Thanks for your explanation. I'll study the code of this example to get fully understanding.
Best regards, xiupeng On Wed, Mar 31, 2010 at 9:21 AM, Jonathan Guyer <[email protected]> wrote: > > > On Mar 30, 2010, at 7:21 PM, Santan William wrote: > > I just successfully install fipy on two ubuntu computers. One is a >> workstation with 4 cpus, another is a common desktop with a 2GHz CPU. >> When I try to run some examples, they seems o.k, like diffusion/mesh1D >> etc. But when I run this examples.phase.anisotropy , it takes a very long >> time. For workstation, it takes more than 30mins. For that ordinary desktop, >> it takes 2 hours but still not finished. Anyone knows the reason? Is there >> any software I miss? >> > > How long should it take? > > diffusion.mesh1D solves a single diffusion equation, only some with any > non-linearity, on a 50 cell mesh. > > phase.anisotropy solves a pair of coupled, non-linear PDEs on a 500x500 > mesh. I don't think I find it surprising that it should take longer. 5000x > longer might be a good estimate of a lower bound. > >
