Implementation of a Python MCMC gibbs-sampler with adaptive stepping.
While this is a simple MCMC algorithm, it is robust and stable and well suited to high dimensional problems with many degrees of freedom and very sharp likelihood features. For instance kinematic modelling of datacubes with this code has been found to be orders of magnitude quicker than using more advanced affine-invariant MCMC methods.
You can install GAStimator with pip install gastimator. Alternatively you can download the code here, navigate to the directory you unpack it too, and run python setup.py install.
It requires the following modules:
- numpy
- matplotlib
- plotbin
- joblib
To get you started, see the walk through here: https://github.com/TimothyADavis/GAStimator/blob/master/documentation/GAStimator_Documentation.ipynb
Copyright 2019 Timothy A. Davis
Built by Timothy A. Davis <https://github.com/TimothyADavis>. Licensed under
the GNU General Public License v3 (GPLv3) license (see LICENSE).