pymocd 1.1.3

No project description provided

Python Multi-Objective Community Detection Algorithms

pymocd is a Python library, powered by a Rust backend, for performing efficient multi-objective evolutionary community detection in complex networks. This library is designed to deliver enhanced performance compared to traditional methods, making it particularly well-suited for analyzing large-scale graphs.

Navigate the Documentation for detailed guidance and usage instructions.

Table of Contents

• Understanding Community Detection with HP-MOCD
• Getting Started
  • Key Features
• Contributing
• Citation

---

Understanding Community Detection with HP-MOCD

The HP-MOCD algorithm, central to pymocd, identifies community structures within a graph. It proposes a solution by grouping nodes into distinct communities, as illustrated below:

Original Graph	Proposed Community Structure

Getting Started

Installing the library using pip interface:

pip install pymocd

For an easy usage:

import networkx
import pymocd

G = networkx.Graph() # Your graph
alg = pymocd.HpMocd(G)
communities = alg.run()

[!IMPORTANT] Graphs must be provided in NetworkX or Igraph compatible format**.

Refer to the official Documentation for detailed instructions and more usage examples.

Contributing

We welcome contributions to pymocd! If you have ideas for new features, bug fixes, or other improvements, please feel free to open an issue or submit a pull request. This project is licensed under the GPL-3.0 or later.

---

Citation

If you use pymocd or the HP-MOCD algorithm in your research, please cite the following paper:

@article{santos2025hpmocd,
  author    = {Guilherme O. Santos, Lucas S. Vieira, Giulio Rossetti, Carlos H. G. Ferreira and Gladston J. P. Moreira},
  title     = {HP-MOCD: A High-Performance Multi-Objective Community Detection Algorithm for Large-Scale Networks},
  journal   = {The 17th International Conference on Advances in Social Networks Analysis and Mining},
  year      = {2025},
  abstract  = {Community detection in social networks has traditionally been approached as a single-objective optimization problem, with various heuristics targeting specific community-defining metrics. However, this approach often proves inadequate for capturing the multifaceted nature of communities. We introduce HP-MOCD, a fully parallelized, evolutionary high-performance multi-objective community detection algorithm designed specifically for large-scale networks. Our implementation overcomes the computational challenges that typically limit multi-objective approaches in this domain. While performance may decrease with networks containing high proportions of inter-community edges, extensive evaluations on synthetic datasets demonstrate that HP-MOCD achieves an exceptional balance between scalability and detection accuracy. Available as open-source software, HP-MOCD offers researchers and practitioners a practical, powerful solution for complex network analysis, particularly for applications requiring both efficiency and detection quality.},
  keywords  = {community detection, complex networks, evolutionary algorithms, genetic algorithms, multi-objective},
  note      = {Currently under analysis by the journal}
}