An Algorithm Based on the FFT for a Generalized Chebyshev Interpolation

From MaRDI portal





In the quadrature method of \textit{C. W. Clenshaw} and \textit{A. R. Curtis} (C-C) [Numer. Math. 2, 197-205 (1960; Zbl 0093.140)] the integrand f(t) is interpolated with a sequence of the truncated Chebyshev series, \(p_ N(t)=\sum^{N}_{k=0}a^ N_ kT_ k(t)\). In order to reduce the cost of using a cosine transformation to compute the \(a^ N_ k\), \textit{W. M. Gentleman} (G) [Commun. ACM. 15, 337-346 (1972; Zbl 0234.65024)] incorporated in the program the fast Fourier transformation (FFT). \textit{M. Branders} and \textit{R. Piessens} (B-P) [J. Comput. Appl. Math. 1, 55-65 (1975; Zbl 0303.65020)] extended the method C-C to integrands K(t)f(t) where K(t) is a singular function. The present paper extends the methods G and B-P by an algorithm for recursively generating a sequence of \(p_ N(t)\) which allows an increase of N more moderate than doubling it at each step; it also uses the FFT. Numerical comparison with G and B-P is given.




Cited in
(29)


Describes a project that uses

Uses Software






This page was built for publication: An Algorithm Based on the FFT for a Generalized Chebyshev Interpolation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4203686)