netseg

netseg is a Python package for measuring segregation and structural polarization in social networks.

This package extends and complements the original netseg R package (Bojanowski & Corten, 2014; Bojanowski, 2025), which introduced the unified framework for measuring network segregation. The Python port and additional polarization measures were developed, with new features subsequently integrated into both packages.

It was developed to address several key gaps in the current Python ecosystem:

• A Unified Framework: While analyzing network segregation and polarization is common, the myriad of existing measures makes cross-network comparisons exceedingly difficult. netseg provides a centralized toolkit for standardizing these measurements.

• Extensively Documented: netseg is extensively documented. From various case by case usage examples to the advanced null-model selection, it offers a clear guide for your research.

• Advanced Null-Model Comparison: Many existing approaches rely on purely random graphs to quantify excess polarization or segregation, a notoriously low baseline. netseg allows you to use customized network ensembles as null models, enabling robust, comparative segregation analysis.

• Extensions and Improved Optimization netseg offers extensions for already existing measures, calculating the structural polarization and segregation in directed and undirected networks with multiple groups. netseg is build on igraph and NumPy , it offers near bare-metal speed with NumPy's vectorization and igraph's C backend.

Support and Contact

If you come across any issues, please visit the repository page https://codeberg.org/OnurB/netseg. Suggestions & Discussions: bal_onur@phd.ceu.edu

netseg R package is maintained at https://mbojan.github.io/netseg/.

Installation

You can install netseg through PyPI with

pip install netseg

Or you can install directly from the source without the blobs with

git clone --filter=blob:none --sparse https://codeberg.org/OnurB/netseg.git
cd netseg
git sparse-checkout set src
pip install .

Documentation

You can reach to the documentation here. The rendered documentation is served from Codeberg Pages (and will stay here due to ethical reasons). That service is independent of codeberg.org and occasionally unavailable while the repository itself remains reachable. The same built HTML lives on the pages branch, so you can always read the docs locally:

git clone --branch pages --single-branch --depth 1 https://codeberg.org/OnurB/netseg.git netseg-docs
cd netseg-docs
python3 -m http.server 8000

Then open http://localhost:8000.

Citation

If you are using netseg, please cite as:

Bal, O. T., & Bojanowski, M. (2026). netseg: A Python package for measuring structural polarization and segregation in social networks (arXiv:2609.16088). arXiv. https://doi.org/10.48550/arXiv.2609.16088

or as BibTeX:

@misc{bal2026netsegpythonpackagemeasuring,
      title={netseg: a Python Package for Measuring Structural Polarization and Segregation in Social Networks}, 
      author={Onur Tuncay Bal and Michał Bojanowski},
      year={2026},
      eprint={2609.16088},
      archivePrefix={arXiv},
      primaryClass={cs.SI},
      url={https://arxiv.org/abs/2609.16088}, 
}

References

Bojanowski M (2025). Measures of Network Segregation and Homophily. R package version 1.0-3, https://mbojan.github.io/netseg/.

Bojanowski M, Corten R (2014). “Measuring Segregation in Social Networks.” Social Networks, 39, 14-32. doi:10.1016/j.socnet.2014.04.001 https://doi.org/10.1016/j.socnet.2014.04.001.