bigfloats of special functions (e.g. Bessel).. wasRe: integrate bessel_j errcatch?



On 12/8/2011 11:05 AM, Raymond Toy wrote:
>  I looked sometime ago for some algorithms for bessel functions and 
> only found one using Hadamard series 
> (http://www.sciencedirect.com/science/article/pii/S0377042708001799) 
> and some variations thereof.  Never got them to converge but I only 
> spent a short time on them.  The algorithms weren't difficult, but I 
> was not smart enough to get them to converge.
>
> Ray
I just glanced at it and although I may be misunderstanding, my 
impression is that
this paper is aimed at finding an acceptable absolute error, which does 
not really do the
job -- you need acceptable relative error (precision, that is).  How 
hard is it
to evaluate those series accurately??

I suppose the same analysis techniques could be used though... yet the 
approaches from
numerical analysts seem so different from this.

For example, evaluation of a Bessel function near a zero of that 
function needs some
special work.  One approach is to expand in a Taylor series about that 
zero...