PyKronecker: A Python Library for the Efficient Manipulation of Kronecker Products and Related Structures

Edward Antonian*, Gareth W. Peters, Michael John Chantler

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Matrix operators composed of Kronecker products and related objects, such as Kronecker sums, arise in many areas of applied mathematics including signal processing, semidefinite programming, and quantum computing (Loan, 2000). As such, a computational toolkit for manipulating Kronecker-based systems, in a way that is both efficient and idiomatic, has the potential to aid research in many fields. PyKronecker aims to deliver this in the Python programming language by providing a simple API that integrates well with the widely-used NumPy library (Harris et al., 2020), and that supports automatic differentiation and accelerated computation on GPU/TPU hardware using Jax (Bradbury et al., 2018)
Original languageEnglish
Number of pages3
JournalThe Journal of Open Source Software
Volume8
Issue number81
DOIs
Publication statusPublished - 30 Jan 2023

Fingerprint

Dive into the research topics of 'PyKronecker: A Python Library for the Efficient Manipulation of Kronecker Products and Related Structures'. Together they form a unique fingerprint.

Cite this