SciML / OrdinaryDiffEq.jl

High performance ordinary differential equation (ODE) and differential-algebraic equation (DAE) solvers, including neural ordinary differential equations (neural ODEs) and scientific machine learning (SciML)
https://diffeq.sciml.ai/latest/
Other
521 stars 198 forks source link

added step_limiter to more methods #2190

Closed cwittens closed 1 month ago

cwittens commented 1 month ago

Checklist

Additional context

added step_limiter and appropriate tests to the following methods: QNDF1, QNDF2, QNDF, FBDF, ImplicitEuler, ImplicitMidpoint, Trapezoid, TRBDF2, Feagin10, Feagin12, Feagin14, KenCarp3, KenCarp4, KenCarp5, Kvaerno3, Kvaerno4, Kvaerno5,Rosenbrock23, Rosenbrock32, ROS3P, Rodas3, Rodas23W, Rodas3P, Rodas4, Rodas42, Rodas4P, Rodas4P2, Rodas5, Rodas5P, Rodas5Pe, Rodas5Pr

and changed location of the cache for Kvaerno3 from sdirk_caches.jl to kencarp_kvaerno_caches.jl.

(one commit talkes about "non adaptive rosenbrock". What I mean is methods where the code for the cache and perform_step is not generated with a marco)

ChrisRackauckas commented 1 month ago

The test failures are real.