def extract_sequence(self, path="./sample_data/testlog1_no_noise.csv"):
eventlog = Eventlog.from_txt(path, sep=',')
eventlog = eventlog.assign_caseid('Case ID')
eventlog = eventlog.assign_activity('Activity')
trace = eventlog.get_event_trace(workers=4, value='Activity')
trace = trace.values()
trace = ['_'.join(x) for x in trace]
trace_count=collections.Counter(trace)
test_trace_count = trace_count
return trace_count
def preparation(path="../result/generatedlog1.csv"):
eventlog = Eventlog.from_txt(path, sep=',')
eventlog = eventlog.assign_caseid('CaseID')
eventlog = eventlog.assign_activity('Activity')
eventlog = eventlog.assign_resource('Resource')
#eventlog = eventlog.assign_timestamp('Start Timestamp', name='Timestamp', format = '%Y/%m/%d %H:%M:%S')
#eventlog = eventlog.clear_columns()
eventlog['Duration'] = eventlog['Complete'] - eventlog['Start']
return eventlog
def prod_eventlog(self, start_point, columns=['CASE_ID', 'Activity', 'Resource', 'Start', 'Complete']): eventlog = pd.DataFrame.from_records(self.eventlist, columns=columns) eventlog['StartTimestamp']=start_point eventlog['CompleteTimestamp']=start_point eventlog['StartTimestamp'] = pd.to_datetime(eventlog['StartTimestamp'], format = '%Y-%m-%d %H:%M:%S', errors='ignore') eventlog['StartTimestamp'] += pd.to_timedelta(eventlog['Start'], unit='m') eventlog['CompleteTimestamp'] = pd.to_datetime(eventlog['CompleteTimestamp'], format = '%Y-%m-%d %H:%M:%S', errors='ignore') eventlog['CompleteTimestamp'] += pd.to_timedelta(eventlog['Complete'], unit='m') eventlog.sort_values(['CASE_ID', 'StartTimestamp'], inplace=True) eventlog = Eventlog(eventlog) return eventlog
def load_real_data(self,path):
eventlog = Eventlog.from_txt(path, sep=',')
eventlog = eventlog.assign_caseid('CASE_ID')
eventlog = eventlog.assign_activity('Activity')
eventlog['Resource'] = eventlog['Resource'].astype(int)
eventlog = eventlog.assign_resource('Resource')
eventlog = eventlog.assign_timestamp(name='StartTimestamp', new_name='StartTimestamp', _format = '%Y.%m.%d %H:%M:%S', errors='raise')
def to_minute(x):
t = x.time()
minutes = t.hour * 60 + t.minute
return minutes
eventlog['Start'] = eventlog['StartTimestamp'].apply(to_minute)
return eventlog
from PyProM.src.data.xes_reader import XesReader
from PyProM.src.mining.transition_matrix import TransitionMatrix
from PyProM.src.mining.dependency_graph import DependencyGraph
from PyProM.src.model.fsm import FSM_Miner
from PyProM.src.visualization.svg_widget import Visualization
from PyQt5 import QtSvg, QtCore, QtGui, Qt, QtWidgets
import multiprocessing
if __name__ == '__main__':
#eventlog
eventlog = Eventlog.from_txt(
'/Users/GYUNAM/Documents/example/repairExample.txt')
eventlog = eventlog.assign_caseid('Case ID')
eventlog = eventlog.assign_activity('Activity')
eventlog = eventlog.assign_resource('Resource')
eventlog = eventlog.assign_timestamp('Complete Timestamp')
eventlog = eventlog.clear_columns()
#preprocessing
#Transition Matrix
TM = TransitionMatrix()
transition_matrix = TM.get_transition_matrix(eventlog,
4,
type='sequence',
horizon=2)
transition_matrix = TM.annotate_transition_matrix(eventlog, 4,
def load_data(self,path):
eventlog = Eventlog.from_txt(path, sep=',')
eventlog = eventlog.assign_caseid('CASE_ID')
eventlog = eventlog.assign_activity('Activity')
eventlog = eventlog.assign_resource('Resource')
return eventlog
def remove_micro_seconds(x):
if len(x) > 19:
x = x[:19]
return x
def to_minute(x):
if np.isnan(x.seconds):
return x
#return int(x.seconds/60)
return math.ceil(x.seconds / 60)
if __name__ == '__main__':
path = '../sample_data/BPI_Challenge_2012.csv'
eventlog = Eventlog.from_txt(path, sep=',')
eventlog = eventlog.assign_caseid('Case ID')
eventlog = eventlog.assign_activity('Activity')
#eventlog = eventlog.assign_resource('Resource')
eventlog['transition'] = eventlog['lifecycle:transition']
eventlog['CompleteTimestamp'] = eventlog['Complete Timestamp']
eventlog['CompleteTimestamp'] = eventlog['CompleteTimestamp'].apply(
remove_micro_seconds)
eventlog['CompleteTimestamp'] = eventlog['CompleteTimestamp'].str.replace(
'.', '/', regex=False)
eventlog['Amount'] = eventlog['(case) AMOUNT_REQ']
#eventlog = eventlog.assign_timestamp('Start Timestamp', name='Timestamp', format = '%Y/%m/%d %H:%M:%S')
eventlog = eventlog.loc[
(eventlog['Activity'].str.contains('W_', regex=False))
& ~(eventlog['Activity'].str.contains('SCHEDULE'))]
caseid = ''
def remove_micro_seconds(x):
if len(x) > 19:
x = x[:19]
return x
def to_minute(x):
if np.isnan(x.seconds):
return x
#return int(x.seconds/60)
return math.ceil(x.seconds / 60)
if __name__ == '__main__':
path = './sample_data/BPI_Challenge_2012.csv'
eventlog = Eventlog.from_txt(path, sep=',')
eventlog = eventlog.assign_caseid('Case ID')
eventlog = eventlog.assign_activity('Activity')
#eventlog = eventlog.assign_resource('Resource')
eventlog['transition'] = eventlog['lifecycle:transition']
eventlog['CompleteTimestamp'] = eventlog['Complete Timestamp']
eventlog['CompleteTimestamp'] = eventlog['CompleteTimestamp'].apply(
remove_micro_seconds)
eventlog['CompleteTimestamp'] = eventlog['CompleteTimestamp'].str.replace(
'.', '/', regex=False)
eventlog['Amount'] = eventlog['(case) AMOUNT_REQ']
#eventlog = eventlog.assign_timestamp('Start Timestamp', name='Timestamp', format = '%Y/%m/%d %H:%M:%S')
eventlog = eventlog.loc[
(eventlog['Activity'].str.contains('W_', regex=False))
& ~(eventlog['Activity'].str.contains('SCHEDULE'))]
caseid = ''
from PyProM.src.data.Eventlog import Eventlog
def remove_micro_seconds(x):
if len(x) > 19:
x = x[:19]
return x
def to_minute(x):
if np.isnan(x.seconds):
return x
#return int(x.seconds/60)
return math.ceil(x.seconds / 60)
if __name__ == '__main__':
path = '../sample_data/BPIC15_1_unfiltered.csv'
eventlog = Eventlog.from_txt(path, sep=',')
eventlog = eventlog.assign_caseid('CASEOID')
eventlog = eventlog.assign_activity('ACTIVITYOID')
eventlog = eventlog.assign_timestamp('ENDAT')
#eventlog = eventlog.assign_resource('Resource')
#eventlog['transition'] = eventlog['lifecycle:transition']
import datetime
date_after = datetime.date(2011, 1, 1)
date_before = datetime.date(2014, 12, 31)
eventlog = eventlog.loc[(eventlog['TIMESTAMP'] > date_after)
& (eventlog['TIMESTAMP'] < date_before)]
print(eventlog)
eventlog.to_csv('../sample_data/BPIC15_1.csv')
from PyProM.src.data.Eventlog import Eventlog
import pandas as pd
class Preprocessor(object):
def load_eventlog(self,
path,
case,
activity,
timestamp,
encoding=None,
clear=True):
eventlog = pd.read_csv(path,
sep=',',
engine='python',
encoding=encoding)
eventlog.sort_values([case, timestamp], inplace=True)
eventlog.reset_index(drop=True, inplace=True)
eventlog = Eventlog(eventlog)
eventlog = eventlog.assign_caseid(case)
eventlog = eventlog.assign_activity(activity)
eventlog = eventlog.assign_timestamp(timestamp)
if clear == True:
eventlog = eventlog.clear_columns()
return eventlog