miniVite | OpenMP implementation of the first phase

miniVite is a C++ library. miniVite has no bugs, it has no vulnerabilities, it has a Permissive License and it has low support. You can download it from GitHub.

miniVite[*] is a proxy app that implements a single phase of Louvain method in distributed memory for graph community detection. Please refer to the following paper for a detailed discussion on distributed memory Louvain method implementation: Apart from real world graphs, users can use specific options to generate a Random Geometric Graph (RGG) in parallel. RGGs have been known to have good community structure: The way we have implemented a parallel RGG generator, vertices owned by a process will only have cross edges with its logical neighboring processes (each process owning 1x1/p chunk of the 1x1 unit square). If MPI process mapping is such that consecutive processes (for e.g., p and p+1) are physically close to each other, then there is not much communication stress in the application. Therefore, we allow an option to add extra edges between randomly chosen vertices, whose owners may be physically far apart. We require the total number of processes to be a power of 2 and total number of vertices to be perfectly divisible by the number of processes when parallel RGG generation options are used. This constraint does not apply to real world graphs passed to miniVite. We also allow users to pass any real world graph as input. However, we expect an input graph to be in a certain binary format, which we have observed to be more efficient than reading ASCII format files. The code for binary conversion (from a variety of common graph formats) is packaged separately with Vite, which is our full implementation of Louvain method in distributed memory. Please follow instructions in Vite README for binary file conversion. Vite could be downloaded from (please don't use the past PNNL/PNL link to download Vite, the following GitHub link is the correct one): Unlike Vite, we do not implement any heuristics to improve the performance of Louvain method. miniVite is a baseline parallel version, implementing only the first phase of Louvain method. This code requires an MPI library (preferably MPI-3 compatible) and C++11 compliant compiler for building.