Python interface to bnlearn and other probabilistic graphical model libraries
Before you can install this library you have to have a working python3 version and a working R version plus rpy2 pre-installed.
Have a look at pybnl/environment/Dockerfile for details. This docker image is based on jupyter/r-notebook, which is basically a miniconda environment plus some R basics like:
conda install -c r r-essentials rpy2
Once you have the pre-requisites fulfilled installation can be done as follows:
python -m pip install --upgrade pip setuptools wheel
git clone https://github.com/cs224/pybnl.git
cd pybnl
pip install -e .
If you do not plan to do development work on this library yourself then you can also install it directly from github:
python -m pip install --upgrade pip setuptools wheel
pip --no-cache-dir install git+https://github.com/cs224/pybnl.git@master