New exponential variable transform methods for functions with endpoint singularities

DOI10.1137/130920460MaRDI QIDQ2927838zbMATH OpenOpenAlexFDO

Authors Ben Adcock, Mark Richardson

Publication date 4 November 2014

Published in SIAM Journal on Numerical Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1305.2643

zbMATH Keywords

Chebyshev interpolationconformal mapendpoint singularityresolution analysis

Mathematics Subject Classification ID

Interpolation in approximation theory (41A05) Rate of convergence, degree of approximation (41A25) Numerical interpolation (65D05)

Abstract: The focus of this article is the approximation of functions which are analytic on a compact interval except at the endpoints. Typical numerical methods for approximating such functions depend upon the use of particular conformal maps from the original interval to either a semi-infinite or an infinite interval, followed by an appropriate approximation procedure on the new region. We first analyse the convergence of these existing methods and show that, in a precisely defined sense, they are sub-optimal. Specifically, they exhibit poor resolution properties, by which we mean that many more degrees of freedom are required to resolve oscillatory functions than standard approximation schemes for analytic functions such as Chebyshev interpolation. To remedy this situation, we introduce two new transforms; one for each of the above settings. We provide full convergence results for these new approximations and then demonstrate that, for a particular choice of parameters, these methods lead to substantially better resolution properties. Finally, we show that optimal resolution power can be achieved by an appropriate choice of parameters, provided one forfeits classical convergence. Instead, the resulting method attains a finite, but user-controlled accuracy specified by the parameter choice.

Review text

Using the Schwarz-Christoffel formula, two parameterized exponential changes of variables are constructed for mapping analytic functions from their original domain \((0,1]\) to \((-\infty,0]\), or from \((0,1)\) to \((-\infty, \infty)\). The transplanted function on either the infinite or semi-infinite interval is then approximated using domain truncation, followed by Chebyshev interpolation. A thorough analysis of the resulting approximation methods obtains exponential convergence rates for their \(\infty\)-norm error, as well as asymptotic bounds for their resolution power (expressed in terms of the number of degrees of freedom required to represent an oscillatory function to a fixed precision). The results demonstrate that, compared to existing variable transform methods, the new methods not only achieve similar convergence, but can also lead, subject to appropriate parameter choice, to superior resolution properties. Numerical experiments that verify the results are provided and several directions for future work are proposed.

Recommendations

• Resolution-optimal exponential and double-exponential transform methods for functions with endpoint singularities
• Polynomial series versus sinc expansions for functions with corner or endpoint singularities
• Near optimality of the sinc approximation
• Further extension of a class of periodizing variable transformations for numerical integration
• Analytic continuation of some special functions by variable transformation

Cited in

(8)

• Polynomial series versus sinc expansions for functions with corner or endpoint singularities
• Exponential node clustering at singularities for rational approximation, quadrature, and PDEs
• Accurate function sinc interpolation and derivative estimations over finite intervals
• Resolution-optimal exponential and double-exponential transform methods for functions with endpoint singularities
• Analytic continuation of some special functions by variable transformation
• Eliminating the unbounded behavior of function derivative expansions in terms of sinc bases
• Hermite spectral approximation for functions with endpoint singularities using exponential transforms
• Recovering exponential accuracy in Fourier spectral methods involving piecewise smooth functions with unbounded derivative singularities

Describes a project that uses

Uses Software

• Sinc-Pack

This page was built for publication: New exponential variable transform methods for functions with endpoint singularities

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2927838)