Fix method ambiguity errors with PDMats 0.10

Manifest.toml

@@ -9,11 +9,6 @@ version = "2.0.2"
 [[Base64]]
 uuid = "2a0f44e3-6c83-55bd-87e4-b1978d98bd5f"
 
-[[CEnum]]
-git-tree-sha1 = "215a9aa4a1f23fbd05b92769fdd62559488d70e9"
-uuid = "fa961155-64e5-5f13-b03f-caf6b980ea82"
-version = "0.4.1"
-
 [[CommonSubexpressions]]
 deps = ["MacroTools", "Test"]
 git-tree-sha1 = "7b8a93dba8af7e3b42fecabf646260105ac373f7"
@@ -26,12 +21,6 @@ git-tree-sha1 = "7c4f882c41faa72118841185afc58a2eb00ef612"
 uuid = "e66e0078-7015-5450-92f7-15fbd957f2ae"
 version = "0.3.3+0"
 
-[[CpuId]]
-deps = ["Markdown", "Test"]
-git-tree-sha1 = "f0464e499ab9973b43c20f8216d088b61fda80c6"
-uuid = "adafc99b-e345-5852-983c-f28acb93d879"
-version = "0.2.2"
-
 [[Dates]]
 deps = ["Printf"]
 uuid = "ade2ca70-3891-5945-98fb-dc099432e06a"
@@ -52,12 +41,6 @@ version = "1.0.1"
 deps = ["Random", "Serialization", "Sockets"]
 uuid = "8ba89e20-285c-5b6f-9357-94700520ee1b"
 
-[[DocStringExtensions]]
-deps = ["LibGit2", "Markdown", "Pkg", "Test"]
-git-tree-sha1 = "c5714d9bcdba66389612dc4c47ed827c64112997"
-uuid = "ffbed154-4ef7-542d-bbb7-c09d3a79fcae"
-version = "0.8.2"
-
 [[ForwardDiff]]
 deps = ["CommonSubexpressions", "DiffResults", "DiffRules", "NaNMath", "Random", "SpecialFunctions", "StaticArrays"]
 git-tree-sha1 = "1d090099fb82223abc48f7ce176d3f7696ede36d"
@@ -68,12 +51,6 @@ version = "0.10.12"
 deps = ["Markdown"]
 uuid = "b77e0a4c-d291-57a0-90e8-8db25a27a240"
 
-[[LLVM]]
-deps = ["CEnum", "Libdl", "Printf", "Unicode"]
-git-tree-sha1 = "a662366a5d485dee882077e8da3e1a95a86d097f"
-uuid = "929cbde3-209d-540e-8aea-75f648917ca0"
-version = "2.0.0"
-
 [[LibGit2]]
 deps = ["Printf"]
 uuid = "76f85450-5226-5b5a-8eaa-529ad045b433"
@@ -88,12 +65,6 @@ uuid = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 [[Logging]]
 uuid = "56ddb016-857b-54e1-b83d-db4d58db5568"
 
-[[LoopVectorization]]
-deps = ["DocStringExtensions", "LinearAlgebra", "OffsetArrays", "SIMDPirates", "SLEEFPirates", "UnPack", "VectorizationBase"]
-git-tree-sha1 = "b595e15d20e45d2eb36c6b4462d2a34143872a45"
-uuid = "bdcacae8-1622-11e9-2a5c-532679323890"
-version = "0.8.15"
-
 [[MacroTools]]
 deps = ["Markdown", "Random"]
 git-tree-sha1 = "f7d2e3f654af75f01ec49be82c231c382214223a"
@@ -104,27 +75,16 @@ version = "0.5.5"
 deps = ["Base64"]
 uuid = "d6f4376e-aef5-505a-96c1-9c027394607a"
 
-[[NNPACK_jll]]
-deps = ["Libdl", "Pkg"]
-git-tree-sha1 = "c3d1a616362645754b18e12dbba96ec311b0867f"
-uuid = "a6bfbf70-4841-5cb9-aa18-3a8ad3c413ee"
-version = "2018.6.22+0"
-
 [[NNlib]]
-deps = ["Libdl", "LinearAlgebra", "LoopVectorization", "NNPACK_jll", "Pkg", "Requires", "Statistics"]
-git-tree-sha1 = "f593bdb98b00a4f5b87cc2c18231b81433111590"
+deps = ["Libdl", "LinearAlgebra", "Pkg", "Requires", "Statistics"]
+git-tree-sha1 = "8ec4693a5422f0b064ce324f59351f24aa474893"
 uuid = "872c559c-99b0-510c-b3b7-b6c96a88d5cd"
-version = "0.7.2"
+version = "0.7.4"
 
 [[NaNMath]]
-git-tree-sha1 = "928b8ca9b2791081dc71a51c55347c27c618760f"
+git-tree-sha1 = "c84c576296d0e2fbb3fc134d3e09086b3ea617cd"
 uuid = "77ba4419-2d1f-58cd-9bb1-8ffee604a2e3"
-version = "0.3.3"
-
-[[OffsetArrays]]
-git-tree-sha1 = "4ba4cd84c88df8340da1c3e2d8dcb9d18dd1b53b"
-uuid = "6fe1bfb0-de20-5000-8ca7-80f57d26f881"
-version = "1.1.1"
+version = "0.3.4"
 
 [[OpenSpecFun_jll]]
 deps = ["CompilerSupportLibraries_jll", "Libdl", "Pkg"]
@@ -157,18 +117,6 @@ version = "1.0.1"
 [[SHA]]
 uuid = "ea8e919c-243c-51af-8825-aaa63cd721ce"
 
-[[SIMDPirates]]
-deps = ["VectorizationBase"]
-git-tree-sha1 = "dae629e96c1819d77882256e6cb29736f493bc30"
-uuid = "21efa798-c60a-11e8-04d3-e1a92915a26a"
-version = "0.8.13"
-
-[[SLEEFPirates]]
-deps = ["Libdl", "SIMDPirates", "VectorizationBase"]
-git-tree-sha1 = "c750d618b7c8268a97e55c70e8c88e56080d30fa"
-uuid = "476501e8-09a2-5ece-8869-fb82de89a1fa"
-version = "0.5.4"
-
 [[Serialization]]
 uuid = "9e88b42a-f829-5b0c-bbe9-9e923198166b"
 
@@ -203,16 +151,5 @@ uuid = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 deps = ["Random", "SHA"]
 uuid = "cf7118a7-6976-5b1a-9a39-7adc72f591a4"
 
-[[UnPack]]
-git-tree-sha1 = "d4bfa022cd30df012700cf380af2141961bb3bfb"
-uuid = "3a884ed6-31ef-47d7-9d2a-63182c4928ed"
-version = "1.0.1"
-
 [[Unicode]]
 uuid = "4ec0a83e-493e-50e2-b9ac-8f72acf5a8f5"
-
-[[VectorizationBase]]
-deps = ["CpuId", "LLVM", "Libdl", "LinearAlgebra"]
-git-tree-sha1 = "bb72c58beab6c9e544851f5373fcd72f8f1f157a"
-uuid = "3d5dd08c-fd9d-11e8-17fa-ed2836048c2f"
-version = "0.12.21"

Project.toml

@@ -1,6 +1,6 @@
 name = "Tracker"
 uuid = "9f7883ad-71c0-57eb-9f7f-b5c9e6d3789c"
-version = "0.2.10"
+version = "0.2.11"
 
 [deps]
 Adapt = "79e6a3ab-5dfb-504d-930d-738a2a938a0e"
@@ -15,7 +15,6 @@ Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 Requires = "ae029012-a4dd-5104-9daa-d747884805df"
 SpecialFunctions = "276daf66-3868-5448-9aa4-cd146d93841b"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
-Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
 [compat]
 Adapt = "1,2"
@@ -27,3 +26,10 @@ NaNMath = "0"
 Requires = "0.5, 1.0"
 SpecialFunctions = "0"
 julia = "1.3"
+
+[extras]
+PDMats = "90014a1f-27ba-587c-ab20-58faa44d9150"
+Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
+
+[targets]
+test = ["PDMats", "Test"]

src/Tracker.jl

@@ -69,6 +69,7 @@ include("numeric.jl")
 include("lib/real.jl")
 include("lib/array.jl")
 include("forward.jl")
+@init @require PDMats="90014a1f-27ba-587c-ab20-58faa44d9150" include("lib/pdmats.jl")
 
 """
     hook(f, x) -> x′

src/lib/array.jl

@@ -329,23 +329,25 @@ end
 A::TrackedArray     / B::TrackedArray     = Tracker.track(/, A, B)
 A::AbstractVecOrMat / B::TrackedArray     = Tracker.track(/, A, B)
 A::TrackedArray     / B::AbstractVecOrMat = Tracker.track(/, A, B)
-@grad function (A / B)
+@grad function Base.:/(A, B)
     return Tracker.data(A) / Tracker.data(B), function (Δ)
-        Binv = inv(B)
-        ∇B = - Binv' * A' * Δ * Binv'
-        return (Δ * Binv',  ∇B)
+        ∇A = Δ / B'
+        ∇B = - (A / B)' * ∇A
+        return (∇A, ∇B)
     end
 end
 
 #       (\) ldivide  (left vec divide needs more work to resolve dispatch ambiguity)
 A::TrackedArray     \ B::TrackedArray     = Tracker.track(\, A, B)
 A::AbstractArray    \ B::TrackedArray     = Tracker.track(\, A, B)
 A::TrackedArray     \ B::AbstractVecOrMat = Tracker.track(\, A, B)
-@grad function (A \ B)
+A::AbstractMatrix   \ B::TrackedVecOrMat  = Tracker.track(\, A, B)
+A::TrackedMatrix    \ B::TrackedVecOrMat  = Tracker.track(\, A, B)
+@grad function Base.:\(A, B)
     return Tracker.data(A) \ Tracker.data(B), function (Δ)
-        Ainv = inv(A)
-        ∇A = - Ainv' * Δ * B' * Ainv'
-        return (∇A,  Ainv' * Δ)
+        ∇B = A' \ Δ
+        ∇A = - ∇B * (A \ B)'
+        return (∇A, ∇B)
     end
 end
 

src/lib/pdmats.jl

@@ -0,0 +1,5 @@
+using .PDMats
+
+Base.:\(A::PDMat, B::TrackedVecOrMat) = Tracker.track(\, A, B)
+Base.:\(A::PDiagMat, B::TrackedVecOrMat) = Tracker.track(\, A, B)
+Base.:\(A::ScalMat, B::TrackedVecOrMat) = Tracker.track(\, A, B)

test/tracker.jl

@@ -1,8 +1,9 @@
 using Tracker, Test, NNlib
 using Tracker: TrackedReal, gradient, gradcheck, grad, checkpoint, forwarddiff
 using NNlib: conv, ∇conv_data, depthwiseconv
+using PDMats
 using Printf: @sprintf
-using LinearAlgebra: diagm, dot, LowerTriangular, norm, det, logdet, logabsdet
+using LinearAlgebra: diagm, dot, LowerTriangular, norm, det, logdet, logabsdet, I
 using Statistics: mean, std
 using Random
 # using StatsBase
@@ -152,7 +153,20 @@ end
 @test gradtest(W -> inv(log.(W * W)), (5,5))
 @test gradtest((A, B) -> A / B , (1,5), (5,5))
 @test gradtest((A, B) -> log.(A * A) / exp.(B * B), (5,5), (5,5))
+@test gradtest((A, B) -> A \ B, (5, 5), (5,))
+@test let A=rand(5, 5)
+    gradtest(B -> A \ B, (5,))
+end
+@test let B=rand(5,)
+    gradtest(A -> A \ B, (5, 5))
+end
 @test gradtest((A, B) -> log.(A * A) \ exp.(B * B), (5,5), (5,5))
+@test let A=rand(5, 5)
+    gradtest(B -> log.(A * A) \ exp.(B * B), (5, 5))
+end
+@test let B=rand(5, 5)
+    gradtest(A -> log.(A * A) \ exp.(B * B), (5, 5))
+end
 
 @testset "mean" begin
   @test gradtest(mean, rand(2, 3))
@@ -432,4 +446,16 @@ end
   @test back([1, 1]) == (32,)
 end
 
+@testset "PDMats" begin
+    B = rand(5, 5)
+    S = PDMat(I + B * B')
+    @test gradtest(A -> S / A, (5, 5))
+
+    S = PDiagMat(rand(5))
+    @test gradtest(A -> S / A, (5, 5))
+
+    S = ScalMat(5, rand())
+    @test gradtest(A -> S / A, (5, 5))
+end
+
 end #testset