-
Notifications
You must be signed in to change notification settings - Fork 2
Fitting Models with Stan
Stan is a powerful software tool for the implementation of Bayesian models. Because Stan is written in C++, it is compatible with a suite of programming languages including Python, R, MATLAB, Julia, Scala, and more. For detailed installation instructions, see here.
Stan code uses syntax similar to C++. Stan code files are identified by the .stan file extension. See the Stan language documentation for an overview of the language. Case studies and tutorials are available on the Stan website as well. The stan_code directory in this repo contains the code for all of the models presented in the manuscript.
PyStan is the Python Stan interface. All results of this work were generated using PyStan. In this repo, the notebook fit_models.ipynb in the notebooks directory implements our models and outputs the results in the results directory. The data used for model fitting, ground truth, and cross validation is found in the data directory and its corresponding sub-directories. The notebook figures.ipynb in the notebooks directory takes the model results and outputs the figures we use in the paper into the figures directory.