Installation

The Python package MacSGP is publicly available at https://github.com/YangLabHKUST/MacSGP/. It’s recommended to first create a virtual environment.

conda create -n MacSGP python=3.11
conda activate MacSGP

Requirements

MacSGP requires pytorch <https://pytorch.org/get-started/locally/> and PyG <https://pytorch-geometric.readthedocs.io/en/latest/install/installation.html>.

For PyG, MacSGP also requires its additional libraries, their installation requires specifications for torch version and CUDA version. Users could use nvcc --version to check the CUDA version for installation.

Here we provide with an example of the CUDA 12.8 installation code.

pip install torch_geometric

# Additional dependencies:
pip install pyg_lib torch_scatter torch_sparse torch_cluster torch_spline_conv -f https://data.pyg.org/whl/torch-2.8.0+cu128.html

Installing MacSGP from PyPI

MacSGP can be directly installed from PyPI

pip install MacSGP

Installing MacSGP from Github

Alternatively, MacSGP can be downloaded from GitHub:

# Clone the repository
git clone https://github.com/YangLabHKUST/MacSGP.git
cd MacSGP

# Install the required packages
install -r requirements.txt

# Install MacSGP
python setup.py build
python setup.py install