English

On a Geometric Approach to the Estimation of Interpolation Projectors

Metric Geometry 2023-09-21 v1

Abstract

Suppose Ω\Omega is a closed bounded subset of Rn,{\mathbb R}^n, SS is an nn-dimensional non-degenerate simplex, ξ(Ω;S):=min{σ1:ΩσS}\xi(\Omega;S):=\min \left\{\sigma\geq 1: \, \Omega\subset \sigma S\right\}. Here σS\sigma S is the result of homothety of SS with respect to the center of gravity with coefficient σ\sigma. Let dn+1,d\geq n+1, φ1(x),,φd(x)\varphi_1(x),\ldots,\varphi_d(x) be linearly independent monomials in nn variables, φ1(x)1,\varphi_1(x)\equiv 1, φ2(x)=x1, , φn+1(x)=xn.\varphi_2(x)=x_1,\ \ldots, \ \varphi_{n+1}(x)=x_n. Put Π:=lin(φ1,,φd).\Pi:={\rm lin}(\varphi_1,\ldots,\varphi_d). The interpolation projector P:C(Ω)ΠP: C(\Omega)\to \Pi with a set of nodes x(1),,x(d)x^{(1)},\ldots, x^{(d)} Ω \in \Omega is defined by equalities Pf(x(j))=f(x(j)).Pf\left(x^{(j)}\right)=f\left(x^{(j)}\right). Denote by PΩ\|P\|_{\Omega} the norm of PP as an operator from C(Ω)C(\Omega) to C(Ω)C(\Omega). Consider the mapping T:RnRd1T:{\mathbb R}^n\to {\mathbb R}^{d-1} of the form T(x):=(φ2(x),,φd(x)).T(x):=(\varphi_2(x),\ldots,\varphi_d(x)). We have the following inequalities: 12(1+1d1)(PΩ1)+1 \frac{1}{2}\left(1+\frac{1}{d-1}\right)\left(\|P\|_{\Omega}-1\right)+1 ξ(T(Ω);S)d2(PΩ1)+1. \leq \xi(T(\Omega);S)\leq \frac{d}{2}\left(\|P\|_{\Omega}-1\right)+1. Here SS is the (d1)(d-1)-dimensional simplex with vertices T(x(j)).T\left(x^{(j)}\right). We discuss this and other relations for polynomial interpolation of functions continuous on a segment. The results of numerical analysis are presented.

Keywords

Cite

@article{arxiv.2307.13780,
  title  = {On a Geometric Approach to the Estimation of Interpolation Projectors},
  author = {Mikhail Nevskii and Alexey Ukhalov},
  journal= {arXiv preprint arXiv:2307.13780},
  year   = {2023}
}

Comments

13 pages, 2 figures