Improved FunctionalAffect alternative #2922
Conversation
BenChung
force-pushed
the
careful-event-affects
branch
from
8640378 to
9d176cf
Compare
|
I'm increasingly happy with this implementation; the main caveat to that statement is that the use of ComponentArrays for MutatingFunctionalAffects means that the values passed in and out are homogenously typed (that is they're all float64s, which is not great). |
|
This looks like a rebase gone wrong 😅 |
|
Yep, alas. I'll fix it with a force push |
BenChung
force-pushed
the
careful-event-affects
branch
2 times, most recently
from
3653246 to
e96ca88
Compare
…amedTuples for heterotyped operation support.
|
Rebase. |
ChrisRackauckas
left a comment
ChrisRackauckas
left a comment
There was a problem hiding this comment.
I'm not super happy with the implementation, I think it seems a bit more complicated than maybe it could be, but I'm willing to see how this goes because it's marked as experimental
src/systems/diffeqs/odesystem.jl
Outdated
| end | ||
|
|
||
| output_expr = isscalar ? ts[1] : | ||
| (array_type <: Vector ? MakeArray(ts, output_type) : MakeTuple(ts)) |
There was a problem hiding this comment.
If I pass StaticVector{Float64, 3} here, I still get a Vector right? Seems slightly confusing, although the docstring is rather clear on what options are available.
There is an option to get static arrays out from build_function, would it make sense to follow this same interface for build_explicit_observed_function?
https://github.com/JuliaSymbolics/Symbolics.jl/blob/10a61ee505038f4faaf9e1edbcda1caeed8d4cc7/src/build_function.jl#L285
There was a problem hiding this comment.
Yes it would make sense to do the same thing here.
|
@ChrisRackauckas In terms of simplification, what'd be your thoughts on how to trim this down in that case? |
|
It's just hard to review because there's multiple things going on here. There's an extension to SymbolicContinuousCallback, there's changing options in Also, since the ImperativeAffect is marked as experimental, it should all be isolated to a file so that if it's removed it can be done cleanly. It seems like there are utils for ImperativeAffect that are mixed in with the other callbacks? |
src/systems/callbacks.jl
Outdated
|
|
||
| finalize_affects(cb::SymbolicContinuousCallback) = cb.initialize | ||
| function finalize_affects(cbs::Vector{SymbolicContinuousCallback}) | ||
| mapreduce(finalize_affects, vcat, cbs, init = Equation[]) |
There was a problem hiding this comment.
| (cb, fn) -> begin | |
| function (cb, fn) |
anonymous functions can be written this way as well, which maybe looks slightly nicer
AayushSabharwal
left a comment
AayushSabharwal
left a comment
There was a problem hiding this comment.
Just a minor doc thing I found
Co-authored-by: Aayush Sabharwal <[email protected]>
4 participants
Checklist
contributor guidelines, in particular the SciML Style Guide and
COLPRAC.
Additional context
This introduces
MutatingFunctionalAffect(better name suggestions much appreciated) which presents a higher level interface when compared to FunctionalAffect. In particular, it decouples the inputs and outputs to the affect from the precise problem structure post-simplification.This isn't quite complete yet; I need to implement it marking unknowns mutating by affects irreducible so that they're always there come runtime. Close, however.