On Monday, February 2, 2015 at 11:17:36 AM UTC-5, Jiahao Chen wrote: > > The entire point of orthogonal polynomials is that they can be evaluated > very efficiently using their recurrence relations. I would make use of this > fact to the extent possible allowed by numerical roundoff. >
(In a serious package, I would tend to want to see additional tricks, e.g hard-coding the low-order base cases with @evalpoly, combining multiple recurrence stages to coarsen the loop body, etc.)