negative-weights

Bellman Ford Algorithm Implementation using Python

Optimal-transport-based cell resampling for negative MC event weights (EMD / spectral EMD metrics, XMD bias measure). Code for arXiv:2607.08723.

Floyd–Warshall all-pairs shortest paths with optional O(n^2) incremental relaxation. / ワーシャル・フロイド法（全点対最短経路）。辺追加時の距離行列更新をO(n^2)で実行可能。

Exact shortest paths with negative edge weights (e.g. EV energy with regenerative braking) using Customizable Contraction Hierarchies — no Bellman–Ford, no potentials. Includes negative-cycle detection, live updates, a battery-constrained variant, baselines (Johnson, ALT, bidirectional A*), and DIMACS benchmarks. Pure Python.