Installation

biSBM is a Python library, and it depends on C++ programs for faster SBM inference. We provide two reference engines for this purpose; they have been added as submodules to this repository. But you will still need CMake and Boost Libraries to compile them. Please refer to their official pages for installation instructions. For macOS users, you may want to run brew install for cmake and boost.

To clone this project along with the submodules, do:

git clone https://github.com/junipertcy/bipartiteSBM.git --recursive

Now enter the directory bipartiteSBM. Since the submodules we cloned in the engines folder may be out-dated, let’s run this command to ensure we have all the newest submodule’s content:

git submodule update

These modules are C++ subroutines for graph partitioning. To compile them, please run this shell script:

sh scripts/compile_engines.sh

Lastly, we have to install Python library dependencies, by simply running this command:

python -m pip install -r requirements.txt

If you are good so far, then we are good to go!