Lonestar
Lonestar is a collection of widely-used real-world applications that exhibit irregular behavior. Since version 3.0, the Lonestar benchmark suite is an integral component of the Galois release and thus can be downloaded here.
The last standalone version of Lonestar, version 2.1, is still available, but it is no longer actively maintained.
Changelog
2018-06-22 Lonestar <lonestar@ices.utexas.edu>
* version 4.0
* released along with Galois v4.0
* added Betweenness Centrality
* added Breadth-first Search
* added Connected Components
* added PageRank
* added Independent Set
* added Maximum Cardinality Bipartite Matching
* added Points-to Analysis
* added Triangle Counting
* added Stochastic Gradient Descent
2011-06-01 Lonestar <lonestar@ices.utexas.edu>
* version 3.0
* merged with Galois v2.0
* added GMetis
* added Kruskal's Algorithm
* added Single-source Shortest Paths
2009-12-01 Lonestar <lonestar@ices.utexas.edu>
* version 2.1
* added Boruvka
* added Eventdriven Simulation
* added Preflow Push
* added Prim
* reimplemented Survey Propagation
2009-04-01 Lonestar <lonestar@ices.utexas.edu>
* version 2.0
* added Agglomerative Clustering
* added Survey Propagation
* reimplemented Delaunay Triangulation
* updated Barnes-Hut N-Body Simulation
* updated Delaunay Mesh Refinement
2008-04-01 Lonestar <lonestar@ices.utexas.edu>
* version 1.1
* updated Focused Communities
2008-02-01 Lonestar <lonestar@ices.utexas.edu>
* version 1.0
* added Focused Communities
2007-12-01 Lonestar <lonestar@ices.utexas.edu>
* version 0.9
* added Barnes-Hut N-Body Simulation
* added Delaunay Mesh Refinement
* added Delaunay Triangulation
Related Publications
If you find this software useful in academic work, please acknowledge Lonestar and cite following publication: