RESEARCH USE
Where Stan fits
Stan expresses probabilistic models in a dedicated language and runs them through CmdStan or interfaces for R, Python, Julia, and other environments. Define data, parameter constraints, and generated quantities explicitly, then evaluate prior and posterior predictions. Report chains, warmup, iterations, R-hat, effective sample sizes, and any divergences.
Research tasks
- Write reusable Bayesian probability models
- Run HMC and NUTS posterior sampling
- Diagnose models and generate predictions through language interfaces
What to evaluate before use
- Divergences, low effective sample sizes, or problematic R-hat values require investigation; simply drawing more samples may not fix the model.
- Parameterization, priors, and variable scales strongly affect identifiability and computation, and the model still requires domain validation.
Verification note
This entry summarizes the tool's role without assessing scientific accuracy or endorsing its outputs. Features and terms can change; consult the official source before adopting it for consequential work.
Last verified: 2026-09-05
Source: official documentation ↗