Hamiltonian Monte Carlo in Stata

About the book

Synopsis

Harness modern Bayesian computation without leaving Stata. Hamiltonian Monte Carlo in Stata introduces a unified framework for Bayesian regression modeling through the bayeshmc package, which brings Stan’s Hamiltonian Monte Carlo and the No-U-Turn Sampler to the familiar bayes: prefix idiom — the same models, fitted with a far better sampler.

Built for applied researchers, it spans forty-eight regression model families with full posterior distributions, automatic NUTS adaptation, generated Stan code for every model, four-panel MCMC diagnostics, five covariance priors for random effects, parallel execution, and WAIC and LOO-CV model comparison — bridging everyday Stata workflows with CmdStan-powered computation.