Merge pull request #279 from avik-pal/ap/better_default

ChrisRackauckas · web-flow · commit c2d007759c00 · 2023-11-07T11:31:58.000+01:00
Use Jacobian Based Algorithms if user supplies custom Jacobian
diff --git a/Project.toml b/Project.toml
@@ -1,7 +1,7 @@
 name = "NonlinearSolve"
 uuid = "8913a72c-1f9b-4ce2-8d82-65094dcecaec"
 authors = ["SciML"]
-version = "2.8.0"
+version = "2.8.1"
 
 [deps]
 ADTypes = "47edcb42-4c32-4615-8424-f2b9edc5f35b"
diff --git a/src/broyden.jl b/src/broyden.jl
@@ -8,7 +8,7 @@ An implementation of `Broyden` with reseting and line search.
 
   - `max_resets`: the maximum number of resets to perform. Defaults to `3`.
   - `reset_tolerance`: the tolerance for the reset check. Defaults to
-    `sqrt(eps(eltype(u)))`.
+    `sqrt(eps(real(eltype(u))))`.
   - `linesearch`: the line search algorithm to use. Defaults to [`LineSearch()`](@ref),
     which means that no line search is performed. Algorithms from `LineSearches.jl` can be
     used here directly, and they will be converted to the correct `LineSearch`. It is
@@ -67,7 +67,7 @@ function SciMLBase.__init(prob::NonlinearProblem{uType, iip}, alg::GeneralBroyde
     u = alias_u0 ? u0 : deepcopy(u0)
     fu = evaluate_f(prob, u)
     J⁻¹ = __init_identity_jacobian(u, fu)
-    reset_tolerance = alg.reset_tolerance === nothing ? sqrt(eps(eltype(u))) :
+    reset_tolerance = alg.reset_tolerance === nothing ? sqrt(eps(real(eltype(u)))) :
                       alg.reset_tolerance
     reset_check = x -> abs(x) ≤ reset_tolerance
 
diff --git a/src/default.jl b/src/default.jl
@@ -166,7 +166,7 @@ end
 
 """
     RobustMultiNewton(; concrete_jac = nothing, linsolve = nothing, precs = DEFAULT_PRECS,
-                        adkwargs...)
+        adkwargs...)
 
 A polyalgorithm focused on robustness. It uses a mixture of Newton methods with different
 globalizing techniques (trust region updates, line searches, etc.) in order to find a
@@ -212,7 +212,7 @@ end
 
 """
     FastShortcutNonlinearPolyalg(; concrete_jac = nothing, linsolve = nothing,
-                                   precs = DEFAULT_PRECS, adkwargs...)
+        precs = DEFAULT_PRECS, must_use_jacobian::Val = Val(false), adkwargs...)
 
 A polyalgorithm focused on balancing speed and robustness. It first tries less robust methods
 for more performance and then tries more robust techniques if the faster ones fail.
@@ -238,15 +238,25 @@ for more performance and then tries more robust techniques if the faster ones fa
     [LinearSolve.jl documentation](https://docs.sciml.ai/LinearSolve/stable/).
 """
 function FastShortcutNonlinearPolyalg(; concrete_jac = nothing, linsolve = nothing,
-        precs = DEFAULT_PRECS, adkwargs...)
-    algs = (GeneralKlement(; linsolve, precs),
-        GeneralBroyden(),
-        NewtonRaphson(; concrete_jac, linsolve, precs, adkwargs...),
-        NewtonRaphson(; concrete_jac, linsolve, precs, linesearch = BackTracking(),
-            adkwargs...),
-        TrustRegion(; concrete_jac, linsolve, precs, adkwargs...),
-        TrustRegion(; concrete_jac, linsolve, precs,
-            radius_update_scheme = RadiusUpdateSchemes.Bastin, adkwargs...))
+        precs = DEFAULT_PRECS, must_use_jacobian::Val{JAC} = Val(false),
+        adkwargs...) where {JAC}
+    if JAC
+        algs = (NewtonRaphson(; concrete_jac, linsolve, precs, adkwargs...),
+            NewtonRaphson(; concrete_jac, linsolve, precs, linesearch = BackTracking(),
+                adkwargs...),
+            TrustRegion(; concrete_jac, linsolve, precs, adkwargs...),
+            TrustRegion(; concrete_jac, linsolve, precs,
+                radius_update_scheme = RadiusUpdateSchemes.Bastin, adkwargs...))
+    else
+        algs = (GeneralKlement(; linsolve, precs),
+            GeneralBroyden(),
+            NewtonRaphson(; concrete_jac, linsolve, precs, adkwargs...),
+            NewtonRaphson(; concrete_jac, linsolve, precs, linesearch = BackTracking(),
+                adkwargs...),
+            TrustRegion(; concrete_jac, linsolve, precs, adkwargs...),
+            TrustRegion(; concrete_jac, linsolve, precs,
+                radius_update_scheme = RadiusUpdateSchemes.Bastin, adkwargs...))
+    end
     return NonlinearSolvePolyAlgorithm(algs, Val(:NLS))
 end
 
@@ -288,12 +298,22 @@ end
 
 ## Defaults
 
+## TODO: In the long run we want to use an `Assumptions` API like LinearSolve to specify
+##       the conditioning of the Jacobian and such
+## Defaults to a fast and robust poly algorithm in most cases. If the user went through
+## the trouble of specifying a custom jacobian function, we should use algorithms that
+## can use that!
+
 function SciMLBase.__init(prob::NonlinearProblem, ::Nothing, args...; kwargs...)
-    return SciMLBase.__init(prob, FastShortcutNonlinearPolyalg(), args...; kwargs...)
+    must_use_jacobian = Val(prob.f.jac !== nothing)
+    return SciMLBase.__init(prob, FastShortcutNonlinearPolyalg(; must_use_jacobian),
+        args...; kwargs...)
 end
 
 function SciMLBase.__solve(prob::NonlinearProblem, ::Nothing, args...; kwargs...)
-    return SciMLBase.__solve(prob, FastShortcutNonlinearPolyalg(), args...; kwargs...)
+    must_use_jacobian = Val(prob.f.jac !== nothing)
+    return SciMLBase.__solve(prob, FastShortcutNonlinearPolyalg(; must_use_jacobian),
+        args...; kwargs...)
 end
 
 function SciMLBase.__init(prob::NonlinearLeastSquaresProblem, ::Nothing, args...; kwargs...)
diff --git a/src/klement.jl b/src/klement.jl
@@ -142,7 +142,7 @@ function perform_step!(cache::GeneralKlementCache{true})
     mul!(cache.Jdu, J, _vec(du))
     cache.fu .= cache.fu2 .- cache.fu
     cache.fu .= _restructure(cache.fu,
-        (_vec(cache.fu) .- cache.Jdu) ./ max.(cache.Jᵀ²du, eps(T)))
+        (_vec(cache.fu) .- cache.Jdu) ./ max.(cache.Jᵀ²du, eps(real(T))))
     mul!(cache.J_cache, _vec(cache.fu), _vec(du)')
     cache.J_cache .*= J
     mul!(cache.J_cache2, cache.J_cache, J)
@@ -202,7 +202,7 @@ function perform_step!(cache::GeneralKlementCache{false})
     cache.Jdu = J * _vec(cache.du)
     cache.fu = cache.fu2 .- cache.fu
     cache.fu = _restructure(cache.fu,
-        (_vec(cache.fu) .- cache.Jdu) ./ max.(cache.Jᵀ²du, eps(T)))
+        (_vec(cache.fu) .- cache.Jdu) ./ max.(cache.Jᵀ²du, eps(real(T))))
     cache.J_cache = ((_vec(cache.fu) * _vec(cache.du)') .* J) * J
     cache.J = J .+ cache.J_cache
 
diff --git a/src/lbroyden.jl b/src/lbroyden.jl
@@ -8,7 +8,7 @@ An implementation of `LimitedMemoryBroyden` with reseting and line search.
 
   - `max_resets`: the maximum number of resets to perform. Defaults to `3`.
   - `reset_tolerance`: the tolerance for the reset check. Defaults to
-    `sqrt(eps(eltype(u)))`.
+    `sqrt(eps(real(eltype(u))))`.
   - `threshold`: the number of vectors to store in the low rank approximation. Defaults
     to `10`.
   - `linesearch`: the line search algorithm to use. Defaults to [`LineSearch()`](@ref),
@@ -82,7 +82,7 @@ function SciMLBase.__init(prob::NonlinearProblem{uType, iip}, alg::LimitedMemory
     threshold = min(alg.threshold, maxiters)
     U, Vᵀ = __init_low_rank_jacobian(u, fu, threshold)
     du = copy(fu)
-    reset_tolerance = alg.reset_tolerance === nothing ? sqrt(eps(eltype(u))) :
+    reset_tolerance = alg.reset_tolerance === nothing ? sqrt(eps(real(eltype(u)))) :
                       alg.reset_tolerance
     reset_check = x -> abs(x) ≤ reset_tolerance
 
diff --git a/src/utils.jl b/src/utils.jl
@@ -302,7 +302,7 @@ _issingular(x::Number) = iszero(x)
     hasmethod(issingular, Tuple{T}) && return :(issingular(x))
     return :(__issingular(x))
 end
-__issingular(x::AbstractMatrix{T}) where {T} = cond(x) > inv(sqrt(eps(T)))
+__issingular(x::AbstractMatrix{T}) where {T} = cond(x) > inv(sqrt(eps(real(T))))
 __issingular(x) = false ## If SciMLOperator and such
 
 # If factorization is LU then perform that and update the linsolve cache
diff --git a/test/23_test_problems.jl b/test/23_test_problems.jl
@@ -3,7 +3,8 @@ using NonlinearSolve, LinearAlgebra, LinearSolve, NonlinearProblemLibrary, Test
 problems = NonlinearProblemLibrary.problems
 dicts = NonlinearProblemLibrary.dicts
 
-function test_on_library(problems, dicts, alg_ops, broken_tests, ϵ = 1e-4)
+function test_on_library(problems, dicts, alg_ops, broken_tests, ϵ = 1e-4;
+        skip_tests = nothing)
     for (idx, (problem, dict)) in enumerate(zip(problems, dicts))
         x = dict["start"]
         res = similar(x)
@@ -15,6 +16,11 @@ function test_on_library(problems, dicts, alg_ops, broken_tests, ϵ = 1e-4)
                         termination_condition = AbsNormTerminationMode())
                     problem(res, sol.u, nothing)
 
+                    skip = skip_tests !== nothing && idx in skip_tests[alg]
+                    if skip
+                        @test_skip norm(res) ≤ ϵ
+                        continue
+                    end
                     broken = idx in broken_tests[alg] ? true : false
                     @test norm(res)≤ϵ broken=broken
                 catch
@@ -90,7 +96,10 @@ end
     broken_tests = Dict(alg => Int[] for alg in alg_ops)
     broken_tests[alg_ops[1]] = [1, 2, 4, 5, 6, 11, 12, 13, 14]
 
-    test_on_library(problems, dicts, alg_ops, broken_tests)
+    skip_tests = Dict(alg => Int[] for alg in alg_ops)
+    skip_tests[alg_ops[1]] = [22]
+
+    test_on_library(problems, dicts, alg_ops, broken_tests; skip_tests)
 end
 
 @testset "GeneralKlement 23 Test Problems" begin
