>>12625422It depends on the function.
If finding recursive derivatives at a boundary point of the function is inexpensive, and the interval of integration isn’t too long, then this technique can work.
On the other hand, if finding recursive derivatives is as expensive, or more expensive than function evaluations, and if the interval of integration is very long, then taylor series is a pretty dumb way of solving the problem; using interpolating polynomials based on intelligently chosen nodes, like Chebyshev nodes, can get much better results.
