Daniel J Sebald schrieb:
> I've tried the routines ode45 and ode23 recently.  For a linear ODE
> for which I can compute the analytical solution, I found an 11 to 16%
> discrepancy between the numerical and analytical solutions.  Granted,
> I was working with a stiff ODE, but stability seems fine; it's just
> that the solution appears incorrect.

This also strongly depends on which options you set?
RelTol, AbsTol, NormControl, InitialStep etc.

> Thomas Treichl appears to be most associated with the ODE package.
> Thomas, when working on those routines, did you test them in anyway?

I made a few test functions and put them into OdePkg. Their names are 
odepkg_testsuite_*.m, 3 problems are stiff and 3 problems are not stiff and the 
rest are implicit. Try these functions with different solvers...

> I'm wondering if the RK formula may be incorrect, or if it is merely
> that these routines have an issue with stiff ODEs.

The RK algorithms are not very well suited for solving real stiff problems. You 
better use other solver functions from the package that do a better job with 
stiff systems...

Regards,

   Thomas

------------------------------------------------------------------------------
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily build your RIAs with Flex Builder, the Eclipse(TM)based development
software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
_______________________________________________
Octave-dev mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/octave-dev

Reply via email to