Enable symbolic remake for optimization problems #1972
Conversation
ValentinKaisermayer
changed the title
src/variables.jl
Outdated
| """ | ||
| $(SIGNATURES) | ||
|
|
||
| Intercept the call to `process_p_u0_symbolic` and process symbolic maps of `p` and/or `u0` if the | ||
| user has `ModelingToolkit` loaded. | ||
| """ | ||
| function SciMLBase.process_p_u0_symbolic(prob::ODEProblem, p, u0) | ||
| function SciMLBase.process_p_u0_symbolic(prob::Union{ODEProblem, OptimizationProblem}, p, |
There was a problem hiding this comment.
we still need to do a lot more problem types 😓
There was a problem hiding this comment.
Why does this function not also cover DEProblem in full and NonlinearProblem?
There was a problem hiding this comment.
Good question. I think it can. It only expects parameters and states to be defined for the underlying system in prob.f.
There was a problem hiding this comment.
What's a DEProblem?
There was a problem hiding this comment.
Do you mean an AbstractDEProblem?
|
That's an odd failure I haven't seen before... ? |
Yeah, added a parameter and forgot to change the test. I will fix it. |
test/nonlinearsystem.jl
Outdated
| @test prob_.p == [1.1, 1.2, 1.3] | ||
|
|
||
| prob_ = remake(prob, u0 = Dict(sys.y => 2.0), p = Dict(sys.a => 2.0)) | ||
| @test_broken prob_.u0 == [1.0, 2.0, 1.0] # TODO: needs a `remake(prob::NonlinearProblem, ...)` in SciMLBase |
There was a problem hiding this comment.
remember
There was a problem hiding this comment.
|
Thanks for seeing this through. It's absolutely fantastic to see this completed! |
|
thx |
Uh oh!
There was an error while loading. Please reload this page.