Sam Isaacson
Sam Isaacson
Need to check the parameter mapping is fully valid.
As least one user would find it helpful: https://discourse.julialang.org/t/implement-matlab-simbiology-concepts-e-g-compartments-repeated-assignments-in-julia-catalyst/114266 Post v14 we should perhaps consider a CatalystStandardLibrary for adding such components for users.
At some point we should add this as `Reaction` metadata (but we'd also need to figure out how it would interact with setting it at the system level -- probably...
Lots of deprecation warnings in the V14 branch tests. These should be fixed before V14.
https://github.com/SciML/Catalyst.jl/actions/runs/9063058490/job/24898215803?pr=823#step:6:1804 This seems due to using `replace` from Symbolics, which is not exported and probably shouldn't be used. I have marked the relevant tests as broken for now. CC: @TorkelE
https://github.com/SciML/Catalyst.jl/blob/2276e1cbd53d8ade2b7b2b6d822ab7df50fa3a53/src/reactionsystem.jl#L1632 seems problematic. We shouldn't be assuming a type. This whole code path should probably just delete the derivative terms instead of setting them to zero to avoid making assumptions...
https://github.com/SciML/Catalyst.jl/blob/2276e1cbd53d8ade2b7b2b6d822ab7df50fa3a53/src/reactionsystem.jl#L1626 we shouldn't be using what seems likely to be a truly internal Symbolics function that could change at any time as in that code.
https://github.com/SciML/ModelingToolkit.jl/pull/2798 We need to propagate this throughout Catalyst where needed (probably some tutorials, tests, and the `symmap_to_varmap` code). Also need to update to depend on MTK 9.18 or later only.
Given they are not likely to support `PeriodicCallback`s any time soon due to https://github.com/SciML/JumpProcesses.jl/issues/417,
Right now we have a lot of places using `@set!` and such to generate new, modified `ReactionSystem`s. It might be easier to just create an internal `remake` type function, which...