Merge pull request #712 from JuliaDiff/dw/backport

Backport bug and test fixes to release-0.10 to prepare for new 0.10 release

Author: devmotion

.github/workflows/ci.yml

@@ -21,7 +21,7 @@ jobs:
           - '1.0'
           - '1.6'
           - '1'
-          - 'nightly'
+          # - 'nightly'
         os:
           - ubuntu-latest
         arch:

.gitignore

@@ -4,5 +4,6 @@
 *.DS_Store
 /docs/build/
 /docs/site/
+/docs/Manifest.toml
 /benchmark_data/
 /Manifest.toml

Project.toml

@@ -21,6 +21,7 @@ CommonSubexpressions = "0.3"
 DiffResults = "0.0.1, 0.0.2, 0.0.3, 0.0.4, 1.0.1"
 DiffRules = "1.4.0"
 DiffTests = "0.0.1, 0.1"
+IrrationalConstants = "0.1, 0.2"
 LogExpFunctions = "0.3"
 NaNMath = "0.2.2, 0.3, 1"
 Preferences = "1"
@@ -35,12 +36,13 @@ ForwardDiffStaticArraysExt = "StaticArrays"
 Calculus = "49dc2e85-a5d0-5ad3-a950-438e2897f1b9"
 DiffTests = "de460e47-3fe3-5279-bb4a-814414816d5d"
 InteractiveUtils = "b77e0a4c-d291-57a0-90e8-8db25a27a240"
+IrrationalConstants = "92d709cd-6900-40b7-9082-c6be49f344b6"
 SparseArrays = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
 StaticArrays = "90137ffa-7385-5640-81b9-e52037218182"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
 [targets]
-test = ["Calculus", "DiffTests", "SparseArrays", "Test", "InteractiveUtils", "StaticArrays"]
+test = ["Calculus", "DiffTests", "IrrationalConstants", "SparseArrays", "Test", "InteractiveUtils", "StaticArrays"]
 
 [weakdeps]
-StaticArrays = "90137ffa-7385-5640-81b9-e52037218182"
+StaticArrays = "90137ffa-7385-5640-81b9-e52037218182"

docs/Project.toml

@@ -1,3 +1,6 @@
 [deps]
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
 ForwardDiff = "f6369f11-7733-5829-9624-2563aa707210"
+
+[compat]
+Documenter = "1"

docs/make.jl

@@ -11,7 +11,8 @@ makedocs(modules=[ForwardDiff],
                     "Upgrading from Older Versions" => "user/upgrade.md"],
                   "Developer Documentation" => [
                     "How ForwardDiff Works" => "dev/how_it_works.md",
-                    "How to Contribute" => "dev/contributing.md"]])
+                    "How to Contribute" => "dev/contributing.md"]],
+         checkdocs=:exports)
 
 deploydocs(
     repo = "github.com/JuliaDiff/ForwardDiff.jl.git"

ext/ForwardDiffStaticArraysExt.jl

@@ -7,7 +7,7 @@ using ForwardDiff: Dual, partials, GradientConfig, JacobianConfig, HessianConfig
                    gradient, hessian, jacobian, gradient!, hessian!, jacobian!,
                    extract_gradient!, extract_jacobian!, extract_value!,
                    vector_mode_gradient, vector_mode_gradient!,
-                   vector_mode_jacobian, vector_mode_jacobian!, valtype, value, _lyap_div!
+                   vector_mode_jacobian, vector_mode_jacobian!, valtype, value
 using DiffResults: DiffResult, ImmutableDiffResult, MutableDiffResult
 
 @generated function dualize(::Type{T}, x::StaticArray) where T
@@ -23,27 +23,25 @@ end
 
 @inline static_dual_eval(::Type{T}, f, x::StaticArray) where T = f(dualize(T, x))
 
+# To fix method ambiguity issues:
 function LinearAlgebra.eigvals(A::Symmetric{<:Dual{Tg,T,N}, <:StaticArrays.StaticMatrix}) where {Tg,T<:Real,N}
-    λ,Q = eigen(Symmetric(value.(parent(A))))
-    parts = ntuple(j -> diag(Q' * getindex.(partials.(A), j) * Q), N)
-    Dual{Tg}.(λ, tuple.(parts...))
+    return ForwardDiff._eigvals(A)
 end
-
 function LinearAlgebra.eigen(A::Symmetric{<:Dual{Tg,T,N}, <:StaticArrays.StaticMatrix}) where {Tg,T<:Real,N}
-    λ = eigvals(A)
-    _,Q = eigen(Symmetric(value.(parent(A))))
-    parts = ntuple(j -> Q*ForwardDiff._lyap_div!(Q' * getindex.(partials.(A), j) * Q - Diagonal(getindex.(partials.(λ), j)), value.(λ)), N)
-    Eigen(λ,Dual{Tg}.(Q, tuple.(parts...)))
+    return ForwardDiff._eigen(A)
 end
 
+# For `MMatrix` we can use the in-place method
+ForwardDiff._lyap_div!!(A::StaticArrays.MMatrix, λ::AbstractVector) = ForwardDiff._lyap_div!(A, λ)
+
 # Gradient
-@inline ForwardDiff.gradient(f, x::StaticArray)                      = vector_mode_gradient(f, x)
-@inline ForwardDiff.gradient(f, x::StaticArray, cfg::GradientConfig) = gradient(f, x)
-@inline ForwardDiff.gradient(f, x::StaticArray, cfg::GradientConfig, ::Val) = gradient(f, x)
+@inline ForwardDiff.gradient(f::F, x::StaticArray) where F                      = vector_mode_gradient(f, x)
+@inline ForwardDiff.gradient(f::F, x::StaticArray, cfg::GradientConfig) where F = gradient(f, x)
+@inline ForwardDiff.gradient(f::F, x::StaticArray, cfg::GradientConfig, ::Val) where F = gradient(f, x)
 
-@inline ForwardDiff.gradient!(result::Union{AbstractArray,DiffResult}, f, x::StaticArray) = vector_mode_gradient!(result, f, x)
-@inline ForwardDiff.gradient!(result::Union{AbstractArray,DiffResult}, f, x::StaticArray, cfg::GradientConfig) = gradient!(result, f, x)
-@inline ForwardDiff.gradient!(result::Union{AbstractArray,DiffResult}, f, x::StaticArray, cfg::GradientConfig, ::Val) = gradient!(result, f, x)
+@inline ForwardDiff.gradient!(result::Union{AbstractArray,DiffResult}, f::F, x::StaticArray) where F = vector_mode_gradient!(result, f, x)
+@inline ForwardDiff.gradient!(result::Union{AbstractArray,DiffResult}, f::F, x::StaticArray, cfg::GradientConfig) where F = gradient!(result, f, x)
+@inline ForwardDiff.gradient!(result::Union{AbstractArray,DiffResult}, f::F, x::StaticArray, cfg::GradientConfig, ::Val) where F = gradient!(result, f, x)
 
 @generated function extract_gradient(::Type{T}, y::Real, x::S) where {T,S<:StaticArray}
     result = Expr(:tuple, [:(partials(T, y, $i)) for i in 1:length(x)]...)
@@ -65,13 +63,13 @@ end
 end
 
 # Jacobian
-@inline ForwardDiff.jacobian(f, x::StaticArray) = vector_mode_jacobian(f, x)
-@inline ForwardDiff.jacobian(f, x::StaticArray, cfg::JacobianConfig) = jacobian(f, x)
-@inline ForwardDiff.jacobian(f, x::StaticArray, cfg::JacobianConfig, ::Val) = jacobian(f, x)
+@inline ForwardDiff.jacobian(f::F, x::StaticArray) where F = vector_mode_jacobian(f, x)
+@inline ForwardDiff.jacobian(f::F, x::StaticArray, cfg::JacobianConfig) where F  = jacobian(f, x)
+@inline ForwardDiff.jacobian(f::F, x::StaticArray, cfg::JacobianConfig, ::Val)  where F = jacobian(f, x)
 
-@inline ForwardDiff.jacobian!(result::Union{AbstractArray,DiffResult}, f, x::StaticArray) = vector_mode_jacobian!(result, f, x)
-@inline ForwardDiff.jacobian!(result::Union{AbstractArray,DiffResult}, f, x::StaticArray, cfg::JacobianConfig) = jacobian!(result, f, x)
-@inline ForwardDiff.jacobian!(result::Union{AbstractArray,DiffResult}, f, x::StaticArray, cfg::JacobianConfig, ::Val) = jacobian!(result, f, x)
+@inline ForwardDiff.jacobian!(result::Union{AbstractArray,DiffResult}, f::F, x::StaticArray)  where F = vector_mode_jacobian!(result, f, x)
+@inline ForwardDiff.jacobian!(result::Union{AbstractArray,DiffResult}, f::F, x::StaticArray, cfg::JacobianConfig) where F  = jacobian!(result, f, x)
+@inline ForwardDiff.jacobian!(result::Union{AbstractArray,DiffResult}, f::F, x::StaticArray, cfg::JacobianConfig, ::Val) where F  = jacobian!(result, f, x)
 
 @generated function extract_jacobian(::Type{T}, ydual::StaticArray, x::S) where {T,S<:StaticArray}
     M, N = length(ydual), length(x)
@@ -110,18 +108,18 @@ end
 end
 
 # Hessian
-ForwardDiff.hessian(f, x::StaticArray) = jacobian(y -> gradient(f, y), x)
-ForwardDiff.hessian(f, x::StaticArray, cfg::HessianConfig) = hessian(f, x)
-ForwardDiff.hessian(f, x::StaticArray, cfg::HessianConfig, ::Val) = hessian(f, x)
+ForwardDiff.hessian(f::F, x::StaticArray) where F = jacobian(y -> gradient(f, y), x)
+ForwardDiff.hessian(f::F, x::StaticArray, cfg::HessianConfig) where F = hessian(f, x)
+ForwardDiff.hessian(f::F, x::StaticArray, cfg::HessianConfig, ::Val) where F = hessian(f, x)
 
-ForwardDiff.hessian!(result::AbstractArray, f, x::StaticArray) = jacobian!(result, y -> gradient(f, y), x)
+ForwardDiff.hessian!(result::AbstractArray, f::F, x::StaticArray) where F = jacobian!(result, y -> gradient(f, y), x)
 
-ForwardDiff.hessian!(result::MutableDiffResult, f, x::StaticArray) = hessian!(result, f, x, HessianConfig(f, result, x))
+ForwardDiff.hessian!(result::MutableDiffResult, f::F, x::StaticArray) where F = hessian!(result, f, x, HessianConfig(f, result, x))
 
-ForwardDiff.hessian!(result::ImmutableDiffResult, f, x::StaticArray, cfg::HessianConfig) = hessian!(result, f, x)
-ForwardDiff.hessian!(result::ImmutableDiffResult, f, x::StaticArray, cfg::HessianConfig, ::Val) = hessian!(result, f, x)
+ForwardDiff.hessian!(result::ImmutableDiffResult, f::F, x::StaticArray, cfg::HessianConfig) where F = hessian!(result, f, x)
+ForwardDiff.hessian!(result::ImmutableDiffResult, f::F, x::StaticArray, cfg::HessianConfig, ::Val) where F = hessian!(result, f, x)
 
-function ForwardDiff.hessian!(result::ImmutableDiffResult, f, x::StaticArray)
+function ForwardDiff.hessian!(result::ImmutableDiffResult, f::F, x::StaticArray) where F
     T = typeof(Tag(f, eltype(x)))
     d1 = dualize(T, x)
     d2 = dualize(T, d1)

src/ForwardDiff.jl

@@ -7,7 +7,12 @@ if VERSION >= v"1.6"
 end
 using Random
 using LinearAlgebra
-
+if VERSION < v"1.2.0-DEV.125" # 1da48c2e4028c1514ed45688be727efbef1db884
+    require_one_based_indexing(A...) = !Base.has_offset_axes(A...) || throw(ArgumentError(
+        "offset arrays are not supported but got an array with index other than 1"))
+else
+    using Base: require_one_based_indexing
+end
 import Printf
 import NaNMath
 import SpecialFunctions

src/derivative.jl

@@ -22,8 +22,9 @@ stored in `y`.
 
 Set `check` to `Val{false}()` to disable tag checking. This can lead to perturbation confusion, so should be used with care.
 """
-@inline function derivative(f!, y::AbstractArray, x::Real,
-                            cfg::DerivativeConfig{T} = DerivativeConfig(f!, y, x), ::Val{CHK}=Val{true}()) where {T, CHK}
+@inline function derivative(f!::F, y::AbstractArray, x::Real,
+                            cfg::DerivativeConfig{T} = DerivativeConfig(f!, y, x), ::Val{CHK}=Val{true}()) where {F, T, CHK}
+    require_one_based_indexing(y)
     CHK && checktag(T, f!, x)
     ydual = cfg.duals
     seed!(ydual, y)
@@ -42,6 +43,7 @@ This method assumes that `isa(f(x), Union{Real,AbstractArray})`.
 """
 @inline function derivative!(result::Union{AbstractArray,DiffResult},
                              f::F, x::R) where {F,R<:Real}
+    result isa DiffResult || require_one_based_indexing(result)
     T = typeof(Tag(f, R))
     ydual = f(Dual{T}(x, one(x)))
     result = extract_value!(T, result, ydual)
@@ -58,8 +60,9 @@ called as `f!(y, x)` where the result is stored in `y`.
 Set `check` to `Val{false}()` to disable tag checking. This can lead to perturbation confusion, so should be used with care.
 """
 @inline function derivative!(result::Union{AbstractArray,DiffResult},
-                             f!, y::AbstractArray, x::Real,
-                             cfg::DerivativeConfig{T} = DerivativeConfig(f!, y, x), ::Val{CHK}=Val{true}()) where {T, CHK}
+                             f!::F, y::AbstractArray, x::Real,
+                             cfg::DerivativeConfig{T} = DerivativeConfig(f!, y, x), ::Val{CHK}=Val{true}()) where {F, T, CHK}
+    result isa DiffResult ? require_one_based_indexing(y) : require_one_based_indexing(result, y)
     CHK && checktag(T, f!, x)
     ydual = cfg.duals
     seed!(ydual, y)

src/dual.jl

@@ -78,6 +78,10 @@ end
 @inline Dual{T,V,N}(x::Number) where {T,V,N} = convert(Dual{T,V,N}, x)
 @inline Dual{T,V}(x) where {T,V} = convert(Dual{T,V}, x)
 
+# Fix method ambiguity issue by adapting the definition in Base to `Dual`s
+Dual{T,V,N}(x::Base.TwicePrecision) where {T,V,N} =
+    (Dual{T,V,N}(x.hi) + Dual{T,V,N}(x.lo))::Dual{T,V,N}
+
 ##############################
 # Utility/Accessor Functions #
 ##############################
@@ -340,7 +344,6 @@ else
     Base.div(x::Dual, y::Dual) = div(value(x), value(y))
 end
 
-Base.hash(d::Dual) = hash(value(d))
 Base.hash(d::Dual, hsh::UInt) = hash(value(d), hsh)
 
 function Base.read(io::IO, ::Type{Dual{T,V,N}}) where {T,V,N}
@@ -416,7 +419,7 @@ function Base.promote_rule(::Type{Dual{T,A,N}},
     return Dual{T,promote_type(A, B),N}
 end
 
-for R in (Irrational, Real, BigFloat, Bool)
+for R in (AbstractIrrational, Real, BigFloat, Bool)
     if isconcretetype(R) # issue #322
         @eval begin
             Base.promote_rule(::Type{$R}, ::Type{Dual{T,V,N}}) where {T,V,N} = Dual{T,promote_type($R, V),N}
@@ -703,7 +706,11 @@ end
 # Symmetric eigvals #
 #-------------------#
 
-function LinearAlgebra.eigvals(A::Symmetric{<:Dual{Tg,T,N}}) where {Tg,T<:Real,N}
+# To be able to reuse this default definition in the StaticArrays extension
+# (has to be re-defined to avoid method ambiguity issues)
+# we forward the call to an internal method that can be shared and reused
+LinearAlgebra.eigvals(A::Symmetric{<:Dual{Tg,T,N}}) where {Tg,T<:Real,N} = _eigvals(A)
+function _eigvals(A::Symmetric{<:Dual{Tg,T,N}}) where {Tg,T<:Real,N}
     λ,Q = eigen(Symmetric(value.(parent(A))))
     parts = ntuple(j -> diag(Q' * getindex.(partials.(A), j) * Q), N)
     Dual{Tg}.(λ, tuple.(parts...))
@@ -721,8 +728,19 @@ function LinearAlgebra.eigvals(A::SymTridiagonal{<:Dual{Tg,T,N}}) where {Tg,T<:R
     Dual{Tg}.(λ, tuple.(parts...))
 end
 
-# A ./ (λ - λ') but with diag special cased
-function _lyap_div!(A, λ)
+# A ./ (λ' .- λ) but with diag special cased
+# Default out-of-place method
+function _lyap_div!!(A::AbstractMatrix, λ::AbstractVector)
+    return map(
+        (a, b, idx) -> a / (idx[1] == idx[2] ? oneunit(b) : b),
+        A,
+        λ' .- λ,
+        CartesianIndices(A),
+    )
+end
+# For `Matrix` (and e.g. `StaticArrays.MMatrix`) we can use an in-place method
+_lyap_div!!(A::Matrix, λ::AbstractVector) = _lyap_div!(A, λ)
+function _lyap_div!(A::AbstractMatrix, λ::AbstractVector)
     for (j,μ) in enumerate(λ), (k,λ) in enumerate(λ)
         if k ≠ j
             A[k,j] /= μ - λ
@@ -731,17 +749,21 @@ function _lyap_div!(A, λ)
     A
 end
 
-function LinearAlgebra.eigen(A::Symmetric{<:Dual{Tg,T,N}}) where {Tg,T<:Real,N}
+# To be able to reuse this default definition in the StaticArrays extension
+# (has to be re-defined to avoid method ambiguity issues)
+# we forward the call to an internal method that can be shared and reused
+LinearAlgebra.eigen(A::Symmetric{<:Dual{Tg,T,N}}) where {Tg,T<:Real,N} = _eigen(A)
+function _eigen(A::Symmetric{<:Dual{Tg,T,N}}) where {Tg,T<:Real,N}
     λ = eigvals(A)
     _,Q = eigen(Symmetric(value.(parent(A))))
-    parts = ntuple(j -> Q*_lyap_div!(Q' * getindex.(partials.(A), j) * Q - Diagonal(getindex.(partials.(λ), j)), value.(λ)), N)
+    parts = ntuple(j -> Q*_lyap_div!!(Q' * getindex.(partials.(A), j) * Q - Diagonal(getindex.(partials.(λ), j)), value.(λ)), N)
     Eigen(λ,Dual{Tg}.(Q, tuple.(parts...)))
 end
 
 function LinearAlgebra.eigen(A::SymTridiagonal{<:Dual{Tg,T,N}}) where {Tg,T<:Real,N}
     λ = eigvals(A)
     _,Q = eigen(SymTridiagonal(value.(parent(A))))
-    parts = ntuple(j -> Q*_lyap_div!(Q' * getindex.(partials.(A), j) * Q - Diagonal(getindex.(partials.(λ), j)), value.(λ)), N)
+    parts = ntuple(j -> Q*_lyap_div!!(Q' * getindex.(partials.(A), j) * Q - Diagonal(getindex.(partials.(λ), j)), value.(λ)), N)
     Eigen(λ,Dual{Tg}.(Q, tuple.(parts...)))
 end
 

src/gradient.jl

@@ -13,7 +13,8 @@ This method assumes that `isa(f(x), Real)`.
 
 Set `check` to `Val{false}()` to disable tag checking. This can lead to perturbation confusion, so should be used with care.
 """
-function gradient(f, x::AbstractArray, cfg::GradientConfig{T} = GradientConfig(f, x), ::Val{CHK}=Val{true}()) where {T, CHK}
+function gradient(f::F, x::AbstractArray, cfg::GradientConfig{T} = GradientConfig(f, x), ::Val{CHK}=Val{true}()) where {F, T, CHK}
+    require_one_based_indexing(x)
     CHK && checktag(T, f, x)
     if chunksize(cfg) == length(x)
         return vector_mode_gradient(f, x, cfg)
@@ -32,6 +33,7 @@ This method assumes that `isa(f(x), Real)`.
 
 """
 function gradient!(result::Union{AbstractArray,DiffResult}, f::F, x::AbstractArray, cfg::GradientConfig{T} = GradientConfig(f, x), ::Val{CHK}=Val{true}()) where {T, CHK, F}
+    result isa DiffResult ? require_one_based_indexing(x) : require_one_based_indexing(result, x)
     CHK && checktag(T, f, x)
     if chunksize(cfg) == length(x)
         vector_mode_gradient!(result, f, x, cfg)

src/hessian.jl

@@ -11,7 +11,8 @@ This method assumes that `isa(f(x), Real)`.
 
 Set `check` to `Val{false}()` to disable tag checking. This can lead to perturbation confusion, so should be used with care.
 """
-function hessian(f, x::AbstractArray, cfg::HessianConfig{T} = HessianConfig(f, x), ::Val{CHK}=Val{true}()) where {T,CHK}
+function hessian(f::F, x::AbstractArray, cfg::HessianConfig{T} = HessianConfig(f, x), ::Val{CHK}=Val{true}()) where {F, T,CHK}
+    require_one_based_indexing(x)
     CHK && checktag(T, f, x)
     ∇f = y -> gradient(f, y, cfg.gradient_config, Val{false}())
     return jacobian(∇f, x, cfg.jacobian_config, Val{false}())
@@ -27,7 +28,8 @@ This method assumes that `isa(f(x), Real)`.
 
 Set `check` to `Val{false}()` to disable tag checking. This can lead to perturbation confusion, so should be used with care.
 """
-function hessian!(result::AbstractArray, f, x::AbstractArray, cfg::HessianConfig{T} = HessianConfig(f, x), ::Val{CHK}=Val{true}()) where {T,CHK}
+function hessian!(result::AbstractArray, f::F, x::AbstractArray, cfg::HessianConfig{T} = HessianConfig(f, x), ::Val{CHK}=Val{true}()) where {F,T,CHK}
+    require_one_based_indexing(result, x)
     CHK && checktag(T, f, x)
     ∇f = y -> gradient(f, y, cfg.gradient_config, Val{false}())
     jacobian!(result, ∇f, x, cfg.jacobian_config, Val{false}())
@@ -61,7 +63,7 @@ because `isa(result, DiffResult)`, `cfg` is constructed as `HessianConfig(f, res
 
 Set `check` to `Val{false}()` to disable tag checking. This can lead to perturbation confusion, so should be used with care.
 """
-function hessian!(result::DiffResult, f, x::AbstractArray, cfg::HessianConfig{T} = HessianConfig(f, result, x), ::Val{CHK}=Val{true}()) where {T,CHK}
+function hessian!(result::DiffResult, f::F, x::AbstractArray, cfg::HessianConfig{T} = HessianConfig(f, result, x), ::Val{CHK}=Val{true}()) where {F,T,CHK}
     CHK && checktag(T, f, x)
     ∇f! = InnerGradientForHess(result, cfg, f)
     jacobian!(DiffResults.hessian(result), ∇f!, DiffResults.gradient(result), x, cfg.jacobian_config, Val{false}())