Probabilistic Programming with Densities in SlicStan: Efficient, Flexible, and Deterministic

Maria I. Gorinova; Andrew D. Gordon; Charles Sutton

doi:10.1145/3290348

Probabilistic Programming with Densities in SlicStan: Efficient, Flexible, and Deterministic

Maria I. Gorinova, Andrew D. Gordon, Charles Sutton

Research output: Contribution to journal › Article › peer-review

Abstract / Description of output

Stan is a probabilistic programming language that has been increasingly used for real-world scalable projects. However, to make practical inference possible, the language sacrifices some of its usability by adopting a block syntax, which lacks compositionality and flexible user-defined functions. Moreover, the semantics of the language has been mainly given in terms of intuition about implementation, and has not been formalised.

This paper provides a formal treatment of the Stan language, and introduces the probabilistic programming language SlicStan --- a compositional, self-optimising version of Stan. Our main contributions are (1) the formalisation of a core subset of Stan through an operational density-based semantics; (2) the design and semantics of the Stan-like language SlicStan, which facilities better code reuse and abstraction through its compositional syntax, more flexible functions, and information-flow type system; and (3) a formal, semantic-preserving procedure for translating SlicStan to Stan.

Original language	English
Article number	35
Pages (from-to)	35:1-35:30
Number of pages	30
Journal	Proceedings of the ACM on Programming Languages
Volume	3
Issue number	POPL
DOIs	https://doi.org/10.1145/3290348
Publication status	Published - 2 Jan 2019

Keywords / Materials (for Non-textual outputs)

information flow analysis
probabilistic programming
Markov chain Monte Carlo method
probabilistic computing

Access to Document

10.1145/3290348Licence: Creative Commons: Attribution (CC-BY)

Probabilistic Programming_GORINOVA_DoA161018_VoR_CC-BY
This work is licensed under a Creative Commons Attribution 4.0 International License. © 2019 Copyright held by the owner/author(s).
Final published version, 478 KBLicence: Creative Commons: Attribution (CC-BY)

https://dl.acm.org/citation.cfm?id=3290348Licence: Creative Commons: Attribution (CC-BY)

Cite this

@article{1a12a86892cf444c8b3b60203f05ebe1,

title = "Probabilistic Programming with Densities in SlicStan: Efficient, Flexible, and Deterministic",

abstract = "Stan is a probabilistic programming language that has been increasingly used for real-world scalable projects. However, to make practical inference possible, the language sacrifices some of its usability by adopting a block syntax, which lacks compositionality and flexible user-defined functions. Moreover, the semantics of the language has been mainly given in terms of intuition about implementation, and has not been formalised.This paper provides a formal treatment of the Stan language, and introduces the probabilistic programming language SlicStan --- a compositional, self-optimising version of Stan. Our main contributions are (1) the formalisation of a core subset of Stan through an operational density-based semantics; (2) the design and semantics of the Stan-like language SlicStan, which facilities better code reuse and abstraction through its compositional syntax, more flexible functions, and information-flow type system; and (3) a formal, semantic-preserving procedure for translating SlicStan to Stan.",

keywords = "information flow analysis, probabilistic programming, Markov chain Monte Carlo method, probabilistic computing",

author = "Gorinova, {Maria I.} and Gordon, {Andrew D.} and Charles Sutton",

year = "2019",

month = jan,

day = "2",

doi = "10.1145/3290348",

language = "English",

volume = "3",

pages = "35:1--35:30",

journal = "Proceedings of the ACM on Programming Languages",

issn = "2475-1421",

publisher = "Association for Computing Machinery (ACM)",

number = "POPL",

}

TY - JOUR

T1 - Probabilistic Programming with Densities in SlicStan: Efficient, Flexible, and Deterministic

AU - Gorinova, Maria I.

AU - Gordon, Andrew D.

AU - Sutton, Charles

PY - 2019/1/2

Y1 - 2019/1/2

N2 - Stan is a probabilistic programming language that has been increasingly used for real-world scalable projects. However, to make practical inference possible, the language sacrifices some of its usability by adopting a block syntax, which lacks compositionality and flexible user-defined functions. Moreover, the semantics of the language has been mainly given in terms of intuition about implementation, and has not been formalised.This paper provides a formal treatment of the Stan language, and introduces the probabilistic programming language SlicStan --- a compositional, self-optimising version of Stan. Our main contributions are (1) the formalisation of a core subset of Stan through an operational density-based semantics; (2) the design and semantics of the Stan-like language SlicStan, which facilities better code reuse and abstraction through its compositional syntax, more flexible functions, and information-flow type system; and (3) a formal, semantic-preserving procedure for translating SlicStan to Stan.

AB - Stan is a probabilistic programming language that has been increasingly used for real-world scalable projects. However, to make practical inference possible, the language sacrifices some of its usability by adopting a block syntax, which lacks compositionality and flexible user-defined functions. Moreover, the semantics of the language has been mainly given in terms of intuition about implementation, and has not been formalised.This paper provides a formal treatment of the Stan language, and introduces the probabilistic programming language SlicStan --- a compositional, self-optimising version of Stan. Our main contributions are (1) the formalisation of a core subset of Stan through an operational density-based semantics; (2) the design and semantics of the Stan-like language SlicStan, which facilities better code reuse and abstraction through its compositional syntax, more flexible functions, and information-flow type system; and (3) a formal, semantic-preserving procedure for translating SlicStan to Stan.

KW - information flow analysis

KW - probabilistic programming

KW - Markov chain Monte Carlo method

KW - probabilistic computing

U2 - 10.1145/3290348

DO - 10.1145/3290348

M3 - Article

SN - 2475-1421

VL - 3

SP - 35:1-35:30

JO - Proceedings of the ACM on Programming Languages

JF - Proceedings of the ACM on Programming Languages

IS - POPL

M1 - 35

ER -

Probabilistic Programming with Densities in SlicStan: Efficient, Flexible, and Deterministic

Abstract / Description of output

Keywords / Materials (for Non-textual outputs)

Access to Document

Fingerprint

Cite this