An iterative algorithm for spline interpolation (Q685868): Difference between revisions

The well-known Schoenberg-Whitney Theorem [cf. \textit{I. J. Schoenberg} and \textit{A. Whitney}, Trans. Amer. Math. Soc. 74, 246-259 (1953; Zbl 0051.336)] provides a nice characterization of the choices of points which ensure the uniqueness of the interpolating spline. For the explicit computation of such a spline, \textit{C. de Boor} and \textit{A. Pinkus} [Numer. Math. 27, 485-490 (1977; Zbl 0336.65020)] have shown that the solution of the underlying linear system is possible in a numerically stable way by Gaussian elimination. The author proposes here an iterative procedure for computing the coefficients in the B-spline basis representation of the interpolation spline. A generalization of Newton's interpolation formula for polynomials to the case of splines has also been presented [cf. \textit{G. Mühlbach}, Numer. Math. 31, 97-110 (1978; Zbl 0427.65003)].