ProximityGraphs

ProximityGraphs is a Python-based computational geometry package for constructing and analyzing proximity and biological graphs, facilitating computational experimentation. It provides tools to generate and transform random and structured point sets and to build graphs from them. Its current scope comprises 13 geometric graphs, most of them proximity graphs—such as the Delaunay triangulation, the Gabriel graph, the Relative Neighborhood graph, and the Sphere-of-Influence graph—as well as the complete graph, the Erdős-Rényi random graph and two bio-inspired graphs.

Installation

We are on pipi! https://pypi.org/project/proximitygraphs/

pip install proximitygraphs

Or the editable on github

python -m pip install -e ".[dev, docs, gis]"

and to update the page

python -m sphinx -b html docs/source docs/build/html

Before pushing

python -m pytest
python -m ruff check .

Quickstart

import proximitygraphs as pg

points = pg.SetPoints.grid(shape=(3, 3))
mst = pg.MST(points)
unit_disk = pg.Unit_Disk(points, dist_max=1.01)

print(points.n)                   # 9 vertices
print(mst.m)                      # 8 edges
print(unit_disk.graph.get_edgelist())

A runnable example script is available at examples/quickstart.py.

API Overview

The main entry points are:

• pg.SetPoints for generating or loading planar point sets.
• pg.GeometricGraph for graph operations, analysis helpers, and visualization.
• pg.DelaunayG, pg.GG, pg.RNG, pg.MST, pg.Unit_Disk, pg.Alpha_Shape, and related classes for proximity graph construction.
• pg.Experiment for repeated simulations and metric aggregation.
• pg.PhysarumGraph for the package's current bio-inspired graph model.

GIS helpers such as SetPoints.from_geopandas() and GeometricGraph.to_gpd_lines() require the optional gis extra.

Reproducibility / Installation

This repository is configured and tested for:

• Windows local development
• GitHub Actions on Ubuntu
• Python 3.10, 3.11, 3.12, 3.13 and 3.14

The project uses a src layout: source code lives in src/proximitygraphs/, while the public import remains import proximitygraphs as pg.

The recommended validation sequence is:

python -m pip install -e ".[dev]"
python -m pytest -q
python -m ruff check .
python -m ruff format --check .

Citation

Software citation metadata is provided in CITATION.cff. A JOSS-ready manuscript draft is provided in paper.md.

The Zenodo DOI is still pending. Until archival metadata is finalized, use the versioned software citation in CITATION.cff and update it after a DOI is minted.

License

ProximityGraphs is distributed under the MIT License. See LICENSE.

Reporting Issues

Bug reports and feature requests should be filed through GitHub Issues. Security-sensitive issues should follow SECURITY.md.

Contributing

Development setup and contribution expectations are documented in CONTRIBUTING.md.