Gradient definitions & supertypes

src/zygote.jl

@@ -43,10 +43,28 @@ function ChainRulesCore.rrule(::typeof(getproperty),A::ArrayPartition, s::Symbol
     A.x,literal_ArrayPartition_x_adjoint
 end
 
+#=
+
+# Define a new species of projection operator for this type:
+ChainRulesCore.ProjectTo(x::VectorOfArray) = ProjectTo{VectorOfArray}()
+
+# Gradient from iteration will be e.g. Vector{Vector}, this makes it another AbstractMatrix
+(::ProjectTo{VectorOfArray})(dx::AbstractVector{<:AbstractArray}) = VectorOfArray(dx)
+# Gradient from broadcasting will be another AbstractArray
+(::ProjectTo{VectorOfArray})(dx::AbstractArray) = dx
+
+But this may not be necessary?
+
+=#
+
+
+# These rules duplicate the `rrule` methods above, because Zygote looks for an `@adjoint`
+# definition first, and finds its own before finding those.
+
 ZygoteRules.@adjoint function getindex(VA::AbstractVectorOfArray, i::Union{Int,AbstractArray{Int},CartesianIndex,Colon,BitArray,AbstractArray{Bool}})
   function AbstractVectorOfArray_getindex_adjoint(Δ)
     Δ′ = [ (i == j ? Δ : zero(x)) for (x,j) in zip(VA.u, 1:length(VA))]
-    (Δ′,nothing)
+    (VectorOfArray(Δ′),nothing)
   end
   VA[i],AbstractVectorOfArray_getindex_adjoint
 end
@@ -55,11 +73,13 @@ ZygoteRules.@adjoint function getindex(VA::AbstractVectorOfArray, i::Union{Int,A
   function AbstractVectorOfArray_getindex_adjoint(Δ)
     Δ′ = zero(VA)
     Δ′[i,j...] = Δ
-    (Δ′, i,map(_ -> nothing, j)...)
+    @show Δ′
+    # (Δ′, i,map(_ -> nothing, j)...)  # surely that i is a bug?
+    (Δ′, nothing, map(_ -> nothing, j)...)
+    # (VectorOfArray(Δ′), nothing, map(_ -> nothing, j)...)
   end
   VA[i,j...],AbstractVectorOfArray_getindex_adjoint
 end
-
 ZygoteRules.@adjoint function ArrayPartition(x::S, ::Type{Val{copy_x}} = Val{false}) where {S<:Tuple,copy_x}
   function ArrayPartition_adjoint(_y)
       y = Array(_y)