1995IEEE Transactions on Automatic ControlRequires access

Halley's method for the matrix sector function

Çetin Kaya Koç, Bertan Bakkaloğlu

Open publisher page 15 citations

Abstract

The matrix n-sector function is a generalization of the matrix sign function; it can be used to determine the number of eigenvalues of a matrix in a specific sector of the complex plane and to extract the eigenpairs belonging to this sector without explicitly computing the eigenvalues. It is known that Newton's method, which can be used for computing the matrix sign function, is not globally convergent for the matrix sector function. The only existing algorithm for computing the matrix sector function is based on the continued fraction expansion approximation to the principal nth root of an arbitrary complex matrix. In this paper, we introduce a new algorithm based on Halley's generalized iteration formula for solving nonlinear equations. It is shown that the iteration has good error propagation properties and high accuracy. Finally, we give two application examples and summarize the results of our numerical experiments comparing Newton's, the continued fraction, and Halley's method.>

About this research paper

What this paper is about

The matrix n-sector function is a generalization of the matrix sign function; it can be used to determine the number of eigenvalues of a matrix in a specific sector of the complex plane and to extract the eigenpairs belonging to this sector without explicitly computing the eigenvalues. It is known that Newton's method, which can be used for computing the matrix sign function, is not globally convergent for the matrix sector function. The only existing algorithm for computing the matrix sector function is based on the continued fraction expansion approximation to the principal nth root of an arbitrary complex matrix. In this paper, we introduce a new algorithm based on Halley's generalized iteration formula for solving nonlinear equations. It is shown that the iteration has good error propagation properties and high accuracy. Finally, we give two application examples and summarize the results of our numerical experiments comparing Newton's, the continued fraction, and Halley's method.>

Why it matters

OpenAlex reports 15 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

The matrix n-sector function is a generalization of the matrix sign function; it can be used to determine the number of eigenvalues of a matrix in a specific sector of the complex plane and to extract the eigenpairs belonging to this sector without explicitly computing the eigenvalues. It is known that Newton's method, which can be used for computing the matrix sign function, is not globally convergent for the matrix sector function. The only existing algorithm for computing the matrix sector function is based on the continued fraction expansion approximation to the principal nth root of an arbitrary complex matrix. In this paper, we introduce a new algorithm based on Halley's generalized iteration formula for solving nonlinear equations. It is shown that the iteration has good error propagation properties and high accuracy. Finally, we give two application examples and summarize the results of our numerical experiments comparing Newton's, the continued fraction, and Halley's method.>

Key concepts: Sign function, Matrix function, Eigenvalues and eigenvectors, Matrix (chemical analysis), Mathematics, Function (biology), Applied mathematics, Complex plane

Related papers

Back to paper searchBrowse research topicsOriginal source
Halley's method for the matrix sector function — Research Paper | ScholarLens