On 11/ 2/09 08:11 PM, svanshaar wrote:
Thanks Jason! That works perfectly.
I don't know if any of this uses the GNU Scientific library, but there are
reports of bug in the Bessel functions on there.
gsl_sf_bessel_lnKnu overflows for large nu
which the developers have acknowledged and
integration. The problem here is that bessel_J is not a symbolic
function, and does not know how to deal with symbolic variables. It
would be great if someone submitted a patch to take care of this!
Yes, in an ideal world we would have already taken care of this, since
this sort of
svanshaar wrote:
I am trying to numerically evaluate the integral of bessel functions.
I've tried constructing it various ways. The one that makes most
sense to me is:
a=var('a')
f=bessel_J(1,a)*bessel_J(0,0.1*a)*e^(-5*a)
f.numerical_integral(a,0,infinity)
I get the error: Cannot