With the default solver, deleteat! deletes too many elements

[evan-wehi]

Describe the bug 🐞

A clear and concise description of what the bug is.

delete!(integrator, indxs) should delete only entries at idxs. Instead, delete! is deleting several other entries as well.

A clear and concise description of what you expected to happen.

Only entries at idxs are deleted.

Minimal Reproducible Example 👇

using DifferentialEquations

u0 = collect(range(1.0, 5.0))
test(u, t, integrator) = u[1] - 0.5
function action!(integrator)
  @show integrator.u
  deleteat!(integrator, 2)
  @show integrator.u
end
cb = ContinuousCallback(test, action!)
rhs(du, u, p, t) = du .= -0.1 .* u
prob = ODEProblem(rhs, u0, (0, 100))
solve(prob, callback=cb)

Error & Stacktrace ⚠️

integrator.u = [0.5000000000000004, 1.0000000000000009, 1.5000000000000007, 2.0000000000000018, 2.500000000000002]
integrator.u = [0.5000000000000004]
ERROR: DimensionMismatch: Some axes are not equal, or feature a dynamic broadcast!
Stacktrace:
  [1] _fast_materialize!
    @ ~/.julia/packages/FastBroadcast/wfdTr/src/FastBroadcast.jl:193 [inlined]
  [2] fast_materialize!
    @ ~/.julia/packages/FastBroadcast/wfdTr/src/FastBroadcast.jl:277 [inlined]
  [3] _ode_addsteps!(k::Vector{…}, t::Float64, uprev::Vector{…}, u::Vector{…}, dt::Float64, f::ODEFunction{…}, p::SciMLBase.NullParameters, cache::OrdinaryDiffEqTsit5.Tsit5Cache{…}, always_calc_begin::Bool, allow_calc_end::Bool, force_calc_end::Bool)
    @ OrdinaryDiffEqTsit5 ~/.julia/packages/OrdinaryDiffEqTsit5/DHYtz/src/tsit_perform_step.jl:9
  [4] ode_addsteps!
    @ ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/dense/generic_dense.jl:87 [inlined]
  [5] ode_addsteps!
    @ ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/dense/generic_dense.jl:61 [inlined]
  [6] reeval_internals_due_to_modification!(integrator::OrdinaryDiffEqCore.ODEIntegrator{…}, continuous_modification::Bool)
    @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/integrators/integrator_interface.jl:51
  [7] reeval_internals_due_to_modification!
    @ ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/integrators/integrator_interface.jl:40 [inlined]
  [8] apply_callback!(integrator::OrdinaryDiffEqCore.ODEIntegrator{…}, callback::ContinuousCallback{…}, cb_time::Float64, prev_sign::Float64, event_idx::Int64)
    @ DiffEqBase ~/.julia/packages/DiffEqBase/sCsah/src/callbacks.jl:591
  [9] macro expansion
    @ ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/integrators/integrator_utils.jl:319 [inlined]
 [10] apply_ith_callback!
    @ ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/integrators/integrator_utils.jl:305 [inlined]
 [11] handle_callbacks!
    @ ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/integrators/integrator_utils.jl:344 [inlined]
 [12] _loopfooter!(integrator::OrdinaryDiffEqCore.ODEIntegrator{…})
    @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/integrators/integrator_utils.jl:243
 [13] loopfooter!
    @ ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/integrators/integrator_utils.jl:207 [inlined]
 [14] solve!(integrator::OrdinaryDiffEqCore.ODEIntegrator{…})
    @ OrdinaryDiffEqCore ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/solve.jl:552
 [15] #__solve#62
    @ ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/solve.jl:7 [inlined]
 [16] __solve
    @ ~/.julia/packages/OrdinaryDiffEqCore/3Y2Xj/src/solve.jl:1 [inlined]
 [17] solve_call(_prob::ODEProblem{…}, args::CompositeAlgorithm{…}; merge_callbacks::Bool, kwargshandle::Nothing, kwargs::@Kwargs{…})
    @ DiffEqBase ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:612
 [18] solve_call
    @ ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:569 [inlined]
 [19] solve_up(prob::ODEProblem{…}, sensealg::Nothing, u0::Vector{…}, p::SciMLBase.NullParameters, args::CompositeAlgorithm{…}; kwargs::@Kwargs{…})
    @ DiffEqBase ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:1080
 [20] solve_up
    @ ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:1066 [inlined]
 [21] #solve#51
    @ ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:1003 [inlined]
 [22] solve
    @ ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:993 [inlined]
 [23] #__solve#3
    @ ~/.julia/packages/OrdinaryDiffEqDefault/2qVWT/src/default_alg.jl:46 [inlined]
 [24] __solve
    @ ~/.julia/packages/OrdinaryDiffEqDefault/2qVWT/src/default_alg.jl:45 [inlined]
 [25] #__solve#72
    @ ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:1394 [inlined]
 [26] __solve
    @ ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:1386 [inlined]
 [27] solve_call(::ODEProblem{…}; merge_callbacks::Bool, kwargshandle::Nothing, kwargs::@Kwargs{…})
    @ DiffEqBase ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:612
 [28] solve_up(::ODEProblem{…}, ::Nothing, ::Vector{…}, ::SciMLBase.NullParameters; kwargs::@Kwargs{…})
    @ DiffEqBase ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:1072
 [29] solve_up
    @ ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:1066 [inlined]
 [30] solve(::ODEProblem{…}; sensealg::Nothing, u0::Nothing, p::Nothing, wrap::Val{…}, kwargs::@Kwargs{…})
    @ DiffEqBase ~/.julia/packages/DiffEqBase/sCsah/src/solve.jl:1003
 [31] top-level scope
    @ ~/PycharmProjects/Odetest/b.jl:13
Some type information was truncated. Use `show(err)` to see complete types.

Environment (please complete the following information):

• Output of using Pkg; Pkg.status()

julia> using Pkg; Pkg.status()
Project Odetest v0.1.0
Status `~/PycharmProjects/Odetest/Project.toml`
  [0c46a032] DifferentialEquations v7.14.0
  [961ee093] ModelingToolkit v9.36.0
  [91a5bcdd] Plots v1.40.8
  [d1185830] SymbolicUtils v3.5.0
  [0c5d862f] Symbolics v6.5.0

• Output of versioninfo()

julia> versioninfo()
Julia Version 1.10.1
Commit 7790d6f0641 (2024-02-13 20:41 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: macOS (arm64-apple-darwin22.4.0)
  CPU: 10 × Apple M1 Pro
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-15.0.7 (ORCJIT, apple-m1)
Threads: 8 default, 0 interactive, 4 GC (on 8 virtual cores)
Environment:
  JULIA_EDITOR = code

Additional context

If I explicitly specifiy Tsit5() or Rosenbrock23() it works as expected.

Add any other context about the problem here.

[ChrisRackauckas]

Oh the default solver probably recurses this on the caches.