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
536 stars 205 forks source link

add strip_solution tests #2437

Closed jClugstor closed 1 month ago

jClugstor commented 1 month ago

Checklist

jClugstor commented 1 month ago

has_lazy_interpolation should extend the one in SciMLBase right? Right now the lazy interpolation check doesn't error since it's using the general dispatch in SciMLBase, instead of the one in OrdinaryDiffEq, SciMLBase.has_lazy_interpolation(alg::Union{Vern6, Vern7, Vern8, Vern9}) = true. I think I might need to reexport something?

ChrisRackauckas commented 1 month ago

You shouldn't need to export anything. If you overload it properly (like what you show) that should be fine.

ChrisRackauckas commented 1 month ago

Test failure

ChrisRackauckas commented 1 month ago

For TRBDF2, check the nlsolver is stripped https://github.com/SciML/OrdinaryDiffEq.jl/blob/master/lib/OrdinaryDiffEqSDIRK/src/sdirk_caches.jl#L15

jClugstor commented 1 month ago

Do I need to set nlsolver to nothing for any cache that has it as a field?

ChrisRackauckas commented 1 month ago

yes