This is an old version, view current version.


This part of the manual lays of the specification of the Stan programming language. The language is responsible for defining a log density function conditioned on data. Typically, this is a Bayesian posterior, but it may also be a penalized likelihood function.