Profiling for laziness

Abstract

While many programmers appreciate the benefits of lazy programming at an abstract level, determining which parts of a concrete program to evaluate lazily poses a significant challenge for most of them. Over the past thirty years, experts have published numerous papers on the problem, but developing this level of expertise requires a significant amount of experience. We present a profiling-based technique that captures and automates this expertise for the insertion of laziness annotations into strict programs. To make this idea precise, we show how to equip a formal semantics with a metric that measures waste in an evaluation. Then we explain how to implement this metric as a dynamic profiling tool that suggests where to insert laziness into a program. Finally, we present evidence that our profiler's suggestions either match or improve on an expert's use of laziness in a range of real-world applications.

Promises are made to be broken: migrating R to strict semantics

“…The authors report an average 8.5% speedup (with a maximum speedup of 89%). Chang and Felleisen [2014] solve the complementary problem of suggesting laziness annotations for call-by-value 𝜆 calculus using dynamic analysis. They introduce the notion of laziness potential, a predictor of the benefit obtained by making an expression lazy.…”

Is functional programming better for modularity?

“…We have found similar situations in at least one paper per year, from 2010 to 2014, where Why FP is used as an authoritative reference to the modularity benefits of functional programming. We start with the most recent article, in 2014 at POPL [4]:…”

“…The complete material used in this experiment is available at [9]. Using the topicmodels R package 4 we used the latent Dirichlet allocation (LDA) algorithm on the ghc/compiler subsystem of GHC to construct scattering and tangling curves as in [2].…”

“…Regarding informal citations, Google Scholar shows a constant following since its publication, with a notable increase since 2006. 2 Albeit 5 of the most recent citations in 2014 refer to Why FP in a neutral or historical context, we became intrigued by this quote [4]:…”

Data-driven ecosystem migration: non-intrusive migration of R ecosystem from lazy to strict semantics

“…The authors report an average 8.5% speedup (with a maximum speedup of 89%). Chang and Felleisen [5] solve the complementary problem of suggesting laziness annotations for call-by-value λ calculus using dynamic analysis. They introduce the notion of laziness potential, a predictor of the benefit obtained by making an expression lazy.…”