Installer kahip avec Homebrew, Nix

Liens du projet

• Page d'accueil
• Dépôt
• Page du gestionnaire de paquets

Commandes et alias

• dspac
• edge_evaluator
• edge_list_to_metis_graph
• edge_partitioning
• evaluator
• global_multisection
• graph2binary
• graph2binary_external
• graphchecker
• interface_test
• kaffpa
• kaffpaE
• label_propagation
• node_ordering
• node_separator
• parhip
• partition_to_vertex_separator
• readbgf

Historique du projet

KaHIP came out of the Karlsruhe graph-partitioning research line associated with Peter Sanders and Christian Schulz. The 2013 arXiv user guide describes the framework and its interfaces, and the v3.00 line covered multilevel partitioning, memetic algorithms, distributed and shared-memory parallel algorithms, node separators, node ordering, edge partitioning, and ILP-based improvement.

Historique d'adoption

The project is distributed as source, command-line tools, libraries, Python bindings, and package-manager builds. Its README emphasizes use cases in scientific computing and systems work, including mesh partitioning, social and web graphs, sparse-matrix ordering, distributed partitioning with ParHIP, and process mapping for parallel applications.

Modes d'utilisation

Users commonly feed KaHIP graphs in METIS-style formats, check them with graphchecker, partition with kaffpa or kaffpaE, run distributed partitioning with parhip, evaluate cuts, compute node separators, or call the library and Python interface from larger workflows.

Pourquoi les passionnés de paquets s'y intéressent

KaHIP is the kind of package that exposes research-grade algorithm engineering as installable Unix tooling: many specialized executables, reproducible graph file formats, and a long release train turning academic partitioning methods into scripts that fit HPC and data-processing pipelines.

Chronologie

• 2013-11-07: KaHIP v3.00 user guide first submitted to arXiv.
• 2019-03-07: GitHub release v2.10 published.
• 2020-08-03: GitHub release v3.00 published.
• 2025-11-03: Release v3.21 added pip-oriented KaFFPa use.
• 2026-03-13: Release v3.25 published.

Related projects

• Related tools and methods include METIS, ParMETIS, Scotch/PT-Scotch, KaFFPa, KaFFPaE, KaBaPE, ParHIP, and KaHyPar for hypergraph partitioning.

Sources

• https://algo2.iti.kit.edu/documents/kahip/index.html
• https://api.github.com/repos/KaHIP/KaHIP/releases?per_page=100
• https://arxiv.org/abs/1311.1714
• https://github.com/KaHIP
• https://github.com/KaHIP/KaHIP