Glad to hear that the installation worked out. I will try to look at your
Matlab code later today.
For now just two brief comments:
- We recently changed the interface of the solvers in ODE.jl, which now
matches the one used in Sundials.jl. However, we wanted to do some more
changes before we tag a new version (which will also break some existing
code).
- Internally, Sundials.ode et al always use adaptive time steps. The
(simplified) routines are just set up in a way to return only results at
the specified points (in contrast to ode23 in ODE.jl). But I am not sure if
it is at all possible to access the results at the intermediate integration
steps in Sundials ... If you need intermediate results, maybe you could
subdivide your integration interval, like intervals(1,:) = [0:time_res:
stim_start] with
some sufficiently small time_res?
Best,
Alex.
On Friday, 30 May 2014 01:33:37 UTC+2, Frederick wrote:
>
> This works for me, too.
> I had to also do a Pkg.rm("Sundials") to start out with, though.
> Now that the Sundials tests are working, on to figuring out how to
> 'translate' from the Matlab code.
> Another thing you might want to put into the comments for Sundials, or to
> highlight:
> the order of the arguments is NOT the same as in the ODE package.
>
>
