def payload_extractor_trial(log_name, settings_file):
log = read_XES_log(log_name)
train, test = split_log_train_test(log, 0.8)
print("Lengths of logs train: {}, test: {}".format(len(train), len(test)))
pex = PayloadExtractor()
settings = settings_from_cfg(settings_file)
## Get first forms of data
extracted_train_data = pex.get_overview_of_data(train, settings)
extracted_test_data = pex.get_overview_of_data(test, settings)
train_df, test_df = build_dataframes(extracted_train_data,
extracted_test_data, settings)
return train_df, test_df
def data_declare_main(inp_folder, log_name, ignored):
drc = DRC()
log = read_XES_log(log_name)
# Transform log into suitable data structures
transformed_log = xes_to_data_positional(log)
train_log, test_log = split_log_train_test(transformed_log, 0.8)
#print(train_log[0])
train_case_ids = [tr["name"] for tr in train_log]
test_case_ids = [tr["name"] for tr in test_log]
train_names, train_features, test_names, test_features = drc.create_data_aware_features(
train_log, test_log, ignored)
train_dict = {}
test_dict = {}
for i, tf in enumerate(train_features):
train_dict[train_names[i]] = tf
for i, tf in enumerate(test_features):
test_dict[test_names[i]] = tf
train_df = pd.DataFrame.from_dict(train_dict)
test_df = pd.DataFrame.from_dict(test_dict)
#train_df = pd.DataFrame(train_features, columns=train_names)
#test_df = pd.DataFrame(test_features, columns=test_names)
#print(train_names)
# add Case_ID
train_df["Case_ID"] = train_case_ids
test_df["Case_ID"] = test_case_ids
train_df.to_csv(inp_folder + "/dwd_train.csv", index=False)
test_df.to_csv(inp_folder + "/dwd_test.csv", index=False)
def payload_extractor(inp_folder, log_name, settings_file):
log = read_XES_log(log_name)
train, test = split_log_train_test(log, 0.8)
print("Lengths of logs train: {}, test: {}".format(len(train), len(test)))
pex = PayloadExtractor()
settings = settings_from_cfg(settings_file)
## Get first forms of data
extracted_train_data = pex.get_overview_of_data(train, settings)
extracted_test_data = pex.get_overview_of_data(test, settings)
train_df, test_df = build_dataframes(extracted_train_data,
extracted_test_data, settings)
# Force all to float (except label?)
train_df.to_csv(inp_folder + "/payload_train.csv", index=False)
test_df.to_csv(inp_folder + "/payload_test.csv", index=False)
return train_df, test_df
def baseline(inp_folder, logPath):
log = read_XES_log(logPath)
transformed_log = xes_to_positional(log)
train_log, test_log = split_log_train_test(transformed_log, 0.8)
# Collect all different IA's
activitySet = list(extract_unique_events_transformed(train_log))
# Transform to matrix
print("Train data")
# train data
train_df = transform_log(train_log, activitySet)
print("Test data")
# test data
test_df = transform_log(test_log, activitySet)
train_df.to_csv(inp_folder + "/baseline_train.csv", index=False)
test_df.to_csv(inp_folder + "/baseline_test.csv", index=False)