The ranked programming library is written for use with the Racket programming language, which is based on the Scheme dialect of LISP. For a more general introduction to Racket please consult Racket's own getting started guide. In this document we provide some instructions to get started with ranked programming in Racket. For a more detailed description, please read the paper.Ī reference manual for this package can be found here. However, instead of using probabilities, one expresses uncertainty by specifying what happens 'normally' and what happens 'exceptionally'. Like probabilistic programming, ranked programming provides a simple and flexible way to represent models with uncertain behaviour, and to perform inference with such models. In short, ranked programming is similar to probabilistic programming, except that the underlying uncertainty formalism is replaced with ranking theory, which measures uncertainty using degrees of surprise on the integer scale from 0 to ∞. This paper was presented at the IJCAI 2019 conference (August 10-16, Macao, China). The ranked-programming package is an implementation of the approach described in the paper Ranked Programming. Racket is a programming language based on the Scheme dialect of LISP. This repository contains the ranked-programming package for use with Racket.
