Re: matlab compatibility of od45 with adaptive stepping

[Carlo De Falco]

On 11 Oct 2015, at 13:31, Carlo de Falco <address@hidden> wrote:

> 
> On 11 Oct 2015, at 13:22, Carlo De Falco <address@hidden> wrote:
> 
>> The main difficulty in writing the patch is understanding all the output 
>> flags and maintaining 
>> compatibility so that high level interface functions such as ode45 do not 
>> require too many modifications.
> 
> for example, can someone clearly explain the meaning of the flag 
> "solution.vunhandledtermination" and its purpose?
> what would happe if this were not set correctly by integrate_adaptive?
> 
> c.

The following changesets :

http://hg.savannah.gnu.org/hgweb/octave/rev/a22d8a2eb0e5
http://hg.savannah.gnu.org/hgweb/octave/rev/43822bda4f65

address the issue.

There are now a few test failures with ode45, as the existing tests
still expect the incompatible behaviour.

I'll fix the tests in ode45, but please keep in mind 
the same will have to be done when moving other solvers.

c.