[ODE] Large mass and accuracy

Jon Watte hplus-ode at mindcontrol.org
Tue Jun 15 16:35:29 MST 2004


> The key word here is ratios. 

As far as I understand it, there is a SEPARATE numerical problem, 
even if the ratio is small, when the values themselves are far away 
from 1.0, because you also work with inverse values, and you start 
chewing up your available bits of relative precision pretty quickly.

Cheers,

			/ h+




More information about the ODE mailing list