-
Notifications
You must be signed in to change notification settings - Fork 8
/
05_preprocess.py
235 lines (203 loc) · 8.45 KB
/
05_preprocess.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
import numpy as np
from tqdm import tqdm
import os
import time
import json
import argparse
from glob import glob
import utils
def parse_args():
parser = argparse.ArgumentParser(description='preprocessing help')
parser.add_argument('--data-dir', type=str, default='data/processed',
help='data dir')
return parser.parse_args()
def get_time(t):
try:
t = float(t)
return t
except:
t = str(t).replace('"', '')
t = time.mktime(time.strptime(t,'%Y-%m-%d %H:%M:%S'))
t = int(t/3600)
return t
def generate_file_for_each_patient(args, features_csv):
selected_indices = []
initial_dir = args.initial_dir
os.system('rm -r ' + initial_dir)
utils.mkdir(initial_dir)
for i_line, line in enumerate(open(features_csv)):
if i_line % 10000 == 0:
print( i_line)
if i_line:
line_data = line.strip().split(',')
assert len(line_data) == len(feat_list)
new_line_data = [line_data[i_feat] for i_feat in selected_indices]
new_line = ','.join(new_line_data)
p_file = os.path.join(initial_dir, line_data[0] + '.csv')
if not os.path.exists(p_file):
wf = open(p_file, 'w')
wf.write(new_head)
wf.close()
wf = open(p_file, 'a')
wf.write('\n' + new_line)
wf.close()
else:
feat_list = utils.csv_split(line.strip())
feat_list = [f.strip('"') for f in feat_list]
print('There are {:d} features.'.format(len(feat_list)))
print(feat_list)
if len(selected_indices) == 0:
selected_indices = range(1, len(feat_list))
selected_feat_list = [feat_list[i_feat].replace('"','').replace(',', ';') for i_feat in selected_indices]
new_head = ','.join(selected_feat_list)
def resample_data(args, delta=1, ignore_time=-48):
resample_dir = args.resample_dir
initial_dir = args.initial_dir
os.system('rm -r ' + resample_dir)
utils.mkdir(resample_dir)
count_intervals = [0, 0]
count_dict = dict()
two_sets = [set(), set()]
for i_fi, fi in enumerate(tqdm(os.listdir(initial_dir))):
time_line_dict = dict()
for i_line, line in enumerate(open(os.path.join(initial_dir, fi))):
if i_line:
if len(line.strip()) == 0:
continue
line_data = line.strip().split(',')
assert len(line_data) == len(feat_list)
ctime = get_time(line_data[0])
ctime = delta * int(float(ctime) / delta)
if ctime not in time_line_dict:
time_line_dict[ctime] = []
time_line_dict[ctime].append(line_data)
else:
feat_list = line.strip().split(',')
feat_list[0] = 'time'
wf = open(os.path.join(resample_dir, fi), 'w')
wf.write(','.join(feat_list))
last_time = None
vis = 0
max_t = max(time_line_dict)
for t in sorted(time_line_dict):
if t - max_t < ignore_time:
continue
line_list = time_line_dict[t]
new_line = line_list[0]
for line_data in line_list:
for iv, v in enumerate(line_data):
if len(v.strip()):
new_line[iv] = v
new_line[0] = str(t - max_t)
new_line = '\n' + ','.join(new_line)
wf.write(new_line)
if last_time is not None:
delta_t = t - last_time
if delta_t > delta:
vis = 1
count_intervals[0] += 1
count_dict[t - last_time] = count_dict.get(t - last_time, 0) + 1
two_sets[0].add(fi)
two_sets[1].add(fi)
count_intervals[1] += 1
last_time = t
wf.close()
print('There are {:d}/{:d} collections data with intervals > {:d}.'.format(count_intervals[0], count_intervals[1], delta))
print('There are {:d}/{:d} patients with intervals > {:d}.'.format(len(two_sets[0]), len(two_sets[1]), delta))
def generate_feature_dict(args):
resample_dir = args.resample_dir
files = sorted(glob(os.path.join(resample_dir, '*')))
feature_value_dict = dict()
feature_missing_dict = dict()
for ifi, fi in enumerate(tqdm(files)):
if 'csv' not in fi:
continue
for iline, line in enumerate(open(fi)):
line = line.strip()
if iline == 0:
feat_list = line.split(',')
else:
data = line.split(',')
for iv, v in enumerate(data):
if v in ['NA', '']:
continue
else:
feat = feat_list[iv]
if feat not in feature_value_dict:
feature_value_dict[feat] = []
feature_value_dict[feat].append(float(v))
feature_mm_dict = dict()
feature_ms_dict = dict()
feature_range_dict = dict()
len_time = max([len(v) for v in feature_value_dict.values()])
for feat, vs in feature_value_dict.items():
vs = sorted(vs)
value_split = []
for i in range(args.split_num):
n = int(i * len(vs) / args.split_num)
value_split.append(vs[n])
value_split.append(vs[-1])
feature_range_dict[feat] = value_split
n = int(len(vs) / args.split_num)
feature_mm_dict[feat] = [vs[n], vs[-n - 1]]
feature_ms_dict[feat] = [np.mean(vs), np.std(vs)]
feature_missing_dict[feat] = 1.0 - 1.0 * len(vs) / len_time
json.dump(feature_mm_dict, open(os.path.join(args.files_dir, 'feature_mm_dict.json'), 'w'))
json.dump(feature_ms_dict, open(os.path.join(args.files_dir, 'feature_ms_dict.json'), 'w'))
json.dump(feat_list, open(os.path.join(args.files_dir, 'feature_list.json'), 'w'))
json.dump(feature_missing_dict, open(os.path.join(args.files_dir, 'feature_missing_dict.json'), 'w'))
json.dump(feature_range_dict, open(os.path.join(args.files_dir, 'feature_value_dict_{:d}.json'.format(args.split_num)), 'w'))
def split_data_to_ten_set(args):
resample_dir = args.resample_dir
files = sorted(glob(os.path.join(resample_dir, '*')))
np.random.shuffle(files)
splits = []
for i in range(10):
st = int(len(files) * i / 10)
en = int(len(files) * (i+1) / 10)
splits.append(files[st:en])
json.dump(splits, open(os.path.join(args.files_dir, 'splits.json'), 'w'))
def generate_label_dict(args, task):
label_dict = dict()
for i_line, line in enumerate(open(os.path.join(args.data_dir, '%s.csv' % task))):
if i_line:
data = line.strip().split(',')
pid = data[0]
label = ''.join(data[1:])
pid = str(int(float(pid)))
label_dict[pid] = label
json.dump(label_dict, open(os.path.join(args.files_dir, '%s_dict.json' % task), 'w'))
def generate_demo_dict(args, demo_csv):
demo_dict = dict()
demo_index_dict = dict()
for i_line, line in enumerate(open(demo_csv)):
if i_line:
data = line.strip().split(',')
pid = str(int(float(data[0])))
demo_dict[pid] = []
for demo in data[1:]:
if demo not in demo_index_dict:
demo_index_dict[demo] = len(demo_index_dict)
demo_dict[pid].append(demo_index_dict[demo])
json.dump(demo_dict, open(os.path.join(args.files_dir, 'demo_dict.json'), 'w'))
json.dump(demo_index_dict, open(os.path.join(args.files_dir, 'demo_index_dict.json'), 'w'))
def main():
args = parse_args()
args.files_dir = os.path.join(args.data_dir, 'files')
args.initial_dir = os.path.join(args.data_dir, 'initial_data')
args.resample_dir = os.path.join(args.data_dir, 'resample_dir')
args.split_num = 4000
utils.mkdir(args.files_dir)
utils.mkdir(args.initial_dir)
utils.mkdir(args.resample_dir)
features_csv = os.path.join(args.data_dir, 'features.csv')
demo_csv = os.path.join(args.data_dir, 'demo.csv')
for task in ['mortality', 'readmit', 'llos']:
generate_label_dict(args, task)
generate_demo_dict(args, demo_csv)
generate_file_for_each_patient(args, features_csv)
resample_data(args)
generate_feature_dict(args)
split_data_to_ten_set(args)
if __name__ == '__main__':
main()