Relax type constraints to allow callable parameters in pdeps

[SebastianM-C]

Checklist

• Appropriate tests were added
• Any code changes were done in a way that does not break public API
• All documentation related to code changes were updated
• The new code follows the
  contributor guidelines, in particular the SciML Style Guide and
  COLPRAC.
• Any new documentation only uses public API

Additional context

Currently parameter dependencies assume that we only have BasicSymbolic. I tried to relax that constraint a bit to allow callable parameters to be stored too. If one has both callable parameters and normal ones in the parameter dependencies, I think that there might be a performance issue, but otherwise it should not affect nortmal usage.

@AayushSabharwal is this the right way of handling this issue? I think that for normal parametrs we store things in the non-numeric portion and that's splitted by types (?), so that might be better for the heterogenuous case.

[SebastianM-C]

I'm having some trouble with the test. The codegen doesn't seem to do the replacement of the parameter in prob.f. In the function it does ˍ₋arg2[1], but in the pdep call it's still CallableFoo(p).

julia> prob.f.f.f_iip
RuntimeGeneratedFunction(#=in ModelingToolkit=#, #=using ModelingToolkit=#, :((ˍ₋out, ˍ₋arg1, ˍ₋arg2, t)->begin    
          #= C:\Users\sebastian\.julia\packages\SymbolicUtils\ij6YM\src\code.jl:385 =#
          #= C:\Users\sebastian\.julia\packages\SymbolicUtils\ij6YM\src\code.jl:386 =#
          #= C:\Users\sebastian\.julia\packages\SymbolicUtils\ij6YM\src\code.jl:387 =#
          begin
              begin
                  begin
                      i = CallableFoo(p)
                      begin
                          #= C:\Users\sebastian\.julia\packages\Symbolics\b4I7P\src\build_function.jl:546 =#       
                          #= C:\Users\sebastian\.julia\packages\SymbolicUtils\ij6YM\src\code.jl:434 =# @inbounds begin
                                  #= C:\Users\sebastian\.julia\packages\SymbolicUtils\ij6YM\src\code.jl:430 =#     
                                  ˍ₋out[1] = (+)(ˍ₋arg2[1], i(t))
                                  #= C:\Users\sebastian\.julia\packages\SymbolicUtils\ij6YM\src\code.jl:432 =#     
                                  nothing
                              end
                      end
                  end
              end
          end
      end))

[AayushSabharwal]

but in the pdep call it's still CallableFoo(p).

You probably need to @register_symbolic CallableFoo(p)

[SebastianM-C]

Thanks! That fixed it.

[SebastianM-C]

@ChrisRackauckas test failures seem the same as master. Does this look okay?

[ChrisRackauckas]

I don't like that they are just relaxed, but if @AayushSabharwal thinks it's required then let's go with it.

[SebastianM-C]

What would be the alternative? I understand that we don't want to introduce type parameters (to avoid increasing inference time?).

[AayushSabharwal]

I don't like that they are just relaxed

What's not to like? 😅 Looks good to me, and a fairly simple change allows new features.

[ChrisRackauckas]

    lhss = BasicSymbolic[]
    lhss = []

lost a validation step, but if it's required, sure.