Boost logo

Boost :

Subject: Re: [boost] Fwd: Regarding GSoc'14 Boost.odeint idea
From: Rhys Ulerich (rhys.ulerich_at_[hidden])
Date: 2014-03-10 17:01:40


On Mon, Mar 10, 2014 at 3:40 PM, Karsten Ahnert
<karsten.ahnert_at_[hidden]> wrote:
> Implicit routines require to solve linear systems. I think, the main
> difficulty is to integrate these linear solvers into odeint such that
> they are exchangeable like the existing algebras and operations. This
> would be a good foundation to implement implicit routines and I would be
> happy if the BDF method could be implemented.

To briefly hijack a thread, SDIRK might be a good use case for API design.

- Rhys


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk