Error estimates for interpolation with piecewise exponential splines of order two and four (Q2656114)

The approximation of univariate functions is important in many applications, as well as in mathematical theory. Interpolation and quasi-interpolation are useful tools in this context. For this, and in particular for approximations in one dimension, splines (piecewise polynomials) and exponential splines (piecewise exponential functions) are particularly useful. The former can be expressed as a limiting case of the latter by letting parameters in the exponents tend to zero. Both polynomial splines and the exponential splines that are studied in this paper have significant advantages such as locality, a particularly easy local structure, and the ability to provide stable approximations. In this article, the authors consider the special case of orders two and four; error bounds for the piecewise exponential case (exponential splines) are provided as extensions or generalisations of the univariate error bounds for polynomial splines (cubic splines, for example). In order that the results can be applied, the approximands have to be assumed to be sufficiently smooth functions that are to be interpolated but there are no further restrictions.