We appreciate contributions from the community! Some of the ways to contribute are:
- Bug fixes, whether they be for performance or correctness
- Support for additional input formats
- Additional kernel implementations
Although the GAP Benchmark Suite only specifies 6 graph kernels, we are open to adding more non-benchmark kernels to this repo. It is our goal to provide a set of portable high-performance baselines, so naturally there are plenty of remaining graph problems (e.g. MST, SCC, ALS, etc...).
For code contributions, please focus on code simplicity and readability. This repo follows nearly all of the Google C++ Style Guide. If you open a pull request, our continuous-integration service run by Travis will do an immediate sanity-check.
Before embarking on a large development task, feel to open an issue or email (sbeamer@lbl.gov) to ensure you are on a profitable path.