Graph generator with tunable parameters based on the degree-corrected stochastic blockmodel.
Notes:
- The generator is non-deterministic; it generates a different graph each time it is run.
- The graph parameters are controlled indirectly; they will not match the inputs exactly.
- For more extreme graph parameter combinations, the generator may not be able to fit a matching stochastic blockmodel and the generation may fail. Re-running the generator a few times may fix the issue, but for the most part the best solution is to use less extreme parameters.
Requirements:
- The requirements for running this generator are listed in
requirements.txt. - All requirements should be installable via pip, but from our experience, the
graph-toolpackage can be tricky to install. The following instructions worked for our system:
conda config --add channels conda-forge
conda config --add channels pkgw-forge
conda install -c pkgw-forge gtk3
conda install -c vgauthier graph-tool
For more information, refer to the graph-tool installation instructions
Usage:
To generate a graph locally, use python generate.py <args>. Note that generating graphs with more than a few million vertices/edges can take several hours.
!WARNING: EXPERIMENTAL! To generate a graph on a remote system, use python generate.py --remote <remote_url> <args>. The generator uses the Paramiko and getpass packages for authentication. You will need to provide your username and password on the command-line. Systems that do not support a username + password authentication method are not supported.
© Virginia Polytechnic Institute and State University, 2022.
Please refer to the included LICENSE file.