AlMohy, Awad and Higham, Nicholas J. and Liu, Xiaobo (2021) Arbitrary Precision Algorithms for Computing the Matrix Cosine and its Fréchet Derivative. [MIMS Preprint]
This is the latest version of this item.
Text
paper.pdf Download (724kB) 
Abstract
Existing algorithms for computing the matrix cosine are tightly coupled to a specific precision of floatingpoint arithmetic for optimal efficiency so they do not conveniently extend to an arbitrary precision environment. We develop an algorithm for computing the matrix cosine that takes the unit roundoff of the working precision as input, and so works in an arbitrary precision. The algorithm employs a Taylor approximation with scaling and recovering and it can be used with a Schur decomposition or in a decompositionfree manner. We also derive a framework for computing the \fd, construct an efficient evaluation scheme for computing the cosine and its Fr\'echet derivative simultaneously in arbitrary precision, and show how this scheme can be extended to compute the matrix sine, cosine, and their \fd s all together. Numerical experiments show that the new algorithms behave in a forward stable way over a wide range of precisions. The transformationfree version of the algorithm for computing the cosine is competitive in accuracy with the stateoftheart algorithms in double precision and surpasses existing alternatives in both speed and accuracy in working precisions higher than double.
Item Type:  MIMS Preprint 

Uncontrolled Keywords:  multiprecision algorithm, multiprecision arithmetic, matrix cosine, matrix exponential, matrix function, Fréchet derivative double angle formula, Taylor approximation, forward error analysis, MATLAB 
Subjects:  MSC 2010, the AMS's Mathematics Subject Classification > 15 Linear and multilinear algebra; matrix theory MSC 2010, the AMS's Mathematics Subject Classification > 65 Numerical analysis 
Depositing User:  Nick Higham 
Date Deposited:  26 Nov 2021 12:29 
Last Modified:  26 Nov 2021 12:29 
URI:  https://eprints.maths.manchester.ac.uk/id/eprint/2837 
Available Versions of this Item

Arbitrary Precision Algorithms for Computing the Matrix Cosine and its Fréchet Derivative. (deposited 16 Aug 2021 18:46)
 Arbitrary Precision Algorithms for Computing the Matrix Cosine and its Fréchet Derivative. (deposited 26 Nov 2021 12:29) [Currently Displayed]
Actions (login required)
View Item 