Computational Aspects of Geometric Algebra Products of Two Homogeneous Multivectors

Stéphane Breuils (1) , Vincent Nozick (2) , Akihiro Sugimoto (1)

1 NII - National Institute of Informatics
2 LIGM - Laboratoire d'Informatique Gaspard-Monge

Studies on time and memory costs of products in geometric algebra have been limited to cases where multivectors with multiple grades have only non-zero elements. This allows to design efficient algorithms for a generic purpose; however, it does not reflect the practical usage of geometric algebra. Indeed, in applications related to geometry, multivectors are likely to be full homogeneous, having their non-zero elements over a single grade. In this paper, we provide a complete computational study on geometric algebra products of two full homogeneous multivectors, that is, the outer, inner, and geometric products of two full homogeneous multivectors. We show tight bounds on the number of the arithmetic operations required for these products. We also show that algorithms exist that achieve this number of arithmetic operations.

Mots clés: Mathematics Subject Classification (2010). Primary 99Z99; Secondary 00A00 Geometric Algebra; Clifford Algebra; Computational complexity; Arithmetic operations

Domaines: Logiciel mathématique [cs.MS]

Stéphane Breuils, Vincent Nozick, Akihiro Sugimoto. Computational Aspects of Geometric Algebra Products of Two Homogeneous Multivectors. 2021. ⟨hal-03169013v1⟩