Using a strong zero in Dual(0.0, 1)^0 to avoid NaN (#84)

* test and fix for nan epsilon of dual^UInt64(0)

* "0.6.3" -> "0.6.4"

* removed NaNs, codified behaviour with Ints

* improve testing, remove type instability

* handle Dual(Integer(1), n)^Integer

* version semver bump 0.6.4 -> 0.6.5

Author: jwscook

Project.toml

@@ -1,6 +1,6 @@
 name = "DualNumbers"
 uuid = "fa6b7ba4-c1ee-5f82-b5fc-ecf0adba8f74"
-version = "0.6.4"
+version = "0.6.5"
 
 [deps]
 Calculus = "49dc2e85-a5d0-5ad3-a950-438e2897f1b9"

src/dual.jl

@@ -245,29 +245,40 @@ Base.:/(z::Number, w::Dual) = Dual(z/value(w), -z*epsilon(w)/value(w)^2)
 Base.:/(z::Dual, x::Number) = Dual(value(z)/x, epsilon(z)/x)
 
 for f in [:(Base.:^), :(NaNMath.pow)]
-    @eval function ($f)(z::Dual, w::Dual)
-        if epsilon(w) == 0.0
-            return $f(z, value(w))
-        end
+    @eval function ($f)(z::Dual{T1}, w::Dual{T2}) where {T1, T2}
+        T = promote_type(T1, T2) # for type stability in ? : statements
         val = $f(value(z), value(w))
 
-        du = epsilon(z) * value(w) * $f(value(z), value(w) - 1) +
-             epsilon(w) * $f(value(z), value(w)) * log(value(z))
+        ezvw = epsilon(z) * value(w) # for using in ? : statement
+        du1 = iszero(ezvw) ? zero(T) : ezvw * $f(value(z), value(w) - 1)
+        ew = epsilon(w) # for using in ? : statement
+        # the float is for type stability because log promotes to floats
+        du2 = iszero(ew) ? zero(float(T)) : ew * val * log(value(z))
+        du = du1 + du2
 
         Dual(val, du)
     end
 end
 
 Base.mod(z::Dual, n::Number) = Dual(mod(value(z), n), epsilon(z))
 
-# these two definitions are needed to fix ambiguity warnings
-Base.:^(z::Dual, n::Unsigned) = z^Signed(n)
-Base.:^(z::Dual, n::Integer) = Dual(value(z)^n, epsilon(z)*n*value(z)^(n-1))
-Base.:^(z::Dual, n::Rational) = Dual(value(z)^n, epsilon(z)*n*value(z)^(n-1))
+# introduce a boolean !iszero(n) for hard zero behaviour to combat NaNs
+function pow(z::Dual, n::AbstractFloat)
+    return Dual(value(z)^n, !iszero(n) * (epsilon(z) * n * value(z)^(n - 1)))
+end
+function pow(z::Dual{T}, n::Integer) where T
+    iszero(n) && return Dual(one(T), zero(T)) # avoid DomainError Int^(negative Int)
+    isone(z) && return Dual(one(T), epsilon(z) * n)
+    return Dual(value(z)^n, epsilon(z) * n * value(z)^(n - 1))
+end
+# these first two definitions are needed to fix ambiguity warnings
+for T1 ∈ (:Integer, :Rational, :Number)
+    @eval Base.:^(z::Dual{T}, n::$T1) where T = pow(z, n)
+end
+
 
-Base.:^(z::Dual, n::Number) = Dual(value(z)^n, epsilon(z)*n*value(z)^(n-1))
-NaNMath.pow(z::Dual, n::Number) = Dual(NaNMath.pow(value(z),n), epsilon(z)*n*NaNMath.pow(value(z),n-1))
-NaNMath.pow(z::Number, w::Dual) = Dual(NaNMath.pow(z,value(w)), epsilon(w)*NaNMath.pow(z,value(w))*log(z))
+NaNMath.pow(z::Dual{T}, n::Number) where T = Dual(NaNMath.pow(value(z),n), epsilon(z)*n*NaNMath.pow(value(z),n-1))
+NaNMath.pow(z::Number, w::Dual{T}) where T = Dual(NaNMath.pow(z,value(w)), epsilon(w)*NaNMath.pow(z,value(w))*log(z))
 
 Base.inv(z::Dual) = dual(inv(value(z)),-epsilon(z)/value(z)^2)
 

test/automatic_differentiation_test.jl

@@ -15,8 +15,52 @@ y = x^3.0
 @test value(y) ≈ 2.0^3
 @test epsilon(y) ≈ 3.0*2^2
 
+# taking care with divides by zero where there shouldn't be any on paper
+for (y, n) ∈ Iterators.product((float(x), Dual(0.0, 1)), (0, 0.0))
+  z = y^n
+  @test value(z) == 1
+  @test !isnan(epsilon(z))
+  @test epsilon(z) == 0
+end
+
+# acting on floats works as expected
+for (y, n) ∈ ((float(x), Dual(0.0, 1)), -1:1)
+  @test float(y)^n == float(y)^float(n)
+end
+
+@test !isnan(epsilon(Dual(0, 1)^1))
+@test Dual(0, 1)^1 == Dual(0, 1)
+
+# power_by_squaring error for integers
+# needs to be wrapped to make n a literal
+powwrap(z, n, epspart=0) = Dual(z, epspart)^n
+@test_throws DomainError powwrap(0, -1)
+@test_throws DomainError powwrap(2, -1)
+@test_throws DomainError powwrap(123, -1) # etc
+# these ones don't DomainError
+@test powwrap(0, 0, 0) == Dual(1, 0) # special case is handled
+@test powwrap(0, 0, 1) == Dual(1, 0) # special case is handled
+@test powwrap(1, -1) == powwrap(1.0, -1) # special case is handled
+@test powwrap(1, -2) == powwrap(1.0, -2) # special case is handled
+@test powwrap(1, -123) == powwrap(1.0, -123) # special case is handled
+@test powwrap(1, 0) == Dual(1, 1)
+@test powwrap(123, 0) == Dual(1, 1)
+for i ∈ -3:3
+  @test powwrap(1, i) == Dual(1, i)
+end
+
+# this no longer throws 1/0 DomainError
+@test powwrap(0, Dual(0, 1)) == Dual(1, 0)
+# this never did DomainError because it starts off with a float
+@test 0.0^Dual(0, 1) == Dual(1.0, NaN)
+# and Dual^Dual uses a log and is now type stable
+# because the log promotes ints to floats for all values
+@test typeof(value(powwrap(0, Dual(0, 1)))) == Float64
+@test Dual(0, 1)^Dual(0, 1) == Dual(1, 0)
+
 y = Dual(2.0, 1)^UInt64(0)
 @test !isnan(epsilon(y))
+@test epsilon(y) == 0
 
 y = sin(x)+exp(x)
 @test value(y) ≈ sin(2)+exp(2)