update to new registration process (#47), require Julia v1.0+

Author: dkarrasch

.travis.yml

@@ -4,13 +4,14 @@ os:
   - linux
   - osx
 julia:
-  - 0.7
   - 1.0
   - 1.1
+  - 1.2
   - nightly
 
 matrix:
   allow_failures:
+    - julia: 1.2
     - julia: nightly
 notifications:
     email: false

Project.toml

@@ -0,0 +1,20 @@
+name = "LinearMaps"
+uuid = "7a12625a-238d-50fd-b39a-03d52299707e"
+version = "2.4.0"
+
+[deps]
+LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
+SparseArrays = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
+
+[compat]
+julia = "≥ 1.0.0"
+
+[extras]
+BenchmarkTools = "6e4b80f9-dd63-53aa-95a3-0cdb28fa8baf"
+LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
+Quaternions = "94ee1d12-ae83-5a48-8b1c-48b8ff168ae0"
+SparseArrays = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
+Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
+
+[targets]
+test = ["LinearAlgebra", "SparseArrays", "Test", "BenchmarkTools", "Quaternions"]

REQUIRE

@@ -2,23 +2,7 @@ using Test
 using LinearMaps
 using SparseArrays
 using LinearAlgebra
-
-# adopted from: https://discourse.julialang.org/t/way-to-return-the-number-of-allocations/5167/10
-macro numalloc(expr)
-    return quote
-        let
-            local f
-            function f()
-                n1 = Base.gc_num()
-                $(expr)
-                n2 = Base.gc_num()
-                diff = Base.GC_Diff(n2, n1)
-                Base.gc_alloc_count(diff)
-            end
-            f()
-        end
-    end
-end
+using BenchmarkTools
 
 A = 2 * rand(ComplexF64, (20, 10)) .- 1
 v = rand(ComplexF64, 10)
@@ -61,7 +45,8 @@ AV = A * V
     @test M * v == Av
     @test N * v == Av
     @test @inferred mul!(copy(w), M, v) == mul!(copy(w), A, v)
-    @test ((@allocated mul!(w, M, v)) == 0)
+    b = @benchmarkable mul!(w, M, v)
+    @test run(b, samples=3).allocs == 0
     @test @inferred mul!(copy(w), N, v) == Av
 
     # mat-vec-mul
@@ -201,12 +186,13 @@ end
 CS! = LinearMap(cumsum!, 10; ismutating=true)
 v = rand(10)
 u = similar(v)
-mul!(u, CS!, v)
-@test ((@allocated mul!(u, CS!, v)) == 0)
+b = @benchmarkable mul!(u, CS!, v)
+@test run(b, samples=3).allocs == 0
 n = 10
 L = sum(fill(CS!, n))
 @test mul!(u, L, v) ≈ n * cumsum(v)
-@test ((@numalloc mul!(u, L, v)) <= 1)
+b = @benchmarkable mul!(u, L, v)
+@test run(b, samples=5).allocs <= 1
 
 A = 2 * rand(ComplexF64, (10, 10)) .- 1
 B = rand(size(A)...)
@@ -215,8 +201,8 @@ N = LinearMap(B)
 LC = M + N
 v = rand(ComplexF64, 10)
 w = similar(v)
-mul!(w, M, v)
-@test ((@allocated mul!(w, M, v)) == 0)
+b = @benchmarkable mul!(w, M, v)
+@test run(b, samples=3).allocs == 0
 @testset "linear combinations" begin
     # @test_throws ErrorException LinearMaps.LinearCombination{ComplexF64}((M, N), (1, 2, 3))
     @test @inferred size(3M + 2.0N) == size(A)
@@ -420,10 +406,19 @@ end
     β = UniformScaling(Quaternion.(rand(4)...))
     L = LinearMap(A)
     @test Array(L) == A
+    @test Array(L') == A'
+    @test Array(transpose(L)) == transpose(A)
     @test Array(α * L) == α * A
     @test Array(L * α) == A * α
     @test Array(α * L) == α * A
-    @test Array(L * α) == A * α
+    @test Array(L * α ) == A * α
+    @test Array(α * L') == α * A'
+    @test Array((α * L')') ≈ (α * A')' ≈ A * conj(α)
+    @test L * x ≈ A * x
+    @test L' * x ≈ A' * x
+    @test α * (L * x) ≈ α * (A * x)
+    @test α * L * x ≈ α * A * x
+    @test (α * L') * x ≈ (α * A') * x
     @test (α * L')' * x ≈ (α * A')' * x
     @test (α * L')' * v ≈ (α * A')' * v
     @test Array(@inferred adjoint(α * L * β)) ≈ conj(β) * A' * conj(α)