + data_class + '/Plot/'

    pathlib.Path(plot_folder).mkdir(parents=True, exist_ok=True)

    class_ids = load_dataclass_ids(parent_folder, data_class)
    class_patient_ids = load_dataclass_patient_ids(parent_folder, data_class)

    print('Number of class data:', len(class_ids))

    print('Number of class individuals:', len(class_patient_ids))

    if clean:

        clean_dataclass(data_class)

    class_clean_ids = load_dataclass_clean_ids(parent_folder, data_class)
    class_corrupted_ids = load_dataclass_corrupted_ids(parent_folder, data_class)

    class_clean_patient_ids = load_dataclass_clean_patient_ids(parent_folder, data_class)
    class_corrupted_patient_ids = load_dataclass_corrupted_patient_ids(parent_folder, data_class)

    class_clean_ages = np.load(stats_folder + 'clean_ages.npy')
    class_clean_acquisition_date = np.load(stats_folder + 'clean_acquisition_date.npy')
    class_clean_max_values = np.load(stats_folder + 'clean_max_values.npy')
    class_clean_min_values = np.load(stats_folder + 'clean_min_values.npy')

    class_corrupted_ages = np.load(stats_folder + 'corrupted_ages.npy')
    class_corrupted_acquisition_date = np.load(stats_folder + 'corrupted_acquisition_date.npy')
    class_corrupted_max_values = np.load(stats_folder + 'corrupted_max_values.npy')
    class_corrupted_min_values = np.load(stats_folder + 'corrupted_min_values.npy')

    clean_data_size = len(class_clean_ids)
    corrupted_data_size = len(class_corrupted_ids)
    class_data_size = len(class_ids)

    clean_patient_size = len(class_clean_patient_ids)
    corrupted_patient_size = len(class_corrupted_patient_ids)
    class_patient_size = len(class_patient_ids)

    print('Number of data: ',  class_data_size)
    print('Number of clean data: ', clean_data_size)
    print('Number of corrupted data: ', corrupted_data_size)

    print()

    print('Number of individuals: ',  class_patient_size)
    print('Number of clean individuals: ', clean_patient_size)
    print('Number of corrupted individuals: ', corrupted_patient_size)

    print()

    data_values = []
    data_value_labels = []

    ages = []
    acquisition_dates = []
    acquisition_date_labels = []

    if clean_data_size > 0:

        print('CLEAN DATA STATS')

        print('Max values')
        print('99th percentile: ', np.percentile(class_clean_max_values, 99))
        print('95th percentile: ', np.percentile(class_clean_max_values, 95))
        print('75th percentile: ', np.percentile(class_clean_max_values, 75))
        print('Median: ', np.median(class_clean_max_values))
        print('Mean: ', np.mean(class_clean_max_values))

        print('Min values')
        print('1st percentile: ', np.percentile(class_clean_min_values, 1))
        print('5th percentile: ', np.percentile(class_clean_min_values, 5))
        print('25th percentile: ', np.percentile(class_clean_min_values, 25))
        print('Median: ', np.median(class_clean_min_values))
        print('Mean: ', np.mean(class_clean_min_values))

        data_values += [class_clean_max_values, class_clean_min_values]
        data_value_labels += ['Max values (clean)', 'Min value (clean)']

        ages += [class_clean_ages]
        acquisition_dates += [class_clean_acquisition_date]
        acquisition_date_labels += ['Clean data']

        print()

    if corrupted_data_size > 0:

        print('CORRUPTED DATA STATS')

        print('Max values')
        print('99th percentile: ', np.percentile(class_corrupted_max_values, 99))
        print('95th percentile: ', np.percentile(class_corrupted_max_values, 95))
        print('75th percentile: ', np.percentile(class_corrupted_max_values, 75))
        print('Median: ', np.median(class_corrupted_max_values))
        print('Mean: ', np.mean(class_corrupted_max_values))

        print('Min values')
        print('1st percentile: ', np.percentile(class_corrupted_min_values, 1))
        print('5th percentile: ', np.percentile(class_corrupted_min_values, 5))
        print('25th percentile: ', np.percentile(class_corrupted_min_values, 25))
        print('Median: ', np.median(class_corrupted_min_values))
        print('Mean: ', np.mean(class_corrupted_min_values))

        data_values += [class_corrupted_max_values, class_corrupted_min_values]
        data_value_labels += ['Max values (corrupted)', 'Min value (corrupted)']

        ages += [class_corrupted_ages]
        acquisition_dates += [class_corrupted_acquisition_date]
        acquisition_date_labels += ['Corrupted data']

        print()

    if clean_data_size > 0 and corrupted_data_size > 0:

        print('Plot voltage distribution...')

        plot_violin_distribution(data_values, data_value_labels, 'Value', plot_folder + 'data_values', ylims=[-4, 4])

        print('Plot age distribution...')

        plot_violin_distribution(ages, acquisition_date_labels, 'Age', plot_folder + 'age_distribution')

        print('Plot acquisition distribution...')

        plot_violin_distribution(acquisition_dates, acquisition_date_labels, 'Acquisition date', plot_folder + 'acquisition_distribution')

        print()

    if reset:

        reset_dataclass_training_ids(parent_folder, data_class)

    train_ids, valid_ids, test_ids = load_class_training_ids(parent_folder, data_class)
    train_patient_ids, valid_patient_ids, test_patient_ids = load_class_training_patient_ids(parent_folder, data_class)

    print('Number of train, valid, test data',
          len(train_ids), len(valid_ids), len(test_ids),
          'whose sum is',  len(train_ids) + len(valid_ids) + len(test_ids))

    print()

    print('Number of train, valid, test individuals',
          len(train_patient_ids), len(valid_patient_ids), len(test_patient_ids),
          'whose sum is',  len(train_patient_ids) + len(valid_patient_ids) + len(test_patient_ids))

    print()

    if analyze:

        analyze_dataclass(data_class, analyze=True)


if __name__ == '__main__':

    parser = argparse.ArgumentParser()

    parser.add_argument('-data_class', '--data_class', type=str)
    parser.add_argument('-multi_class', '--multi_class', action='store_const', const=True, default=False)
    parser.add_argument('-clean', '--clean', action='store_const', const=True, default=False)
    parser.add_argument('-analyze', '--analyze', action='store_const', const=True, default=False)
    parser.add_argument('-reset', '--reset', action='store_const', const=True, default=False)

    args = vars(parser.parse_args())

    data_class = args['data_class']

    clean = args['clean']
    analyze = args['analyze']
    reset = args['reset']

    process_dataclass(data_class, clean, analyze, reset)