On Generating Discrete Orthogonal Bivariate Polynomials

被引:0
|
作者
Marko Huhtanen
Rasmus Munk Larsen
机构
[1] Stanford University,SCCM program, Computer Science Department
[2] Stanford University,SCCM program, Computer Science Department, and SOI
来源
BIT Numerical Mathematics | 2002年 / 42卷
关键词
Lanczos algorithm; bivariate polynomial; slowly growing length of the recurrence; curve fitting; least squares approximation; bivariate interpolation;
D O I
暂无
中图分类号
学科分类号
摘要
In this paper we present an algorithm for recursively generating orthogonal bivariate polynomials on a discrete set S ⊂ ∝2. For this purpose we employ commuting pairs of real symmetric matrices H, K ∈ ∝n×n to obtain, in a certain sense, a two dimensional Hermitian Lanczos method. The resulting algorithm relies on a recurrence having a slowly growing length. Practical implementation issues an applications are considered. The method can be generalized to compute orthogonal polynomials depending on an arbitrary number of variables.
引用
收藏
页码:393 / 407
页数:14
相关论文
共 50 条