Cannot simulate a certain ODESystem (probably one where everything gets simplified away)

[TorkelE]

This was working on the Catalyst tests up until the v9.66 release, so probably related.

Simple system, just consisting of a simple algebraic equation. Probably related to why we have an error.

using ModelingToolkit
using ModelingToolkit: t_nounits as t, D_nounits as D
using OrdinaryDiffEqDefault
@variables X(t)
eqs = [2X ~ 6.0 - X]
@mtkbuild osys = ODESystem(eqs, t)
u0 = [X => 1.0]
oprob = ODEProblem(osys, u0, 1.0)
sol = solve(oprob)

ERROR: MethodError: no method matching build_solution(::ODEProblem{…}, ::Nothing, ::Vector{…}, ::Vector{…}, ::SciMLBase.ReturnCode.T)
The function `build_solution` exists, but no method is defined for this combination of argument types.

Closest candidates are:
  build_solution(::SciMLBase.AbstractDAEProblem, ::Any, ::Any, ::Any, ::Any; timeseries_errors, dense, dense_errors, calculate_error, k, interp, retcode, destats, stats, saved_subsystem, kwargs...)
   @ SciMLBase ~/.julia/packages/SciMLBase/Ub9ub/src/solutions/dae_solutions.jl:82
  build_solution(::Union{SciMLBase.AbstractDDEProblem, SciMLBase.AbstractODEProblem}, ::Any, ::Any, ::Any; timeseries_errors, dense, dense_errors, calculate_error, k, alg_choice, interp, retcode, destats, stats, resid, original, saved_subsystem, kwargs...)
   @ SciMLBase ~/.julia/packages/SciMLBase/Ub9ub/src/solutions/ode_solutions.jl:454
  build_solution(::SciMLBase.AbstractOptimizationCache, ::Any, ::Any, ::Any; retcode, original, stats, kwargs...)
   @ SciMLBase ~/.julia/packages/SciMLBase/Ub9ub/src/solutions/optimization_solutions.jl:74
  ...

Stacktrace:
 [1] build_null_solution(::ODEProblem{…}; saveat::Tuple{}, save_everystep::Bool, save_on::Bool, save_start::Bool, save_end::Bool, kwargs::@Kwargs{})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/gGshU/src/solve.jl:719
 [2] build_null_solution(::ODEProblem{…})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/gGshU/src/solve.jl:691
 [3] solve_call(::ODEProblem{…}; merge_callbacks::Bool, kwargshandle::Nothing, kwargs::@Kwargs{})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/gGshU/src/solve.jl:627
 [4] solve_call(::ODEProblem{…})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/gGshU/src/solve.jl:592
 [5] solve_up(::ODEProblem{…}, ::Nothing, ::Nothing, ::MTKParameters{…}; kwargs::@Kwargs{})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/gGshU/src/solve.jl:1128
 [6] solve_up
   @ ~/.julia/packages/DiffEqBase/gGshU/src/solve.jl:1122 [inlined]
 [7] solve(::ODEProblem{…}; sensealg::Nothing, u0::Nothing, p::Nothing, wrap::Val{…}, kwargs::@Kwargs{})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/gGshU/src/solve.jl:1059
 [8] solve(::ODEProblem{Nothing, Tuple{…}, true, MTKParameters{…}, ODEFunction{…}, @Kwargs{}, SciMLBase.StandardODEProblem})
   @ DiffEqBase ~/.julia/packages/DiffEqBase/gGshU/src/solve.jl:1049

If things like this shouldn't be simualteable in the first way, there should probably be an error when one tries to create an ODESystem that cannot be simualted.

[ChrisRackauckas]

@AayushSabharwal I thought you already released the hotfix on this?

[AayushSabharwal]

Was waiting for CI but I guess we can just tag

[AayushSabharwal]

Waiting for registrator now

[TorkelE]

Sounds like things are already under control!