Giving Haskell a promotion

Yorgey, Brent A.; Weirich, Stephanie; Cretin, Julien; Jones, Simon Peyton; Vytiniotis, Dimitrios; Magalhães, José Pedro

doi:10.1145/2103786.2103795

Cited by 141 publications

(84 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, they have also started to be incorporated in more standard programming languages, such as Haskell [40]. In all these approaches the types can express richer properties of the program that can then be ensured automatically by type checking.…”

Section: Related Workmentioning

confidence: 99%

Linear dependent types for differential privacy

GaboardiMarco¹,

HaeberlenAndreas²,

HsuJustin³

et al. 2013

SIGPLAN Not.

View full text Add to dashboard Cite

Differential privacy offers a way to answer queries about sensitive information while providing strong, provable privacy guarantees, ensuring that the presence or absence of a single individual in the database has a negligible statistical effect on the query's result. Proving that a given query has this property involves establishing a bound on the query's sensitivity-how much its result can change when a single record is added or removed.A variety of tools have been developed for certifying that a given query is differentially private. In one approach, Reed and Pierce [34] proposed a functional programming language, Fuzz, for writing differentially private queries. Fuzz uses linear types to track sensitivity and a probability monad to express randomized computation; it guarantees that any program with a certain type is differentially private. Fuzz can successfully verify many useful queries. However, it fails when the sensitivity analysis depends on values that are not known statically.We present DFuzz, an extension of Fuzz with a combination of linear indexed types and lightweight dependent types. This combination allows a richer sensitivity analysis that is able to certify a larger class of queries as differentially private, including ones whose sensitivity depends on runtime information. As in Fuzz, the differential privacy guarantee follows directly from the soundness theorem of the type system. We demonstrate the enhanced expressivity of DFuzz by certifying differential privacy for a broad class of iterative algorithms that could not be typed previously.

show abstract

Section: Related Workmentioning

confidence: 99%

Linear dependent types for differential privacy

GaboardiMarco¹,

HaeberlenAndreas²,

HsuJustin³

et al. 2013

SIGPLAN Not.

View full text Add to dashboard Cite

show abstract

“…A black and white version is available from http://dreixel.net/research/pdf/ trkgp_nocolor.pdf. Yorgey et al (2012) introduce two new features in GHC: userdefined kinds (through datatype promotion) and kind polymorphism. We introduce the two features in this section by means of example uses, and refer the reader to the original paper for implementation details.…”

Section: Notationmentioning

confidence: 99%

“…Visible examples of this are the dataCast 1 /dataCast 2 functions in the Scrap Your Boilerplate approach (SYB, Peyton Jones 2003, 2004), or the Generic/Generic 1 classes of Magalhães et al (2010a). This is all bound to change, though, with the recent addition of kind polymorphism and user-defined kinds to GHC (Yorgey et al 2012). The new features, available as a "technology preview" in GHC version 7.4, and due to be fully supported in version 7.6, bring a whole new range of programming techniques within reach.…”

Section: Introductionmentioning

confidence: 99%

The right kind of generic programming

Magalhães

2012

Proceedings of the 8th ACM SIGPLAN Workshop on Generic Programming

Self Cite

View full text Add to dashboard Cite

Haskell is known for its strong, static type system. A good type system classifies values, constraining the valid terms of the language and preventing many common programming errors. The Glasgow Haskell Compiler (GHC) has further extended the type language of Haskell, adding support for type-level computation and explicit type equality constraints.Type-level programming is used to classify values, but types themselves have remained notoriously unclassified. Recently, however, the kind-level language was extended with support for userdefined kinds and kind polymorphism. In this paper we revisit generic programming approaches that rely heavily on type-level computation, and analyse how to improve them with the extended kind language. For instructive purposes, we list a series of advantages given by the new features, and also a number of shortcomings that prevent desirable use patterns.

show abstract

“…As such, the design of generic-deriving could be kept relatively simple, at the loss of some generality. Recently, with the advent of kind polymorphism in GHC [16], many libraries are being generalised from kind-specific to kind-polymophic (e.g. Typeable [16]).…”

Section: Introductionmentioning

confidence: 99%

“…Recently, with the advent of kind polymorphism in GHC [16], many libraries are being generalised from kind-specific to kind-polymophic (e.g. Typeable [16]). In this paper we describe an elegant generalisation of generic-deriving that works with multiple parameters, lifting the one-parameter restriction without requiring the full-blown power (and complexity) of indexed functors [7].…”

Section: Introductionmentioning

confidence: 99%

Generic Programming with Multiple Parameters

Magalhães

2014

Functional and Logic Programming

Self Cite

View full text Add to dashboard Cite

Abstract. Generic programming, a form of abstraction in programming languages that serves to reduce code duplication by exploiting the regular structure of algebraic datatypes, has been present in the Haskell language in different forms for many years. Lately, a library for generic deriving has been given native support in the compiler, allowing programmers to write functions such as fmap that abstract over one datatype parameter generically. The power of this approach is limited to dealing with one parameter per datatype, however. In this paper, we lift this restriction by providing a generalisation of generic deriving that supports multiple parameters, making essential use of datatype promotion and kind polymorphism. We show example encodings of datatypes, and how to define a map function that operates on multiple parameters simultaneously.

show abstract

Giving Haskell a promotion

Cited by 141 publications

References 29 publications

Linear dependent types for differential privacy

Linear dependent types for differential privacy

The right kind of generic programming

Generic Programming with Multiple Parameters

Contact Info

Product

Resources

About