scanwidth 0.2.1

Package for computing the edge- and node-scanwidth of a directed acyclic graph (DAG)

scanwidth

scanwidth is a Python package for computing edge-scanwidth and node-scanwidth of directed acyclic graphs (DAGs). It provides exact and heuristic algorithms, plus reduction pipelines that make practical computation on larger instances more tractable.

Key Features

• Tree-extension and Extension classess: classes that support (tree-)extensions of DAGs, let one compute scanwidth-bags and convert to canonical tree-extenions
• Exact and heuristic solvers: XP, brute-force, partition-based exact methods (edge), ILP backend selection (node), and multiple heuristics for edge- and node-scanwidth.
• Reduction framework: configurable ReducerConfig/Reducer pipelines for edge and node scanwidth, including optional parallel s-block solving.

Installation

Install the base package:

pip install scanwidth

Install optional ILP dependencies:

# SciPy backend for node_scanwidth(..., algorithm="ilp", backend="scipy")
pip install scanwidth[scipy]

# Gurobi backend for node_scanwidth(..., algorithm="ilp", backend="gurobi")
pip install scanwidth[gurobi]

# Both ILP backends
pip install scanwidth[ilp]

gurobipy requires a working Gurobi installation and a valid Gurobi license (commercial or academic, depending on your setup).

Documentation

For installation instructions, quickstart examples, and full API reference, see the scanwidth docs.

Citation

If you use scanwidth in your research, please cite:

Niels Holtgrefe, Leo van Iersel, and Mark Jones. Exact and heuristic computation of the scanwidth of directed acyclic graphs. Journal of Computer and System Sciences, 160:103802, 2026. doi: 10.1016/j.jcss.2026.103802

License

MIT License - see LICENSE.