Oct 2026· Proceedings of the 11th ACM SIGPLAN International Workshop on Numerical and Symbolic Abstract Domains· pp. 18-25· 0 citations· 19 references
TL;DR
This paper presents a parametric abstract domain for R vectors, defined over flowR, a static analysis framework for R, and evaluates it on a suite of 61 handcrafted programs covering six categories of vector operations, demonstrating the feasibility and precision of the approach.
Abstract
R is a dynamically typed, vector-oriented language widely used for data analysis. Its vector semantics, including automatic type coercion, recycling, and flexible selection, are pervasive, non-trivial, and deeply intertwined with the semantics of virtually every R operation. The complexity and counter-intuitive nature of this semantics make static reasoning about R programs particularly challenging, and existing tools offer only shallow analyses that fall short of reasoning about R vector-manipulating programs. In this paper, we present a parametric abstract domain for R vectors, defined over , a core calculus we design to capture the important vector operations in R. An abstract vector simultaneously captures the possible lengths of the vector, the values of its elements, and its potential attributes. The domain is parametric in the element abstract domain. We define abstract operators for all operations and equip the domain with a widening operator to ensure termination. We implement the domain on top of flowR, a static analysis framework for R, and evaluate it on a suite of 61 handcrafted programs covering six categories of vector operations, demonstrating the feasibility and precision of the approach.
Rapid Type Analysis (RTA) is an important algorithm used in constructing whole-program call graphs. RTA occupies a special middle ground between precision and speed, making it an algorithm of choice for many industry-scale downstream program analysis tasks. RTA’s core subtyping query, which asks whether a concrete type...
Elton Pinto, Milind Chabbi· Proceedings of the ACM on Pr...· 1 citation
NumPy is a widely used Python library for numerical scientific computing, known for its declarative APIs and its optimized implementations. However, writing efficient NumPy programs, which often entails using vectorized array operations instead of explicit Python loops, may not be straightforward. This can be difficult...
This work introduces DueList, an abstraction-refinement approach geared towards list reasoning, which is implemented on top of off-the-shelf SMT solvers and extends reasoning facilities of existing solvers, allowing to conclude about the (un)satisfiability of a larger range of problems, while outperforming existing sol...
Pierre Goutagny, Aymeric Fromherz, Raphaël Monat· 0 citations
Modern application systems depend on dynamic typing, object binding, deserialization, reflection, symbolic dispatch, plugin discovery, and runtime module loading. These mechanisms are useful because they translate flexible values and metadata into executable behavior, but they also create security failures in which an...
Benchy is a semantic language and execution engine for benchmarking AI programs. A benchmark is completely specified by a program, a scoring function, and a dataset, B=(P,S,D), and is separate from the AI-system taking it; a run binds the two, R=(B,AI). Benchmarks are authored as canonical YAML in which each semantic c...
Francis F. Daniel, Mauro Ibañez, Francis Perelman et al.· 0 citations
Data processing using traditional pattern matching syntax and direct recursive functions is straightforward to write but becomes awkward in ambiguous (i.e., nondeterministic) cases: when programmers wish to avoid backtracking, they often end up having to write complicated code that sacrifices clarity and modularity. Ho...
Lu-Yu Cheng, Florent Ferrari, Michael D. Adams et al.· Proceedings of the ACM on Pr...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.