sum of 1/A007504(n)

[Max Alekseyev]

On 5/16/07, Don Reble <djr at nk.ca> wrote:
>     Does anyone know how to calculate the digamma function, _very_
>     accurately, for arbitrary (irrational) arguments?

I do not know specifics of digamma, but there is a generic trick that
can help track the precision of numerical computations in a
(black-box) math system (like maple, mathematica etc.).
First, your compute something with certain precision. Then you double
the precision and compute the very same thing. Comparing the former
and the latter results will give an idea of how accurate are your
computations.

Regards,
Max