文章目录

Ref
Quick Start
1. 虚拟环境配置
1. 读取log
- 1.1 读取XES格式log
- - pm4py.read_xes(log)
  - xes_importer.apply(log)
- 1.2 读取CSV格式log
2. miner
- Process model - BPMN
- Petri net
- Process tree
3. view
代码

Ref

【pm4py 官网】https://pm4py.fit.fraunhofer.de/
【ProM官网】http://promtools.org/doku.php

Quick Start

A sneak peak of pmp4y’s simple application:

import pm4py
log = pm4py.read_xes('<path-to-xes-log-file.xes>')
process_model = pm4py.discover_bpmn_inductive(log)
pm4py.view_bpmn(process_model)

即读取XES log，基于inductive miner生成bpmn，并进行预览。例如：

在这里插入图片描述

1. 虚拟环境配置

创建虚拟环境：

virtualenv -p python3.7 ~/venv/processmining
source ~/venv/processminging/bin/activate

安装pm4py：

python3.7 -m pip install pm4py

pip list

pm4py为最新版本2.2.27

在这里插入图片描述

1. 读取log

可读取XES和CSV格式log：

1.1 读取XES格式log

pm4py.read_xes(log)

import pm4py
log = pm4py.read_xes('tests/input_data/running-example.xes')

此时可查看log，可看出log包含25个eventlog（即ProM中的cases），第一个eventlog包含92个traces（即Prom中第一个case包含92个event），第2个eventlog包含62个traces；这25个eventlog包含的traces数目最大为92，最小为34。

其中第一个eventlog的第一个trace（即ProM中第一个case的第一个event），concept:name为Start，lifecycle:transition为start……
在这里插入图片描述

下图即为ProM软件（【ProM官网】http://promtools.org/doku.php）对同一XES日志文件的读取分析数据，与pm4py相同。

在这里插入图片描述

xes_importer.apply(log)

除了pm4py官网上所显示的pm4py.read_xes('tests/input_data/running-example.xes')，还可以通过converter进行读取：

from pm4py.objects.log.importer.xes import importer as xes_importer
log = xes_importer.apply('<path-to-xes-log-file.xes>')

在这里插入图片描述

1.2 读取CSV格式log

import pandas as pd
import pm4pydataframe = pd.read_csv('tests/input_data/running-example.csv', sep=',')
dataframe = pm4py.format_dataframe(dataframe, case_id='case:concept:name', activity_key='concept:name', timestamp_key='time:timestamp')
event_log = pm4py.convert_to_event_log(dataframe)

要注意的是，表头应与case:concept:name, concept:name，time:timestamp对应；

但是当表格中有起始和终止时间时，未确定读取方法。（可在ProM中读取CSV，再转换为XES格式；然后即可通过1.1的方式读取）

2. miner

Process model - BPMN

process_model = pm4py.discover_bpmn_inductive(log)

在这里插入图片描述

Petri net

petri_net, initial_marking, final_marking = pm4py.discover_petri_net_inductive(log)

Process tree

process_tree = pm4py.discover_process_tree_inductive(log)

在这里插入图片描述

3. view

# BPMN
pm4py.view_bpmn(process_model_bpmn, format="pdf")# Petri net
pm4py.view_petri_net(petri_net)# Process tree
pm4py.view_process_tree(process_tree)

在这里插入图片描述

放大来看：
BPMN：
在这里插入图片描述
Petri net：

Process tree：
在这里插入图片描述

代码

#!/usr/bin/env python3
# -*- coding: utf-8 -*-# @Time : 8/26/22 4:13 PM
# @Author : 
# @File : processmining.py
# @Version : 
# @Software: PyCharm
# @Description : 'process mining'import pm4py
from pm4py.objects.log.importer.xes import importer as xes_importerdef quick_start():log = pm4py.read_xes('/home/deng/cugb/STL/CSTL/Process/PromEventLog/activitylog_uci_detailed_labour.xes')process_model = pm4py.discover_bpmn_inductive(log)pm4py.view_bpmn(process_model)def read_xes():# read XES log# by pm4py.read_xeslog = pm4py.read_xes('./PromEventLog/activitylog_uci_detailed_labour.xes')# by xes_importer.applylog = xes_importer.apply('./PromEventLog/activitylog_uci_detailed_labour.xes')log = xes_importer.apply('./PromEventLog/PurchasingExample_convertedByProM.xes')return logdef miner(log):# BPMNprocess_model_bpmn = pm4py.discover_bpmn_inductive(log)# Petri netpetri_net, initial_marking, final_marking = pm4py.discover_petri_net_inductive(log)# Process treeprocess_tree = pm4py.discover_process_tree_inductive(log)return process_model_bpmn, petri_net, process_treedef view(process_model_bpmn, petri_net, process_tree):# BPMNpm4py.view_bpmn(process_model_bpmn, format="pdf")pm4py.save_vis_bpmn(process_model_bpmn, 'test.pdf')# Petri netpm4py.view_petri_net(petri_net)# Process treepm4py.view_process_tree(process_tree)if __name__ == '__main__':quick_start()log = read_xes()process_model_bpmn, petri_net, process_tree = miner(log)view(process_model_bpmn, petri_net, process_tree)print('done')