Move to Julia 0.7 (#75)

* Update REQUIRE and update code for Julia 0.7
* Remove Compat
* Set seed in autodiff tests

.travis.yml

@@ -4,7 +4,7 @@ os:
   - linux
   - osx
 julia:
-  - 0.6
+  - 0.7
   - nightly
 notifications:
   email: false

REQUIRE

@@ -1,5 +1,4 @@
-julia 0.6
-Compat
+julia 0.7-alpha
 DiffEqDiffTools
 ForwardDiff
 Calculus

appveyor.yml

@@ -1,12 +1,11 @@
 environment:
   matrix:
-  - JULIA_URL: "https://julialang-s3.julialang.org/bin/winnt/x86/0.6/julia-0.6-latest-win32.exe"
-  - JULIA_URL: "https://julialang-s3.julialang.org/bin/winnt/x64/0.6/julia-0.6-latest-win64.exe"
+  - JULIA_URL: "https://julialang-s3.julialang.org/bin/winnt/x86/0.7/julia-0.7-latest-win32.exe"
+  - JULIA_URL: "https://julialang-s3.julialang.org/bin/winnt/x64/0.7/julia-0.7-latest-win64.exe"
 matrix:
   allow_failures:
     - JULIA_URL: "https://julialangnightlies-s3.julialang.org/bin/winnt/x86/julia-latest-win32.exe"
     - JULIA_URL: "https://julialangnightlies-s3.julialang.org/bin/winnt/x64/julia-latest-win64.exe"
-
 branches:
   only:
     - master

src/NLSolversBase.jl

@@ -2,10 +2,8 @@ __precompile__(true)
 
 module NLSolversBase
 
-using Compat
 using DiffEqDiffTools, ForwardDiff, Calculus
-import Compat.Distributed: clear!
-import Compat.LinearAlgebra: gradient
+import Distributed: clear!
 export AbstractObjective,
        NonDifferentiable,
        OnceDifferentiable,

src/objective_types/constraints.jl

@@ -71,8 +71,8 @@ function nconstraints_x(cb::ConstraintBounds)
     me = isempty(cb.eqx) ? 0 : maximum(cb.eqx)
     nmax = max(mi, me)
     hasconstraint = falses(nmax)
-    hasconstraint[cb.ineqx] = true
-    hasconstraint[cb.eqx] = true
+    hasconstraint[cb.ineqx] .= true
+    hasconstraint[cb.eqx] .= true
     sum(hasconstraint)
 end
 

src/objective_types/incomplete.jl

@@ -73,15 +73,15 @@ function NonDifferentiable(t::Union{InplaceObjective, NotInplaceObjective}, x::A
 end
 
 
-function TwiceDifferentiable(t::InplaceObjective{<: Void, <: Void, TH}, x::AbstractArray, F::Real = real(zero(eltype(x))), G::AbstractArray = similar(x), H = alloc_H(x)) where {TH}
+function TwiceDifferentiable(t::InplaceObjective{<: Nothing, <: Nothing, TH}, x::AbstractArray, F::Real = real(zero(eltype(x))), G::AbstractArray = similar(x), H = alloc_H(x)) where {TH}
     f   =     x  -> t.fgh(F, nothing, nothing, x)
     df  = (G, x) -> t.fgh(nothing, G, nothing, x)
     fdf = (G, x) -> t.fgh(F, G, nothing, x)
     h   = (H, x) -> t.fgh(F, nothing, H, x)
     TwiceDifferentiable(f, df, fdf, h, x, F, G, H)
 end
 
-function TwiceDifferentiable(t::InplaceObjective{<: Void, <: Void, TH}, x::AbstractVector, F::Real = real(zero(eltype(x))), G::AbstractVector = similar(x)) where {TH}
+function TwiceDifferentiable(t::InplaceObjective{<: Nothing, <: Nothing, TH}, x::AbstractVector, F::Real = real(zero(eltype(x))), G::AbstractVector = similar(x)) where {TH}
 
     H = alloc_H(x)
     f   =     x  -> t.fgh(F, nothing, nothing, x)

src/objective_types/twicedifferentiablehv.jl

@@ -30,7 +30,7 @@ end
 
 function gradient!!(obj::TwiceDifferentiableHV, x)
     obj.df_calls .+= 1
-    copy!(obj.x_df, x)
+    copyto!(obj.x_df, x)
     obj.fdf(obj.DF, x)
 end
 

test/REQUIRE

@@ -1,2 +1,2 @@
-OptimTestProblems 1.2
+OptimTestProblems 2.0
 RecursiveArrayTools

test/autodiff.jl

@@ -1,4 +1,6 @@
 @testset "autodiff" begin
+    srand(0)
+
     # Should throw, as :wah is not a proper autodiff choice
     @test_throws ErrorException OnceDifferentiable(x->x, rand(10); autodiff=:wah)
     @test_throws ErrorException OnceDifferentiable(x->x, rand(10), 0.0; autodiff=:wah)
@@ -46,9 +48,9 @@
     x = rand(nx)
     f(x) = sum(x.^3)
     fx = f(x)
-    g(G, x) = copyto!(G, 3.*x.^2)
+    g(G, x) = copyto!(G, 3 .* x.^2)
     gx = g(NLSolversBase.alloc_DF(x, 0.0), x)
-    h(H, x) = copyto!(H, Diagonal(6.*x))
+    h(H, x) = copyto!(H, Diagonal(6 .* x))
     hx = h(fill(0.0, nx, nx), x)
     for dtype in (OnceDifferentiable, TwiceDifferentiable)
         for autodiff in (:finite, :forward)
@@ -77,7 +79,7 @@
         end
     end
     for autodiff in (:finite, :forward)
-        td = TwiceDifferentiable(x->sum(x), (G, x)->copyto!(G, ones(x)), copy(x); autodiff = autodiff)
+        td = TwiceDifferentiable(x->sum(x), (G, x)->copyto!(G, fill!(copy(x),1)), copy(x); autodiff = autodiff)
         value(td)
         value!(td, x)
         value_gradient!(td, x)
@@ -93,7 +95,7 @@
             gradient!(td, x)
             hessian!(td, x)
         end
-        for od = (OnceDifferentiable(x->sum(x), (G, x)->copyto!(G, ones(x)), copy(x)), OnceDifferentiable(x->sum(x), (G, x)->copyto!(G, ones(x)), copy(x), 0.0))
+        for od = (OnceDifferentiable(x->sum(x), (G, x)->copyto!(G, fill!(copy(x),1)), copy(x)), OnceDifferentiable(x->sum(x), (G, x)->copyto!(G, fill!(copy(x),1)), copy(x), 0.0))
             td = TwiceDifferentiable(od; autodiff = autodiff)
             value(td)
             value!(td, x)

test/incomplete.jl

@@ -85,7 +85,6 @@
 
 end
 @testset "incomplete objectives vectors" begin
-    import Compat: copyto!
     function tf(x)
         x.^2
     end

test/interface.jl

@@ -153,7 +153,7 @@
                 if prob.istwicedifferentiable
                     hv!(storage::Vector, x::Vector, v::Vector) = begin
                         n = length(x)
-                        H = Matrix{Float64}(n, n)
+                        H = Matrix{Float64}(undef, n, n)
                         MVP.hessian(prob)(H, x)
                         storage .= H * v
                     end

test/objective_types.jl

@@ -17,7 +17,7 @@
     @test od.df_calls == [0]
     od.x_df .= x_seed
     gold = copy(od.DF)
-    xnew = rand(size(x_seed))
+    xnew = rand(eltype(x_seed), size(x_seed))
     gnew = gradient(od, xnew)
     @test od.x_df == x_seed
     @test od.DF == gold

test/runtests.jl

@@ -1,11 +1,13 @@
-using NLSolversBase
-using Compat, Compat.Test, Compat.Random, Compat.LinearAlgebra
+using NLSolversBase, Test
+using Random
+using LinearAlgebra: Diagonal, I
+using SparseArrays
 using OptimTestProblems
 using RecursiveArrayTools
 MVP = OptimTestProblems.MultivariateProblems
 
-# TODO: Use OptimTestProblems
-# TODO: MultivariateProblems.UnconstrainedProblems.exampples["Exponential"]
+# TODO: Use OptimTestProblems (but it does not have exponential_gradient_hession etc.)
+# TODO: MultivariateProblems.UnconstrainedProblems.examples["Exponential"]
 
 # Test example
 function exponential(x)

test/sparse.jl

@@ -1,13 +1,13 @@
 @testset "sparse" begin
     @testset "𝐑ⁿ → 𝐑" begin
         f(x) = sum(x->x.^2, x)
-        g(G, x) = copy!(G, 2.*x)
-        h(H, x) = H .= speye(size(H)...).*2
+        g(G, x) = copy!(G, 2 .* x)
+        h(H, x) = H .= sparse(2.0I, size(H)...)
 
         obj_dense = TwiceDifferentiable(f, g, h, rand(40))
         @test !issparse(obj_dense.H)
 
-        obj_sparse = TwiceDifferentiable(f, g, h, rand(40), 0.0, rand(40), speye(40))
+        obj_sparse = TwiceDifferentiable(f, g, h, rand(40), 0.0, rand(40), sparse(1.0I, 40, 40))
         @test typeof(obj_sparse.H) <: SparseMatrixCSC
 
         function fgh!(F, G, H, x)
@@ -16,20 +16,20 @@
             end
 
             if !(G == nothing)
-                copy!(G, 2.*x)
+                copy!(G, 2 .* x)
             end
             if !(H == nothing)
-                H .= speye(size(H)...).*2
+                H .= sparse(2.0I, size(H)...)
             end
             return fx
         end
 
-        obj_fgh = TwiceDifferentiable(NLSolversBase.only_fgh!(fgh!), rand(40), 0.0, rand(40), speye(40))
+        obj_fgh = TwiceDifferentiable(NLSolversBase.only_fgh!(fgh!), rand(40), 0.0, rand(40), sparse(1.0I, 40, 40))
         @test typeof(obj_fgh.H) <: SparseMatrixCSC
     end
     @testset "𝐑ⁿ → 𝐑ⁿ" begin
-        f(F, x) = copy!(F, 2.*x)
-        j(J, x) = J .= speye(size(J)...).*2
+        f(F, x) = copy!(F, 2 .* x)
+        j(J, x) = J .= sparse(2.0I, size(J)...)
 
         # Test that with no spec on the Jacobian cache it is dense
         obj_dense = OnceDifferentiable(f, j, rand(40), rand(40))
@@ -39,20 +39,20 @@
         @test !issparse(obj_dense.DF)
 
 
-        obj_sparse = OnceDifferentiable(f, j, rand(40), rand(40), speye(40))
+        obj_sparse = OnceDifferentiable(f, j, rand(40), rand(40), sparse(1.0I, 40, 40))
         @test typeof(obj_sparse.DF) <: SparseMatrixCSC
 
         function fj!(F, J, x)
             if !(F == nothing)
-                copy!(G, 2.*x)
+                copy!(G, 2 .* x)
             end
             if !(J == nothing)
-                J .= speye(size(H)...).*2
+                J .= sparse(2.0I, size(J)...)
             end
             return fx
         end
 
-        obj_fj = OnceDifferentiable(NLSolversBase.only_fj!(fj!), rand(40), rand(40), speye(40))
+        obj_fj = OnceDifferentiable(NLSolversBase.only_fj!(fj!), rand(40), rand(40), sparse(1.0I, 40, 40))
         @test typeof(obj_fj.DF) <: SparseMatrixCSC
     end
 end