Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Watch this 31-minute conference presentation from the ARRAY 2025 workshop that introduces Star, a novel calculus for array programming that addresses the limitations of current type systems in array programming languages. Learn how researchers Jakub Bachurski, Alan Mycroft, and Dominic Orchard from the University of Cambridge and University of Kent propose using structural record and variant types with subtyping to create richer array types that capture structure explicitly. Discover how this approach prevents indexing errors without resolving arithmetic problems, offering a middle ground between nearly untyped systems that only distinguish arrays by element type or dimensions and complex dependent type systems. Explore the theoretical foundations of algebraic subtyping and its potential for efficient ML-style polymorphic type inference in practical array programming applications for data science and machine learning. Gain insights into how this research contributes to improving type safety, abstractions, and optimizations in array programming while maintaining practical usability for developers working with dynamic typing paradigms.
