Ruprecht-Karls-Universität Heidelberg

Approximate Bayesian Computation with ABrox


This web page is the home of ABrox, a software that makes Approximate Bayesian Computation easy. The software was written by Ulf Mertens and Stefan Radev.

The corresponding GitHub repository can be found here:

Installation instructions

  1. Downloading and installing Python
  2. Go to and download the latest version of Python 3.x. Make sure you check Add Python to PATH in the installer.

  3. Downloading and installing ABrox
  4. Simply type pip install abrox into the console. If you are using Windows and assuming Python version 3.6, type python3.6 -m pip install abrox

    After downloading, type: abrox-gui. The GUI should open.

If you encounter ANY problems during installation of any of the components, do not hesitate to contact us.

Project examples

We provide several templates (.brox files) that can be imported from the GUI via 'Load Project'.

Currently, we provide:

  1. Two-sample t-test (t_test.bprox)
  2. Levene-Test (levene.bprox)
  3. A comparison of multinomial processing tree models (mptree.bprox)

The project-files can be downloaded here: