Accurate and efficient evaluation of Chebyshev tensor product surface (Q1992395)

Summary: A Chebyshev tensor product surface is widely used in image analysis and numerical approximation. This article illustrates an accurate evaluation for the surface in form of Chebyshev tensor product. This algorithm is based on the application of error-free transformations to improve the traditional Clenshaw Chebyshev tensor product algorithm. Our error analysis shows that the error bound is \(u + \mathcal{O} \left(u^2\right) \times \mathtt{cond} \left(P, x, y\right)\) in contrast to classic scheme \(u \times \mathtt{cond}(P, x, y)\), where \(u\) is working precision and \(\mathtt{cond} \left(P, x, y\right)\) is a condition number of bivariate polynomial \(P(x, y)\), which means that the accuracy of the computed result is similar to that produced by classical approach with twice working precision. Numerical experiments verify that the proposed algorithm is stable and efficient.