Fix and test asymmetric_quadratic example

examples/asymmetric_quadratic.jl

@@ -1,9 +1,11 @@
 using Nabla
+using LinearAlgebra
+using Random
 
 # Generate the values required to compute a matrix quadratic form.
 N = 5
 B = randn(5, 5)
-A = B.'B + UniformScaling(1e-6)
+A = B'B + UniformScaling(1e-6)
 
 # Low-level API computation of derivatives.
 
@@ -12,13 +14,10 @@ x_, y_ = randn(N), randn(N)
 x, y = Leaf.(Tape(), (x_, y_))
 
 # Compute the forward pass.
-z = x.' * (A * y) # Temporary bracketting because we don't support RowVectors yet.
+z = x' * (A * y) # Temporary bracketting because we don't support RowVectors yet.
 
-println("Output of the forward pass is:")
-println(z)
-println()
-println("y is $(Nabla.unbox(z)).")
-println()
+println("Output of the forward pass is:\n $z\n")
+println("y is $(Nabla.unbox(z)).\n")
 
 # Get the reverse tape.
 z̄ = ∇(z)
@@ -42,7 +41,7 @@ println("Gradient of z w.r.t. y at $y_ is $ȳ")
 # Define the function to be differentiated. Parameters w.r.t. which we want gradients must
 # be arguments. Parameters that we don't want gradients w.r.t. should be passed in via a
 # closure.
-@unionise f(x::AbstractVector, y::AbstractVector) = x.'A * y
+@unionise f(x::AbstractVector, y::AbstractVector) = x'A * y
 
 # Compute a function `∇f` which computes the derivative of `f` w.r.t. the inputs.
 ∇f = ∇(f)

test/runtests.jl

@@ -51,4 +51,9 @@ end
     include("checkpointing.jl")
 end
 
+@testset "examples" begin
+    # make sure the examples don't throw errors
+    include("../examples/asymmetric_quadratic.jl")
+end
+
 end