if mode:
# results for each machine type
csv_lines.append([machine_type])
csv_lines.append([
"section", "domain", "AUC", "pAUC", "precision", "recall",
"F1 score"
])
performance = []
dir_names = ["source_test", "target_test"]
for dir_name in dir_names:
#list machine id
section_names = com.get_section_names(target_dir,
dir_name=dir_name)
for section_name in section_names:
# load test file
files, y_true = com.file_list_generator(
target_dir=target_dir,
section_name=section_name,
dir_name=dir_name,
mode=mode)
# setup anomaly score file path
anomaly_score_csv = "{result}/anomaly_score_{machine_type}_{section_name}_{dir_name}.csv".format(
result=param["result_directory"],
machine_type=machine_type,
section_name=section_name,
dir_name=dir_name)
if os.path.exists(model_file_path):
com.logger.info("model exists")
continue
history_img = "{model}/history_{machine_type}.png".format(
model=param["model_directory"], machine_type=machine_type)
# pickle file for storing section names
section_names_file_path = "{model}/section_names_{machine_type}.pkl".format(
model=param["model_directory"], machine_type=machine_type)
# pickle file for storing anomaly score distribution
score_distr_file_path = "{model}/score_distr_{machine_type}.pkl".format(
model=param["model_directory"], machine_type=machine_type)
# get section names from wave file names
section_names = com.get_section_names(target_dir, dir_name="train")
unique_section_names = np.unique(section_names)
n_sections = unique_section_names.shape[0]
# make condition dictionary
joblib.dump(unique_section_names, section_names_file_path)
# generate dataset
print("============== DATASET_GENERATOR ==============")
# number of wave files in each section
# required for calculating y_pred for each wave file
n_files_ea_section = []
data = np.empty(
(0, param["feature"]["n_frames"] * param["feature"]["n_mels"]),
float)