-
Notifications
You must be signed in to change notification settings - Fork 47
/
01_test.py
245 lines (211 loc) · 10.5 KB
/
01_test.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
"""
@file 01_test.py
@brief Script for test
@author Toshiki Nakamura, Yuki Nikaido, and Yohei Kawaguchi (Hitachi Ltd.)
Copyright (C) 2020 Hitachi, Ltd. All right reserved.
"""
########################################################################
# import default python-library
########################################################################
import os
import glob
import csv
import re
import itertools
import sys
########################################################################
########################################################################
# import additional python-library
########################################################################
import numpy
# from import
from tqdm import tqdm
from sklearn import metrics
# original lib
import common as com
import keras_model
########################################################################
########################################################################
# load parameter.yaml
########################################################################
param = com.yaml_load()
#######################################################################
########################################################################
# def
########################################################################
def save_csv(save_file_path,
save_data):
with open(save_file_path, "w", newline="") as f:
writer = csv.writer(f, lineterminator='\n')
writer.writerows(save_data)
def get_machine_id_list_for_test(target_dir,
dir_name="test",
ext="wav"):
"""
target_dir : str
base directory path of "dev_data" or "eval_data"
test_dir_name : str (default="test")
directory containing test data
ext : str (default="wav)
file extension of audio files
return :
machine_id_list : list [ str ]
list of machine IDs extracted from the names of test files
"""
# create test files
dir_path = os.path.abspath("{dir}/{dir_name}/*.{ext}".format(dir=target_dir, dir_name=dir_name, ext=ext))
file_paths = sorted(glob.glob(dir_path))
# extract id
machine_id_list = sorted(list(set(itertools.chain.from_iterable(
[re.findall('id_[0-9][0-9]', ext_id) for ext_id in file_paths]))))
return machine_id_list
def test_file_list_generator(target_dir,
id_name,
dir_name="test",
prefix_normal="normal",
prefix_anomaly="anomaly",
ext="wav"):
"""
target_dir : str
base directory path of the dev_data or eval_data
id_name : str
id of wav file in <<test_dir_name>> directory
dir_name : str (default="test")
directory containing test data
prefix_normal : str (default="normal")
normal directory name
prefix_anomaly : str (default="anomaly")
anomaly directory name
ext : str (default="wav")
file extension of audio files
return :
if the mode is "development":
test_files : list [ str ]
file list for test
test_labels : list [ boolean ]
label info. list for test
* normal/anomaly = 0/1
if the mode is "evaluation":
test_files : list [ str ]
file list for test
"""
com.logger.info("target_dir : {}".format(target_dir+"_"+id_name))
# development
if mode:
normal_files = sorted(
glob.glob("{dir}/{dir_name}/{prefix_normal}_{id_name}*.{ext}".format(dir=target_dir,
dir_name=dir_name,
prefix_normal=prefix_normal,
id_name=id_name,
ext=ext)))
normal_labels = numpy.zeros(len(normal_files))
anomaly_files = sorted(
glob.glob("{dir}/{dir_name}/{prefix_anomaly}_{id_name}*.{ext}".format(dir=target_dir,
dir_name=dir_name,
prefix_anomaly=prefix_anomaly,
id_name=id_name,
ext=ext)))
anomaly_labels = numpy.ones(len(anomaly_files))
files = numpy.concatenate((normal_files, anomaly_files), axis=0)
labels = numpy.concatenate((normal_labels, anomaly_labels), axis=0)
com.logger.info("test_file num : {num}".format(num=len(files)))
if len(files) == 0:
com.logger.exception("no_wav_file!!")
print("\n========================================")
# evaluation
else:
files = sorted(
glob.glob("{dir}/{dir_name}/*{id_name}*.{ext}".format(dir=target_dir,
dir_name=dir_name,
id_name=id_name,
ext=ext)))
labels = None
com.logger.info("test_file num : {num}".format(num=len(files)))
if len(files) == 0:
com.logger.exception("no_wav_file!!")
print("\n=========================================")
return files, labels
########################################################################
########################################################################
# main 01_test.py
########################################################################
if __name__ == "__main__":
# check mode
# "development": mode == True
# "evaluation": mode == False
mode = com.command_line_chk()
if mode is None:
sys.exit(-1)
# make output result directory
os.makedirs(param["result_directory"], exist_ok=True)
# load base directory
dirs = com.select_dirs(param=param, mode=mode)
# initialize lines in csv for AUC and pAUC
csv_lines = []
# loop of the base directory
for idx, target_dir in enumerate(dirs):
print("\n===========================")
print("[{idx}/{total}] {dirname}".format(dirname=target_dir, idx=idx+1, total=len(dirs)))
machine_type = os.path.split(target_dir)[1]
print("============== MODEL LOAD ==============")
# set model path
model_file = "{model}/model_{machine_type}.hdf5".format(model=param["model_directory"],
machine_type=machine_type)
# load model file
if not os.path.exists(model_file):
com.logger.error("{} model not found ".format(machine_type))
sys.exit(-1)
model = keras_model.load_model(model_file)
model.summary()
if mode:
# results by type
csv_lines.append([machine_type])
csv_lines.append(["id", "AUC", "pAUC"])
performance = []
machine_id_list = get_machine_id_list_for_test(target_dir)
for id_str in machine_id_list:
# load test file
test_files, y_true = test_file_list_generator(target_dir, id_str)
# setup anomaly score file path
anomaly_score_csv = "{result}/anomaly_score_{machine_type}_{id_str}.csv".format(
result=param["result_directory"],
machine_type=machine_type,
id_str=id_str)
anomaly_score_list = []
print("\n============== BEGIN TEST FOR A MACHINE ID ==============")
y_pred = [0. for k in test_files]
for file_idx, file_path in tqdm(enumerate(test_files), total=len(test_files)):
try:
data = com.file_to_vector_array(file_path,
n_mels=param["feature"]["n_mels"],
frames=param["feature"]["frames"],
n_fft=param["feature"]["n_fft"],
hop_length=param["feature"]["hop_length"],
power=param["feature"]["power"])
errors = numpy.mean(numpy.square(data - model.predict(data)), axis=1)
y_pred[file_idx] = numpy.mean(errors)
anomaly_score_list.append([os.path.basename(file_path), y_pred[file_idx]])
except:
com.logger.error("file broken!!: {}".format(file_path))
# save anomaly score
save_csv(save_file_path=anomaly_score_csv, save_data=anomaly_score_list)
com.logger.info("anomaly score result -> {}".format(anomaly_score_csv))
if mode:
# append AUC and pAUC to lists
auc = metrics.roc_auc_score(y_true, y_pred)
p_auc = metrics.roc_auc_score(y_true, y_pred, max_fpr=param["max_fpr"])
csv_lines.append([id_str.split("_", 1)[1], auc, p_auc])
performance.append([auc, p_auc])
com.logger.info("AUC : {}".format(auc))
com.logger.info("pAUC : {}".format(p_auc))
print("\n============ END OF TEST FOR A MACHINE ID ============")
if mode:
# calculate averages for AUCs and pAUCs
averaged_performance = numpy.mean(numpy.array(performance, dtype=float), axis=0)
csv_lines.append(["Average"] + list(averaged_performance))
csv_lines.append([])
if mode:
# output results
result_path = "{result}/{file_name}".format(result=param["result_directory"], file_name=param["result_file"])
com.logger.info("AUC and pAUC results -> {}".format(result_path))
save_csv(save_file_path=result_path, save_data=csv_lines)