Benchmarks
Benchmarks are provided in the source distribution of Minion, together with programs to generate new benchmarks from various problem classes.
As well as this, we aim to archive the specific benchmarks used in papers on Minion.
As well as providing sample sets of benchmarks, this will enhance the reproducibility of our research.
Researchers may run Minion on exactly the instances reported in our papers.
- Instances of the Equidistant Frequency Permutation Array problem. A set of instances used in a SARA 2009 submission. (tgz format, approx. 5MB)
- Benchmarks for all-different. A variety of instances containing the all-different constraint, for testing different propagation methods. (tar bz2 format, approx 1.3MB)
- Benchmarks used in our ECAI 2006 paper. Instances from BIBD, Peg Solitaire, Steel Mill, and Golomb Ruler problem classes. (zip format, approx 2.2MB)
- Benchmarks used in our CP 2006 paper. Instances from Langford’s problem, the Prime Queens problem, Quasigroup using the Element constraint, Quasigroup translated from SAT, and Random SAT instances. (zip format, approx 5.9MB)