PM4Py is a python library that supports state-of-the-art process mining algorithms in Python. It is open source and intended to be used in both academia and industry projects.
PM4Py is managed and developed by PIS — Process Intelligence Solutions (https://processintelligence.solutions/), a spin-off from the Fraunhofer Institute for Applied Information Technology FIT where PM4Py was initially developed.
The open-source version of PM4Py, available on GitHub (https://github.com/process-intelligence-solutions/pm4py), is licensed under the GNU Affero General Public License version 3 (AGPL-3.0).
We offer a separate version of PM4Py for commercial use in closed-source environments under a different license. For more information about the licensing options for using PM4Py in closed-source settings, please visit https://processintelligence.solutions/pm4py#licensing.
The documentation of PM4Py can be found at https://processintelligence.solutions/pm4py/.
Here is a simple example to spark your interest:
import pm4py
if __name__ == "__main__":
log = pm4py.read_xes('<path-to-xes-log-file.xes>')
net, initial_marking, final_marking = pm4py.discover_petri_net_inductive(log)
pm4py.view_petri_net(net, initial_marking, final_marking, format="svg")PM4Py can be installed on Python 3.9.x / 3.10.x / 3.11.x / 3.12.x / 3.13.x / 3.14.x by invoking:
pip install -U pm4py
PM4Py is also running on older Python environments with different requirements sets, including:
- Python 3.8 (3.8.10):
third_party/old_python_deps/requirements_py38.txt
PM4Py depends on some other Python packages, with different levels of importance:
- Essential requirements: numpy, pandas, deprecation, networkx
- Normal requirements (installed by default with the PM4Py package, important for mainstream usage): graphviz, intervaltree, lxml, matplotlib, pydotplus, pytz, scipy, tqdm
- Optional requirements (not installed by default): requests, pyvis, jsonschema, workalendar, pyarrow, scikit-learn, polars, openai, pyemd, pyaudio, pydub, pygame, pywin32, pygetwindow, pynput
To track the incremental updates, please refer to the CHANGELOG.md file.
If you want to contribute to PM4Py, please review the contributing guidelines and Contributor License Agreement (CLA).
As scientific library in the Python ecosystem, we rely on external libraries to offer our features.
In the /third_party folder, we list all the licenses of our direct dependencies.
Please check the /third_party/LICENSES_TRANSITIVE file to get a full list of all transitive dependencies and the
corresponding license.
If you are using PM4Py in your scientific work, please cite PM4Py as follows:
Alessandro Berti, Sebastiaan van Zelst, Daniel Schuster. (2023). PM4Py: A process mining library for Python. Software Impacts, 17, 100556. doi: 10.1016/j.simpa.2023.100556
BiBTeX:
@article{pm4py,
title = {PM4Py: A process mining library for Python},
journal = {Software Impacts},
volume = {17},
pages = {100556},
year = {2023},
issn = {2665-9638},
doi = {https://doi.org/10.1016/j.simpa.2023.100556},
url = {https://www.sciencedirect.com/science/article/pii/S2665963823000933},
author = {Alessandro Berti and Sebastiaan van Zelst and Daniel Schuster},
}This repository is managed by Process Intelligence Solutions (PIS). Further information about PIS can be found online at https://processintelligence.solutions.
