Numerical Interpolation: Polynomial and Spline

Fit data exactly with the unique degree-n polynomial (Lagrange/Newton) or a smooth C^2 cubic spline, and bound the polynomial error by f^(n+1)/(n+1)! times the node product.