Stable approximation of functions from equispaced samples via Jacobi frames
In this paper, we study the Jacobi frame approximation with equispaced samples and derive an error estimation. We observe numerically that the approximation accuracy gradually decreases as the extended domain parameter $γ$ increases in the uniform norm, especially for differentiable functions. In addition, we show that when the indexes of Jacobi polynomials $α$ and $β$ are larger (for example $\max\{α,β\} > 10$), it leads to a divergence behavior on the frame approximation error decay.