diffeqdifftools -> FiniteDiff

Author: ChrisRackauckas

Project.toml

@@ -8,8 +8,8 @@ Adapt = "79e6a3ab-5dfb-504d-930d-738a2a938a0e"
 ArrayInterface = "4fba245c-0d91-5ea0-9b3e-6abc04ee57a9"
 Compat = "34da2185-b29b-5c13-b0c7-acf172513d20"
 DataStructures = "864edb3b-99cc-5e75-8d2d-829cb0a9cfe8"
-DiffEqDiffTools = "01453d9d-ee7c-5054-8395-0335cb756afa"
 ForwardDiff = "f6369f11-7733-5829-9624-2563aa707210"
+FiniteDiff = "6a86dc24-6348-571c-b903-95158fe2bd41"
 InteractiveUtils = "b77e0a4c-d291-57a0-90e8-8db25a27a240"
 LightGraphs = "093fc24a-ae57-5d10-9952-331d41423f4d"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
@@ -22,7 +22,7 @@ Adapt = "1"
 ArrayInterface = "1.1, 2.0"
 Compat = "2.2, 3"
 DataStructures = "0.17"
-DiffEqDiffTools = "1.7"
+FiniteDiff = "2"
 ForwardDiff = "0.10"
 LightGraphs = "1.3"
 Requires = "0.5, 1.0"

README.md

@@ -71,8 +71,8 @@ pattern used 1 (pseudo) `f`-evaluation, so the total number of times that
 is 5 times:
 
 ```julia
-using DiffEqDiffTools
-DiffEqDiffTools.finite_difference_jacobian!(jac, f, rand(30), colorvec=colors)
+using FiniteDiff
+FiniteDiff.finite_difference_jacobian!(jac, f, rand(30), colorvec=colors)
 @show fcalls # 5
 ```
 
@@ -141,12 +141,12 @@ The result is a vector which assigns a colorvec to each column (or row) of the m
 ### Colorvec-Assisted Differentiation
 
 Colorvec-assisted differentiation for numerical differentiation is provided by
-DiffEqDiffTools.jl and for automatic differentiation is provided by
+FiniteDiff.jl and for automatic differentiation is provided by
 ForwardDiff.jl.
 
-For DiffEqDiffTools.jl, one simply has to use the provided `colorvec` keyword
+For FiniteDiff.jl, one simply has to use the provided `colorvec` keyword
 argument. See
-[the DiffEqDiffTools Jacobian documentation](https://github.com/JuliaDiff/DiffEqDiffTools.jl#jacobians)
+[the FiniteDiff Jacobian documentation](https://github.com/JuliaDiff/FiniteDiff.jl#jacobians)
 for more details.
 
 For forward-mode automatic differentiation, use of a colorvec vector is provided

src/SparseDiffTools.jl

@@ -1,7 +1,7 @@
 module SparseDiffTools
 
 using Compat
-using DiffEqDiffTools
+using FiniteDiff
 using ForwardDiff
 using LightGraphs
 using Requires

src/differentiation/compute_jacobian_ad.jl

@@ -153,7 +153,7 @@ function forwarddiff_color_jacobian!(J::AbstractMatrix{<:Number},
     maxcolor = maximum(colorvec)
     fill!(J, zero(eltype(J)))
 
-    if DiffEqDiffTools._use_findstructralnz(sparsity)
+    if FiniteDiff._use_findstructralnz(sparsity)
         rows_index, cols_index = ArrayInterface.findstructralnz(sparsity)
     else
         rows_index = nothing
@@ -176,7 +176,7 @@ function forwarddiff_color_jacobian!(J::AbstractMatrix{<:Number},
                 dx .= partials.(fx, j)
                 if ArrayInterface.fast_scalar_indexing(dx)
                     #dx is implicitly used in vecdx
-                    DiffEqDiffTools._colorediteration!(J,sparsity,rows_index,cols_index,vecdx,colorvec,color_i,ncols)
+                    FiniteDiff._colorediteration!(J,sparsity,rows_index,cols_index,vecdx,colorvec,color_i,ncols)
                 else
                     #=
                     J.nzval[rows_index] .+= (colorvec[cols_index] .== color_i) .* dx[rows_index]

src/differentiation/jaches_products.jl

@@ -37,9 +37,9 @@ function num_hesvec!(du,f,x,v,
                      cache1 = similar(v),
                      cache2 = similar(v),
                      cache3 = similar(v))
-    cache = DiffEqDiffTools.GradientCache(v[1],cache1,Val{:central})
+    cache = FiniteDiff.GradientCache(v[1],cache1,Val{:central})
     g = let f=f,cache=cache
-        (dx,x) -> DiffEqDiffTools.finite_difference_gradient!(dx,f,x,cache)
+        (dx,x) -> FiniteDiff.finite_difference_gradient!(dx,f,x,cache)
     end
     T = eltype(x)
     # Should it be min? max? mean?
@@ -52,7 +52,7 @@ function num_hesvec!(du,f,x,v,
 end
 
 function num_hesvec(f,x,v)
-    g = (x) -> DiffEqDiffTools.finite_difference_gradient(f,x)
+    g = (x) -> FiniteDiff.finite_difference_gradient(f,x)
     T = eltype(x)
     # Should it be min? max? mean?
     ϵ = sqrt(eps(real(T))) * max(one(real(T)), abs(norm(x)))
@@ -96,15 +96,15 @@ function autonum_hesvec!(du,f,x,v,
                      cache1 = similar(v),
                      cache2 = ForwardDiff.Dual{DeivVecTag}.(x, v),
                      cache3 = ForwardDiff.Dual{DeivVecTag}.(x, v))
-    cache = DiffEqDiffTools.GradientCache(v[1],cache1,Val{:central})
-    g = (dx,x) -> DiffEqDiffTools.finite_difference_gradient!(dx,f,x,cache)
+    cache = FiniteDiff.GradientCache(v[1],cache1,Val{:central})
+    g = (dx,x) -> FiniteDiff.finite_difference_gradient!(dx,f,x,cache)
     cache2 .= Dual{DeivVecTag}.(x, v)
     g(cache3,cache2)
     du .= partials.(cache3, 1)
 end
 
 function autonum_hesvec(f,x,v)
-    g = (x) -> DiffEqDiffTools.finite_difference_gradient(f,x)
+    g = (x) -> FiniteDiff.finite_difference_gradient(f,x)
     partials.(g(Dual{DeivVecTag}.(x, v)), 1)
 end
 

test/test_ad.jl

@@ -203,7 +203,7 @@ _Jt = similar(Tridiagonal(J))
 forwarddiff_color_jacobian!(_Jt, f, x, colorvec = repeat(1:3,10), sparsity = _Jt)
 @test _Jt ≈ J
 
-#https://github.com/JuliaDiff/DiffEqDiffTools.jl/issues/67#issuecomment-516871956
+#https://github.com/JuliaDiff/FiniteDiff.jl/issues/67#issuecomment-516871956
 function f(out, x)
 	x = reshape(x, 100, 100)
 	out = reshape(out, 100, 100)

test/test_integration.jl

@@ -1,5 +1,5 @@
 using SparseDiffTools
-using DiffEqDiffTools: finite_difference_jacobian, finite_difference_jacobian!
+using FiniteDiff: finite_difference_jacobian, finite_difference_jacobian!
 using SparsityDetection
 
 using LinearAlgebra, SparseArrays, Test

test/test_jaches_products.jl

@@ -1,4 +1,4 @@
-using SparseDiffTools, ForwardDiff, DiffEqDiffTools, Zygote, IterativeSolvers
+using SparseDiffTools, ForwardDiff, FiniteDiff, Zygote, IterativeSolvers
 using LinearAlgebra, Test
 
 using Random
@@ -12,10 +12,10 @@ v = rand(300)
 du = similar(x)
 g(u) = sum(abs2,u)
 function h(x)
-      DiffEqDiffTools.finite_difference_gradient(g,x)
+      FiniteDiff.finite_difference_gradient(g,x)
 end
 function h(dx,x)
-      DiffEqDiffTools.finite_difference_gradient!(dx,g,x)
+      FiniteDiff.finite_difference_gradient!(dx,g,x)
 end
 
 cache1 = ForwardDiff.Dual{SparseDiffTools.DeivVecTag}.(x, v)