Am Donnerstag, den 27.08.2009, 10:54 -0500 schrieb Barton Willis:
> It would be great if Maxima could determine the limits of the digamma
> function. I have some code for summation that would benefit:
>
> (%i29) simpsum : true$
>
> (%i30) sum(1/k,k,1,n);
> (%o30) psi[0](n+1)+%gamma
>
> (%i31) sum(1/k^2,k,12,n);
> (%o31) (25613280*%pi^2-239437889)/153679680-psi[1](n+1)
>
> (%i32) sum(1/k^2,k,12,inf);
> (%o32) (25613280*%pi^2-239437889)/153679680-limit(psi[1](k),k,inf)
I am working at the limit algorithm because I am interested to
strengthen defint.
I have already found some possibilities to improve the limits of the
factorial and the log functions. There is more work possible to
strengthen the limits for more special functions.
I had a look at the psi function too. Perhaps I can present the next
days some code to improve this function too.
Dieter Kaiser