problem with symbolic integration
- To: mathgroup at smc.vnet.net
- Subject: [mg16552] problem with symbolic integration
- From: gpenn at socrates.berkeley.edu (Gregg Penn)
- Date: Tue, 16 Mar 1999 04:00:23 -0500
- Organization: University of California, Berkeley
- Sender: owner-wri-mathgroup at wolfram.com
I am trying to evaluate integrals of the form
Integrate[BesselJ[n,x]^2/(x-z),{x,-Infinity,Infinity}]
where z is a complex number. When I force the integral to be
done purely numerically, it works reasonably well. However,
when using a symbolic solution (involving Meijer G-functions)
the answer is off by a factor of 2 and also blows up for large |z|.
For example, I got nonsense answers for z=-15 I.
What's going on? Is there a better way to evaluate this integral?
Thanks,
Gregg Penn
UC Berkeley