This repository contains code for the paper:
Monnahan, C.C., J.T. Thorson, T.A. Branch (In press). Faster estimation of Bayesian models in ecology using Hamiltonian Monte Carlo. Methods in Ecology and Evolution.
This study can be reproduced by installing necessary software and R packages, and then running the script run_simulation.R. See that file for more information.
The models directory contains case studies, with data (as .RDS files), Stan and JAGS implementations (as .stan and .jags files), and an R script to run them (using some functions defined in startup.R). Within each model there is a folder containing plots of results.
The algorithms directory contains an R script containing versions of the static HMC and NUTS algorithms implemented in R. These versions do not contain the full suite of features used in Stan, but may be useful for learning more about their behavior. The script mcmc.R contains the algorithms and helper functions, and examples.R a short script showing how to use them.