We release our most successful implementations as open source software libraries that are easy to use in real-world applications. Our tools provide easy-to-use interfaces compatible with existing systems, scalability to huge inputs and large numbers of processors, and modular, well-documented code.
CHSZLabLib -- unified Python frontend wrapping most of our libraries into a single pip install.
Partitioning
Cuts
- VieCut -- Vienna Minimum Cuts
- HeiCut -- Hypergraph Minimum Cuts
- HeiConnect -- Connectivity Augmentation
- DMAX -- Data Reduction for Maximum Cut
Clustering
- VieClus -- Vienna Graph Clustering
- SCC -- Scalable Correlation Clustering
- HeidelbergMotifClustering -- Local Motif Clustering
Independent Sets
- KaMIS -- Maximum Independent Sets
- CHILS -- Concurrent MWIS Heuristic
- LearnAndReduce -- GNN-Guided MWIS Reductions
- HyperMIS -- Hypergraph Independent Sets
- red2pack -- 2-Packing Set Solver
Dynamic Graphs
- DynDeltaOrientation -- Dynamic Edge Orientation
- DynMatch -- Dynamic Maximal Matching
- DynWMIS -- Dynamic Max Weight Ind. Set
- DyReach -- Dynamic Reachability
Process Mapping
- VieM -- Mapping & Sparse QAP
- SharedMap -- Shared-Memory Process Mapping
- IntegratedProcessMapping -- Multi-Level Mapping