Modified anti-Gauss and degree optimal average formulas for Gegenbauer measure (Q2469136)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Modified anti-Gauss and degree optimal average formulas for Gegenbauer measure
scientific article

    Statements

    Modified anti-Gauss and degree optimal average formulas for Gegenbauer measure (English)
    0 references
    0 references
    4 February 2008
    0 references
    Let \(Q^G_n[f]=\sum_{k=1}^n w_k^G f(x_k^G)\) be the Gauss formula which approximates the integral \(I_\omega[f]=\int_a^b f(x)\omega(x)\,dx\). Let \(R^G_n[f]=I_\omega[f]-Q^G_n[f]\). Suppose that the functional \(Q_m[f]=\sum_{k=1}^n \tilde{w}_k f(\tilde{x}_k)\) provides an acceptable estimate of the quadrature error: \(|R^G_n[f]|\approx |Q_m[f]-Q^G_n[f]|\). \(Q_m[f]\) is called stopping functional and its study is the core of this paper. More precisely, the properties of the modified anti-Gauss formula \(Q_{n+1}^{AG}\) and the corresponding modified average formula \(Q_{2n+1}=((1+\gamma)Q_n^G+Q_{n+1}^{AG})\), \(\gamma>-1\), for the Gegenbauer measure \(d\mu=(1-x^2)^{\alpha-1/2}\,dx\), \(x\in [-1,1]\), \(\alpha>-1/2\) are investigated. One of the main results in the article is the following Theorem. For \(n\geq 2\), the (\(n+1\))-point stratified extension \(Q^C_{2n+1,\gamma^*}\) of the highest possible degree of the \(n\)-point Gauss-Gegenbauer formula is unique and is obtained for \(\gamma=\gamma^*\), where \[ \gamma^*=\frac{2\alpha (\alpha-1)}{n(n+2\alpha -1)(n+\alpha +1)}. \] Besides \(\deg(Q^C_{2n+1,\gamma^*})=2n+3\). The author remarks that he has followed a method different from that previously introduced by \textit{M. M. Spalević} in [Math. Comput. 76, No. 259, 1483--1492 (2007; Zbl 1113.65025)].
    0 references
    0 references
    0 references
    0 references
    0 references
    Gauss quadrature formulas
    0 references
    anti-Gauss formulas
    0 references
    error estimation
    0 references
    Gegenbauer polynomials
    0 references
    0 references