The Stan Math Library: Reverse-Mode Automatic Differentiation in C++

TL;DR

The paper introduces the Stan Math Library, a comprehensive C++ tool for reverse-mode automatic differentiation that simplifies derivative computation in complex algorithms, emphasizing usability, extensibility, efficiency, and portability.

Contribution

It presents a new, extensive C++ library that makes reverse-mode automatic differentiation accessible, efficient, and customizable for optimization and statistical inference tasks.

Findings

Provides a user-friendly C++ interface for automatic differentiation.

Includes extensive built-in functions for matrix, linear algebra, and probability.

Achieves high efficiency and stability through advanced memory management and stable algorithms.

Abstract

As computational challenges in optimization and statistical inference grow ever harder, algorithms that utilize derivatives are becoming increasingly more important. The implementation of the derivatives that make these algorithms so powerful, however, is a substantial user burden and the practicality of these algorithms depends critically on tools like automatic differentiation that remove the implementation burden entirely. The Stan Math Library is a C++, reverse-mode automatic differentiation library designed to be usable, extensive and extensible, efficient, scalable, stable, portable, and redistributable in order to facilitate the construction and utilization of such algorithms. Usability is achieved through a simple direct interface and a cleanly abstracted functional interface. The extensive built-in library includes functions for matrix operations, linear algebra, differential equation solving, and most common probability functions. Extensibility derives from a straightforward object-oriented framework for expressions, allowing users to easily create custom functions. Efficiency is achieved through a combination of custom memory management, subexpression caching, traits-based metaprogramming, and expression templates. Partial derivatives for compound functions are evaluated lazily for improved scalability. Stability is achieved by taking care with arithmetic precision in algebraic expressions and providing stable, compound functions where possible. For portability, the library is standards-compliant C++ (03) and has been tested for all major compilers for Windows, Mac OS X, and Linux.