This is the Python version of hBayesDM (hierarchical Bayesian modeling of Decision-Making tasks), a user-friendly package that offers hierarchical Bayesian analysis of various computational models on an array of decision-making tasks. hBayesDM in Python uses PyStan (Python interface for Stan) for Bayesian inference.
It supports Python 3.5 or higher versions and requires several packages including: NumPy, SciPy, Pandas, PyStan, Matplotlib, and ArviZ.
Warning
The current Python implementation depends on PyStan 2, which is not the latest version (PyStan 3.*). In the matter of fact, the latest version of PyStan has different interfaces from those in PyStan 2, and it does not support Windows for now. In these points, we developers are concerned that it can affect the availability of hBayesDM for Windows users, so instead of updating hBayesDM to use PyStan 3, we plan to use cmdstanpy for our backend in a near future. Until then, we strongly recommend you to use the R version instead, but you can still use the current Python implementation with PyStan 2.19.1.1. Apologies for the inconvenience, and please stay tuned for the future update.
- Documentation: http://hbayesdm.readthedocs.io/
You can install hBayesDM from PyPI with the following line:
pip install "pystan==2.19.1.1" # Use PyStan 2, for now
pip install hbayesdm # Install using pipIf you want to install the development version:
pip install "git+https://github.com/CCS-Lab/hBayesDM.git@develop#egg=hbayesdm&subdirectory=Python"If you used hBayesDM or some of its codes for your research, please cite this paper:
@article{hBayesDM,
title = {Revealing Neurocomputational Mechanisms of Reinforcement Learning and Decision-Making With the {hBayesDM} Package},
author = {Ahn, Woo-Young and Haines, Nathaniel and Zhang, Lei},
journal = {Computational Psychiatry},
year = {2017},
volume = {1},
pages = {24--57},
publisher = {MIT Press},
url = {doi:10.1162/CPSY_a_00002},
}