search

SED-ML / KiSAO

Ontology of algorithms for analyzing biological models, their parameters, and their outputs
Artistic License 2.0
9 stars 0 forks source link

Create new 'DAE Solver' category; put IDA-like things in it. #115

Closed luciansmith closed 4 months ago

luciansmith commented 4 months ago

IDA solvers are ODE solvers plus algebraic rules, not steady state solvers.

luciansmith commented 4 months ago

This PR stemmed from Logan noticing that when VCell told SED-ML it used IDA to solve some things, it would refuse to try to run those on Copasi or Roadrunner. However, the models in question didn't actually need IDA; they just happened to use IDA on a problem that was actually just ODEs, with no DAEs in it. So now, the IDA family of solvers is now removed from the steady state solvers (where they didn't belong) and is now in its own DAE branch, where everything is tagged as solving DAE problems (KiSAO 373) as well as ODE problems (374). If you ask libkisao for a list of ODE solvers, it gives you both lists, and if you ask it for DAE solvers, it just gives you those.

In addition, the KINSOL solver and the 'method for solving a system of linear equations' branch were tagged as being DAE solvers, which they are not, so I removed that tag from them. It's possible that one of the children of that branch are? Maybe? But this is well beyond my knowledge.