Skip to content

Variational Autoencoders Modular Bayesian Networks (VAMBN)

Welcome to the documentation for Variational Autoencoders Modular Bayesian Networks (VAMBN) 2.0. This version features a PyTorch-based HI-VAE (refer to Nazabal et al.'s paper) and employs Snakemake to manage the workflow of Python and R scripts. Have a look at the VAMBN page for an overview of the project.


Getting Started

  1. Follow the installation instructions.
  2. Go through the example in the walkthrough section.
  3. Copy your input data into the data/raw folder according to the description in the walkthrough section.
  4. Configure the vambn_config.yml file according to your needs. Refer to the configuration section for details.
  5. Execute your pipeline locally or on a cluster.
  6. Analyze your results. Refer to the example for explanations.

Development & Bug Tracking

This project is under active development. Expect changes and potential bugs. Please open an issue for any problems you encounter.

License

This software is licensed under the GNU General Public License (GPL) v3 for non-commercial use. For commercial use, please contact Holger Fröhlich to obtain a commercial license.