How do you do partitioning of a graph? Do you have offline (you have infinite time) techniques as well as online (you’re seeing the graph for the first time and have to do it quickly)?
We use vertex randomization (with a hash function) based 2D partitioning. This is quick and robust to graph updates. We currently do not use a more sophisticated partitioning scheme, but we can replace the default partitioning with a more complex strategy if necessary (we don’t see the need yet). Or if you need graph partitioning as a feature, please let us know.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.