KaGen
KaGen copied to clipboard
KaGen: Communication-free Massively Distributed Graph Generators
It would be nice to have an option for edge weights based on the distance between adjacent vertices.
The test cases are for small graphs and work for 1, 2 and 4 PEs. For other numbers of PEs, segmentation faults can occur irregularly because coordinates or edges are...
Currently, graphs are always generated with 64 bit edge and vertex IDs. Would be nice if we had a CMake flag to use 32 bit data types instead.
Currently, KaGen can only use one thread per MPI process. Since the generators are (mostly) communication free, we could implement additional OpenMPI parallelization without too much hustle.
There might still be happening weird things when the number of vertices to permute is zero, we should rename `most_significant_bit_set` to `most_significant_bit`, and it should probably return an optional. But...
Upgrades the EdgeRange iterator from forward-only to bidirectional, enabling backward iteration through graph edges. ## Changes - **Iterator category**: Upgraded from `std::forward_iterator_tag` to `std::bidirectional_iterator_tag` - **Backward iteration operators**: Implemented `operator--()`...
We maybe want to have a Spack package. This might require some dependency cleanup.
## RMAT, GNM already undirected ## RGGs - draw a random radius per vertex from a radius range and set cell size to the maximum radius. Then we do not...