- From: r.britten at auckland.ac.nz (Randall Britten)
- Subject: [cellml-dev] Pluggable solver for CellML
- Date: Wed, 16 Apr 2008 07:56:32 +1200
Another option is to use COR, which can generate Java code for the model.
There is a good possibility that PCEnv will have this feature too, in the
near future.
>
-----Original Message-----
>
From: cellml-tools-developers-bounces at cellml.org [mailto:cellml-tools-
>
developers-bounces at cellml.org] On Behalf Of Andrew
>
Sent: Tuesday, 15 April 2008 11:26 p.m.
>
To: Michael Cooling
>
Cc: cellml-tools-developers at cellml.org
>
Subject: Re: [cellml-dev] Pluggable solver for CellML
>
>
Michael Cooling wrote:
>
> Hi
>
>
>
> Some collaborators in Newcastle want to run integrators over CellML
>
> model programmatically. What hooks do PCEnv or COR have to do this?
>
Do
>
> they have a Java-accessible bridge/api or similar to enable
>
> integration of the solvers with other code?
>
Hi Mike,
>
>
There is no need to do this through a front end, this is the purpose of
>
the CellML API (which is what PCEnv uses to perform the integration).
>
The CellML API provides a service called the CIS (CellML Integration
>
Service) which is used for integration.
>
>
We do have plans to eventually develop a Java bridge, but this is
>
something that has been pushed to a lower priority several times so has
>
never actually been implemented.
>
>
>
>
> As competition, these people are also assessing SBML workbench which
>
> apparently allows you to do this. I think if they can't easily run
>
the
>
> CellML model they may shift to SBML soon which would be a shame.
>
>
I don't think we should regard SBML as 'competition', SBML and CellML
>
solve different problems because CellML is a more general format for
>
representing mathematical models, while SBML tackles the problem of
>
representing systems of reactions. That said, since CellML is more
>
general, we want it to be able to represent any mathematical system
>
that
>
SBML can represent.
>
>
However, it would seem that for the general problem of integrating
>
models, CellML would normally be the best choice of format, and I
>
believe that, at least in terms of integrator performance, the CellML
>
tools are the most developed.
>
>
One option in the interim until we have a full Java bridge would be the
>
use test program built from the API which allows simulations to be run
>
from the command line, and outputs CSV results to standard output.
>
>
Best regards,
>
Andrew
>
>
_______________________________________________
>
cellml-tools-developers mailing list
>
cellml-tools-developers at cellml.org
>
http://www.cellml.org/mailman/listinfo/cellml-tools-developers
Archive powered by MHonArc 2.6.18.