Merge branch 'master' into ADTypesSwitch

Project.toml

@@ -96,7 +96,7 @@ LinearSolve = "2"
 Logging = "1.9"
 MacroTools = "0.5"
 MuladdMacro = "0.2.1"
-NonlinearSolve = "3"
+NonlinearSolve = "3, 4"
 OrdinaryDiffEqAdamsBashforthMoulton = "1"
 OrdinaryDiffEqBDF = "1"
 OrdinaryDiffEqCore = "1"
@@ -121,8 +121,8 @@ OrdinaryDiffEqQPRK = "1"
 OrdinaryDiffEqRKN = "1"
 OrdinaryDiffEqRosenbrock = "1"
 OrdinaryDiffEqSDIRK = "1"
-OrdinaryDiffEqStabilizedIRK = "1"
 OrdinaryDiffEqSSPRK = "1"
+OrdinaryDiffEqStabilizedIRK = "1"
 OrdinaryDiffEqStabilizedRK = "1"
 OrdinaryDiffEqSymplecticRK = "1"
 OrdinaryDiffEqTsit5 = "1"
@@ -136,7 +136,7 @@ Reexport = "1.0"
 SciMLBase = "2.53.2"
 SciMLOperators = "0.3"
 SciMLStructures = "1"
-SimpleNonlinearSolve = "1"
+SimpleNonlinearSolve = "1, 2"
 SimpleUnPack = "1"
 SparseDiffTools = "2"
 Static = "0.8, 1"

README.md

@@ -15,6 +15,7 @@ ordinary differential equation solvers and utilities. While completely independe
 and usable on its own, users interested in using this
 functionality should check out [DifferentialEquations.jl](https://github.com/SciML/DifferentialEquations.jl).
 
+
 ## Installation
 
 Assuming that you already have Julia correctly installed, it suffices to import

lib/OrdinaryDiffEqCore/Project.toml

@@ -1,7 +1,7 @@
 name = "OrdinaryDiffEqCore"
 uuid = "bbf590c4-e513-4bbe-9b18-05decba2e5d8"
 authors = ["ParamThakkar123 <[email protected]>"]
-version = "1.9.0"
+version = "1.10.0"
 
 [deps]
 ADTypes = "47edcb42-4c32-4615-8424-f2b9edc5f35b"
@@ -34,6 +34,7 @@ SimpleUnPack = "ce78b400-467f-4804-87d8-8f486da07d0a"
 Static = "aedffcd0-7271-4cad-89d0-dc628f76c6d3"
 StaticArrayInterface = "0d7ed370-da01-4f52-bd93-41d350b8b718"
 StaticArraysCore = "1e83bf80-4336-4d27-bf5d-d5a4f845583c"
+SymbolicIndexingInterface = "2efcf032-c050-4f8e-a9bb-153293bab1f5"
 TruncatedStacktraces = "781d530d-4396-4725-bb49-402e4bee1e77"
 
 [weakdeps]
@@ -69,13 +70,14 @@ Random = "<0.0.1, 1"
 RecursiveArrayTools = "2.36, 3"
 Reexport = "1.0"
 SafeTestsets = "0.1.0"
-SciMLBase = "2.56"
+SciMLBase = "2.57.2"
 SciMLOperators = "0.3"
 SciMLStructures = "1"
 SimpleUnPack = "1"
 Static = "0.8, 1"
 StaticArrayInterface = "1.2"
 StaticArraysCore = "1.0"
+SymbolicIndexingInterface = "0.3.31"
 Test = "<0.0.1, 1"
 TruncatedStacktraces = "1.2"
 julia = "1.10"

lib/OrdinaryDiffEqCore/src/OrdinaryDiffEqCore.jl

@@ -76,6 +76,8 @@ import Accessors: @reset
 
 using SciMLStructures: canonicalize, Tunable, isscimlstructure
 
+using SymbolicIndexingInterface: parameter_values, is_variable, variable_index, symbolic_type, NotSymbolic
+
 const CompiledFloats = Union{Float32, Float64}
 import Preferences
 

lib/OrdinaryDiffEqCore/src/integrators/integrator_interface.jl

@@ -3,7 +3,7 @@
 # Hence, we need to have two separate functions.
 
 function _change_t_via_interpolation!(integrator, t,
-        modify_save_endpoint::Type{Val{T}}) where {T}
+        modify_save_endpoint::Type{Val{T}}, reinitialize_alg=nothing) where {T}
     # Can get rid of an allocation here with a function
     # get_tmp_arr(integrator.cache) which gives a pointer to some
     # cache array which can be modified.
@@ -17,7 +17,7 @@ function _change_t_via_interpolation!(integrator, t,
         end
         integrator.t = t
         integrator.dt = integrator.t - integrator.tprev
-        DiffEqBase.reeval_internals_due_to_modification!(integrator)
+        DiffEqBase.reeval_internals_due_to_modification!(integrator; callback_initializealg=reinitialize_alg)
         if T
             solution_endpoint_match_cur_integrator!(integrator)
         end
@@ -28,10 +28,10 @@ function DiffEqBase.change_t_via_interpolation!(integrator::ODEIntegrator,
         t,
         modify_save_endpoint::Type{Val{T}} = Val{
             false,
-        }) where {
+        }, reinitialize_alg=nothing) where {
         T,
 }
-    _change_t_via_interpolation!(integrator, t, modify_save_endpoint)
+    _change_t_via_interpolation!(integrator, t, modify_save_endpoint, reinitialize_alg)
     return nothing
 end
 

lib/OrdinaryDiffEqCore/src/solve.jl

@@ -264,6 +264,30 @@ function DiffEqBase.__init(
     end
 
     ### Algorithm-specific defaults ###
+    if save_idxs === nothing
+        saved_subsystem = nothing
+    else
+        if !(save_idxs isa AbstractArray) || symbolic_type(save_idxs) != NotSymbolic()
+            _save_idxs = [save_idxs]
+        else
+            _save_idxs = save_idxs
+        end
+        saved_subsystem = SciMLBase.SavedSubsystem(prob, parameter_values(prob), _save_idxs)
+        if saved_subsystem !== nothing
+            _save_idxs = SciMLBase.get_saved_state_idxs(saved_subsystem)
+            if isempty(_save_idxs)
+                # no states to save
+                save_idxs = Int[]
+            elseif !(save_idxs isa AbstractArray) || symbolic_type(save_idxs) != NotSymbolic()
+                # only a single state to save, and save it as a scalar timeseries instead of
+                # single-element array
+                save_idxs = only(_save_idxs)
+            else
+                save_idxs = _save_idxs
+            end
+        end
+    end
+
     if save_idxs === nothing
         ksEltype = Vector{rateType}
     else
@@ -427,7 +451,7 @@ function DiffEqBase.__init(
         f, timeseries, ts, ks, alg_choice, dense, cache, differential_vars, false)
     sol = DiffEqBase.build_solution(prob, _alg, ts, timeseries,
         dense = dense, k = ks, interp = id, alg_choice = alg_choice,
-        calculate_error = false, stats = stats)
+        calculate_error = false, stats = stats, saved_subsystem = saved_subsystem)
 
     if recompile_flag == true
         FType = typeof(f)

lib/OrdinaryDiffEqExtrapolation/Project.toml

@@ -1,7 +1,7 @@
 name = "OrdinaryDiffEqExtrapolation"
 uuid = "becaefa8-8ca2-5cf9-886d-c06f3d2bd2c4"
 authors = ["Chris Rackauckas <[email protected]>", "Yingbo Ma <[email protected]>"]
-version = "1.2.0"
+version = "1.2.1"
 
 [deps]
 ADTypes = "47edcb42-4c32-4615-8424-f2b9edc5f35b"

lib/OrdinaryDiffEqFIRK/test/ode_firk_tests.jl

@@ -1,7 +1,7 @@
 using OrdinaryDiffEqFIRK, DiffEqDevTools, Test, LinearAlgebra
 import ODEProblemLibrary: prob_ode_linear, prob_ode_2Dlinear, van
 
-testTol = 0.35
+testTol = 0.5
 
 for prob in [prob_ode_linear, prob_ode_2Dlinear]
     sim21 = test_convergence(1 .// 2 .^ (6:-1:3), prob, RadauIIA5())

lib/OrdinaryDiffEqNonlinearSolve/Project.toml

@@ -1,7 +1,7 @@
 name = "OrdinaryDiffEqNonlinearSolve"
 uuid = "127b3ac7-2247-4354-8eb6-78cf4e7c58e8"
 authors = ["Chris Rackauckas <[email protected]>", "Yingbo Ma <[email protected]>"]
-version = "1.2.1"
+version = "1.2.2"
 
 [deps]
 ADTypes = "47edcb42-4c32-4615-8424-f2b9edc5f35b"

lib/OrdinaryDiffEqNonlinearSolve/src/type.jl

@@ -100,7 +100,7 @@ function NLSolver{iip, tType}(z, tmp, ztmp, γ, c, α, alg, κ, fast_convergence
         tmp,
         tmp2,
         ztmp,
-        γ,
+        float(γ),
         convert(tType, c),
         convert(tType, α),
         alg,

lib/OrdinaryDiffEqRosenbrock/src/OrdinaryDiffEqRosenbrock.jl

@@ -40,7 +40,7 @@ using Reexport
 import OrdinaryDiffEqCore: alg_autodiff
 import OrdinaryDiffEqCore
 
-function rosenbrock_wanner_docstring(description::String,
+function rosenbrock_wolfbrandt_docstring(description::String,
         name::String;
         references::String = "",
         extra_keyword_description = "",
@@ -119,7 +119,7 @@ function rosenbrock_wanner_docstring(description::String,
     end
 
     generic_solver_docstring(
-        description, name, "Rosenbrock-Wanner Method. ", references,
+        description, name, "Rosenbrock-Wanner-W(olfbrandt) Method. ", references,
         keyword_default_description, keyword_default
     )
 end
@@ -203,7 +203,7 @@ function rosenbrock_docstring(description::String,
     end
 
     generic_solver_docstring(
-        description, name, "Rosenbrock Method. ", references,
+        description, name, "Rosenbrock-Wanner Method. ", references,
         keyword_default_description, keyword_default
     )
 end

lib/OrdinaryDiffEqRosenbrock/src/algorithms.jl

@@ -144,7 +144,7 @@ function GeneralRosenbrock(; chunk_size = Val{0}(), autodiff = AutoForwardDiff()
         factorization)
 end
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
     """
     A 4th order L-stable Rosenbrock-W method (fixed step only).
     """,

lib/OrdinaryDiffEqRosenbrock/src/generic_rosenbrock.jl

@@ -911,7 +911,7 @@ function ROS2Tableau() # 2nd order
     RosenbrockAdaptiveTableau(a,C,b,btilde,gamma,d,c)
 end
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 An Order 2/3 L-Stable Rosenbrock-W method which is good for very stiff equations with oscillations at low tolerances. 2nd order stiff-aware interpolation.
 """,
@@ -922,7 +922,7 @@ Scientific Computing, 18 (1), pp. 1-22.
 """,
 with_step_limiter = true) Rosenbrock23
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 An Order 3/2 A-Stable Rosenbrock-W method which is good for mildly stiff equations without oscillations at low tolerances. Note that this method is prone to instability in the presence of oscillations, so use with caution. 2nd order stiff-aware interpolation.
 """,
@@ -944,7 +944,7 @@ references = """
 """,
 with_step_limiter = true) ROS3P
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 An Order 2/3 L-Stable Rosenbrock-W method for stiff ODEs and DAEs in mass matrix form. 2nd order stiff-aware interpolation and additional error test for interpolation.
 """,
@@ -955,7 +955,7 @@ references = """
 """,
 with_step_limiter = true) Rodas23W
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 A 4th order L-stable Rosenbrock-W method.
 """,
@@ -971,7 +971,7 @@ references = """
   publisher={Springer}}
 """) ROS34PW1a
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 A 4th order L-stable Rosenbrock-W method.
 """,
@@ -987,7 +987,7 @@ references = """
   publisher={Springer}}
 """) ROS34PW1b
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 A 4th order stiffy accurate Rosenbrock-W method for PDAEs.
 """,
@@ -1003,7 +1003,7 @@ references = """
   publisher={Springer}}
 """) ROS34PW2
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 A 4th order strongly A-stable (Rinf~0.63) Rosenbrock-W method.
 """,
@@ -1088,7 +1088,7 @@ references = """
 """,
 with_step_limiter=true) Rodas42
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_docstring(
 """
 4th order A-stable stiffly stable Rosenbrock method with a stiff-aware 3rd order interpolant. 4th order
 on linear parabolic problems and 3rd order accurate on nonlinear parabolic problems (as opposed to
@@ -1102,7 +1102,7 @@ references = """
 """,
 with_step_limiter=true) Rodas4P
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 A 4th order L-stable stiffly stable Rosenbrock method with a stiff-aware 3rd order interpolant. 4th order
 on linear parabolic problems and 3rd order accurate on nonlinear parabolic problems. It is an improvement
@@ -1128,7 +1128,7 @@ references = """
 """,
 with_step_limiter=true) Rodas5
 
-@doc rosenbrock_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 A 5th order A-stable stiffly stable Rosenbrock method with a stiff-aware 4th order interpolant.
 Has improved stability in the adaptive time stepping embedding.
@@ -1141,7 +1141,7 @@ references = """
 """,
 with_step_limiter=true) Rodas5P
 
-@doc rosenbrock_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 A 5th order A-stable stiffly stable Rosenbrock method with a stiff-aware 4th order interpolant.
 Has improved stability in the adaptive time stepping embedding.
@@ -1154,7 +1154,7 @@ references = """
 """,
 with_step_limiter=true) Rodas5Pr
 
-@doc rosenbrock_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 A 5th order A-stable stiffly stable Rosenbrock method with a stiff-aware 4th order interpolant.
 Has improved stability in the adaptive time stepping embedding.
@@ -1200,7 +1200,7 @@ references = """
 """,
 with_step_limiter=true) Veldd4
 
-@doc rosenbrock_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 A 4th order A-stable Rosenbrock method.
 """,
@@ -1324,7 +1324,7 @@ function ROS2STableau() # 2nd order
     RosenbrockAdaptiveTableau(a,C,b,btilde,gamma,d,c)
 end
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 2nd order stiffly accurate Rosenbrock-Wanner W-method with 3 internal stages with B_PR consistent of order 2 with (Rinf=0).
 """,
@@ -1619,7 +1619,7 @@ function ROS34PRwTableau() # 3rd order
     RosenbrockAdaptiveTableau(a,C,b,btilde,gamma,d,c)
 end
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 3rd order stiffly accurate Rosenbrock-Wanner W-method with 4 internal stages,
 B_PR consistent of order 2.
@@ -1736,7 +1736,7 @@ function ROK4aTableau() # 4rd order
     RosenbrockAdaptiveTableau(a,C,b,btilde,gamma,d,c)
 end
 
-@doc rosenbrock_wanner_docstring(
+@doc rosenbrock_wolfbrandt_docstring(
 """
 4rd order L-stable Rosenbrock-Krylov method with 4 internal stages,
 with a 3rd order embedded method which is strongly A-stable with Rinf~=0.55. (when using exact Jacobians)

test/downstream/autodiff_events.jl

@@ -24,7 +24,7 @@ prob = ODEProblem(f, eltype(p).([1.0, 0.0]), eltype(p).((0.0, 1.0)), copy(p))
 function test_f(p)
     _prob = remake(prob, p = p)
     solve(_prob, Tsit5(), abstol = 1e-14, reltol = 1e-14, callback = cb,
-        save_everystep = false)[end]
+        save_everystep = false).u[end]
 end
 findiff = Calculus.finite_difference_jacobian(test_f, p)
 findiff

test/downstream/delaydiffeq.jl

@@ -24,8 +24,8 @@ using Test
         @test sol.errors[:l∞] < error
 
         sol_scalar = solve(prob_scalar, ddealg)
-        @test sol.t≈sol_scalar.t atol=1e-6
-        @test sol[1, :] ≈ sol_scalar.u
+        @test sol.t≈sol_scalar.t atol=1e-3
+        @test sol[1, :] ≈ sol_scalar.u atol=1e-3
     end
 end