Skip to content

Faster MIS reductions through parallelism, dependency checking and reduction tracking

Notifications You must be signed in to change notification settings

Hespian/ParFastKer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ParFastKer

This is the Code written for the paper "Scalable Kernelization for Maximum Independent Sets" if you would like to use it for your publication, please cite the following paper:

Hespe, Demian, Christian Schulz, and Darren Strash. "Scalable kernelization for maximum independent sets." Journal of Experimental Algorithmics (JEA) 24.1 (2019): 1-16.

In the above paper, we first kernelize the graph using the LinearTime algorithm from https://github.com/LijunChang/Near-Maximum-Independent-Set

For partitioning, we used ParHIP (https://github.com/schulzchristian/KaHIP)

Installation

mkdir build
cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
make

Usage

./benchmark [input file] --partition_path=[partition file] --output=[output file] --console_log

[input file] should be an unweighted graph in the METIS graph format

[partition file] should be a file containing one line for each vertex in the graph specifying it's partition index, with the first index being 0. The file name should have the form [number of blocks].partition

License

All files are under the MIT license, except for src/MaximumMatching.cpp which was released under the BSD 3-clause license by the original authors

About

Faster MIS reductions through parallelism, dependency checking and reduction tracking

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages