On piecewise-polynomial approximation of functions with a bounded fractional derivative in an \(L_ p\)-norm (Q810781)

The author studies the error in approximating functions with a bounded \((r+\alpha)th\) derivative in an L-norm. Here r is a nonnegative integer, \(\alpha\in [0,1)\), and \(f^{(r+\alpha)}\) is the classical fractional derivative. The author proves that, for any such function f, there exists a piecewise-polynomial of degree s that interpolates f at n equally spaced points and that approximates f with an error (in the sup-norm) \(\| f^{(r+\alpha)}\|_ pO(n^{-(r+\alpha -1/p)}).\) The author also proves that no algorithm based on n function and/or derivative values of f has this error for any f. This implies the optimality of piecewise-polynomial interpolation. These two results generalize well known results on approximating functions with bounded r-th derivative \((\alpha =0)\).