pripel.py

import datetime
import sys

from pm4py.objects.log.exporter.xes import exporter as xes_exporter
from pm4py.objects.log.importer.xes import importer as xes_import_factory

from attributeAnonymizier import AttributeAnonymizer as AttributeAnonymizer
from trace_variant_query import privatize_tracevariants
from tracematcher import TraceMatcher as TraceMatcher


def freq(lst):
    d = {}
    for i in lst:
        if d.get(i):
            d[i] += 1
        else:
            d[i] = 1
    return d


log_path = sys.argv[1]
epsilon = float(sys.argv[2])
N = int(sys.argv[3])
k = int(sys.argv[4])

new_ending = "_epsilon_" + str(epsilon) + "_k" + str(k) + "_anonymized.xes"

result_log_path = log_path.replace(".xes", new_ending)

starttime = datetime.datetime.now()
log = xes_import_factory.apply(log_path)

starttime_tv_query = datetime.datetime.now()
tv_query_log = privatize_tracevariants(log, epsilon, k, N)
if (len(tv_query_log) == 0):
    raise ValueError(
        "Pruning parameter k is too high. The result of the trace variant query is empty. At least k traces must appear "
        "in a noisy variant count to be part of the result of the query.")
endtime_tv_query = datetime.datetime.now()
print("Time of TV Query: " + str((endtime_tv_query - starttime_tv_query)))
starttime_trace_matcher = datetime.datetime.now()
traceMatcher = TraceMatcher(tv_query_log, log)
matchedLog = traceMatcher.matchQueryToLog()
print(len(matchedLog))
endtime_trace_matcher = datetime.datetime.now()
print("Time of TraceMatcher: " + str((endtime_trace_matcher - starttime_trace_matcher)))
distributionOfAttributes = traceMatcher.getAttributeDistribution()
occurredTimestamps, occurredTimestampDifferences = traceMatcher.getTimeStampData()
print(min(occurredTimestamps))
starttime_attribute_anonymizer = datetime.datetime.now()
attributeAnonymizer = AttributeAnonymizer()
anonymizedLog, attributeDistribution = attributeAnonymizer.anonymize(matchedLog, distributionOfAttributes, epsilon,
                                                                     occurredTimestampDifferences, occurredTimestamps)
endtime_attribute_anonymizer = datetime.datetime.now()
print("Time of attribute anonymizer: " + str(endtime_attribute_anonymizer - starttime_attribute_anonymizer))
xes_exporter.apply(anonymizedLog, result_log_path)
endtime = datetime.datetime.now()
print("Complete Time: " + str((endtime - starttime)))
print("Time of TV Query: " + str((endtime_tv_query - starttime_tv_query)))
print("Time of TraceMatcher: " + str((endtime_trace_matcher - starttime_trace_matcher)))
print("Time of attribute anonymizer: " + str(endtime_attribute_anonymizer - starttime_attribute_anonymizer))
print(result_log_path)
print(freq(attributeDistribution))