add sinkhorn_divergence to runtests (#146)

* add sinkhorn_divergence to runtests

* fix issues with julia 1.0 compat

* add more compat tag

* remove usage of eachcol

Project.toml

@@ -27,8 +27,8 @@ PythonOT = "3c485715-4278-42b2-9b5f-8f00e43c12ef"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 ReverseDiff = "37e2e3b7-166d-5795-8a7a-e32c996b4267"
 SafeTestsets = "1bc83da4-3b8d-516f-aca4-4fe02f6d838f"
-Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
+Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
 [targets]
 test = ["Distances", "ForwardDiff", "ReverseDiff", "Pkg", "PythonOT", "Random", "SafeTestsets", "Test", "StatsBase"]

src/utils.jl

@@ -25,7 +25,10 @@ end
 dot_matwise(x::AbstractMatrix, y::AbstractArray) = dot_matwise(y, x)
 
 function dot_vecwise(x::AbstractMatrix, y::AbstractMatrix)
-    return [dot(u, v) for (u, v) in zip(eachcol(x), eachcol(y))]
+    return [
+        dot(u, v) for (u, v) in
+        zip((view(x, :, i) for i in axes(x, 2)), (view(y, :, i) for i in axes(y, 2)))
+    ]
 end
 
 dot_vecwise(x::AbstractMatrix, y::AbstractVector) = x' * y

test/entropic/sinkhorn_divergence.jl

@@ -5,7 +5,6 @@ using ForwardDiff
 using ReverseDiff
 using LogExpFunctions
 using PythonOT: PythonOT
-using StatsBase
 using LinearAlgebra
 using Random
 using Test
@@ -54,18 +53,18 @@ Random.seed!(100)
             for reg in (true, false)
                 loss_batch = sinkhorn_divergence(μ, ν, C, ε; regularization=reg)
                 @test loss_batch ≈ [
-                    sinkhorn_divergence(x, y, C, ε; regularization=reg) for
-                    (x, y) in zip(eachcol(μ), eachcol(ν))
+                    sinkhorn_divergence(μ[:, i], ν[:, i], C, ε; regularization=reg) for
+                    i in 1:M
                 ]
                 loss_batch_μ = sinkhorn_divergence(μ, ν[:, 1], C, ε; regularization=reg)
                 @test loss_batch_μ ≈ [
-                    sinkhorn_divergence(x, ν[:, 1], C, ε; regularization=reg) for
-                    x in eachcol(μ)
+                    sinkhorn_divergence(μ[:, i], ν[:, 1], C, ε; regularization=reg) for
+                    i in 1:M
                 ]
                 loss_batch_ν = sinkhorn_divergence(μ[:, 1], ν, C, ε; regularization=reg)
                 @test loss_batch_ν ≈ [
-                    sinkhorn_divergence(μ[:, 1], y, C, ε; regularization=reg) for
-                    y in eachcol(ν)
+                    sinkhorn_divergence(μ[:, 1], ν[:, i], C, ε; regularization=reg) for
+                    i in 1:M
                 ]
             end
         end
@@ -98,7 +97,7 @@ Random.seed!(100)
         Cμν = pairwise(SqEuclidean(), μ_spt', ν_spt'; dims=2)
         Cμ = pairwise(SqEuclidean(), μ_spt'; dims=2)
         Cν = pairwise(SqEuclidean(), ν_spt'; dims=2)
-        ε = 0.1 * max(mean(Cμν), mean(Cμ), mean(Cν))
+        ε = 1.0
 
         @testset "basic" begin
             for reg in (true, false)

test/runtests.jl

@@ -28,6 +28,9 @@ const GROUP = get(ENV, "GROUP", "All")
             @safetestset "Sinkhorn barycenter" begin
                 include(joinpath("entropic", "sinkhorn_barycenter.jl"))
             end
+            @safetestset "Sinkhorn divergence" begin
+                include(joinpath("entropic", "sinkhorn_divergence.jl"))
+            end
         end
 
         @safetestset "Quadratically regularized OT" begin