Remove onlyreal macro

Author: devmotion

Project.toml

@@ -4,6 +4,5 @@ uuid = "872c559c-99b0-510c-b3b7-b6c96a88d5cd"
 [deps]
 Libdl = "8f399da3-3557-5675-b5ff-fb832c97cbdb"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
-MacroTools = "1914dd2f-81c6-5fcd-8719-6d5c9610ff09"
 Requires = "ae029012-a4dd-5104-9daa-d747884805df"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

REQUIRE

@@ -1,3 +1,2 @@
 julia 0.7-
 Requires
-MacroTools

src/NNlib.jl

@@ -2,8 +2,6 @@ module NNlib
 
 using Requires, Libdl
 
-using MacroTools: @capture
-
 export σ, sigmoid, relu, leakyrelu, elu, gelu, swish, selu, softplus, softsign, logσ, logsigmoid,
   softmax, logsoftmax, maxpool, meanpool
 

src/activation.jl

@@ -1,27 +1,15 @@
-macro onlyreal(ex)
-  @capture(ex, (f_(x, a__) = body_) | (function f_(x, a__) body_ end)) ||
-    error("expected a function with initial argument `x`")
-
-  errmsg = "Use explicit invocations such as `$(f).(x)` to apply activation functions to tensors!"
-
-  quote
-    Base.@__doc__ $(f)(x::Real, $(a...)) = $body
-    $(f)(x::AbstractArray, $(a...)) = error($errmsg)
-  end |> esc
-end
-
 """
     σ(x) = 1 / (1 + exp(-x))
 
 Classic [sigmoid](https://en.wikipedia.org/wiki/Sigmoid_function) activation
 function.
 """
-@onlyreal σ(x) = one(x) / (one(x) + exp(-x))
+σ(x::Real) = one(x) / (one(x) + exp(-x))
 
 const sigmoid = σ
 
 # ForwardDiff numerical stability hack
-@onlyreal σ_stable(x) = ifelse(x < -80, zero(x), one(x) / (one(x) + exp(-x)))
+σ_stable(x::Real) = ifelse(x < -80, zero(x), one(x) / (one(x) + exp(-x)))
 
 σ(x::Float32) = σ_stable(x)
 
@@ -42,7 +30,7 @@ Return `log(σ(x))` which is computed in a numerically stable way.
       -10.0
        -0.0
 """
-@onlyreal function logσ(x)
+function logσ(x::Real)
   max_v = max(zero(x), -x)
   z = exp(-max_v) + exp(-x-max_v)
   -(max_v + log(z))
@@ -56,7 +44,7 @@ const logsigmoid = logσ
 [Rectified Linear Unit](https://en.wikipedia.org/wiki/Rectifier_(neural_networks))
 activation function.
 """
-@onlyreal relu(x) = max(zero(x), x)
+relu(x::Real) = max(zero(x), x)
 
 
 """
@@ -66,7 +54,7 @@ Leaky [Rectified Linear Unit](https://en.wikipedia.org/wiki/Rectifier_(neural_ne
 activation function.
 You can also specify the coefficient explicitly, e.g. `leakyrelu(x, 0.01)`.
 """
-@onlyreal leakyrelu(x, a = oftype(x/1, 0.01)) = max(a*x, x/1)
+leakyrelu(x::Real, a = oftype(x/1, 0.01)) = max(a*x, x/1)
 
 """
     elu(x, α = 1) =
@@ -76,15 +64,15 @@ Exponential Linear Unit activation function.
 See [Fast and Accurate Deep Network Learning by Exponential Linear Units](https://arxiv.org/abs/1511.07289).
 You can also specify the coefficient explicitly, e.g. `elu(x, 1)`.
 """
-@onlyreal elu(x, α = one(x)) = ifelse(x ≥ 0, x/1, α * (exp(x) - one(x)))
+elu(x, α = one(x)) = ifelse(x ≥ 0, x/1, α * (exp(x) - one(x)))
 
 """
     gelu(x) = 0.5x*(1 + tanh(√(2/π)*(x + 0.044715x^3)))
 
 [Gaussian Error Linear Unit](https://arxiv.org/pdf/1606.08415.pdf)
 activation function.
 """
-@onlyreal function gelu(x)
+function gelu(x::Real)
     λ = oftype(x/1, √(2/π))
     α = oftype(x/1, 0.044715)
     h = oftype(x/1, 0.5)
@@ -98,7 +86,7 @@ end
 Self-gated actvation function.
 See [Swish: a Self-Gated Activation Function](https://arxiv.org/pdf/1710.05941.pdf).
 """
-@onlyreal swish(x) = x * σ(x)
+swish(x::Real) = x * σ(x)
 
 """
     selu(x) = λ * (x ≥ 0 ? x : α * (exp(x) - 1))
@@ -109,7 +97,7 @@ See [Swish: a Self-Gated Activation Function](https://arxiv.org/pdf/1710.05941.p
 Scaled exponential linear units.
 See [Self-Normalizing Neural Networks](https://arxiv.org/pdf/1706.02515.pdf).
 """
-@onlyreal function selu(x)
+function selu(x::Real)
   λ = oftype(x/1, 1.0507009873554804934193349852946)
   α = oftype(x/1, 1.6732632423543772848170429916717)
   λ * ifelse(x > 0, x/1, α * (exp(x) - 1))
@@ -120,12 +108,18 @@ end
 
 See [Quadratic Polynomials Learn Better Image Features](http://www.iro.umontreal.ca/~lisa/publications2/index.php/attachments/single/205).
 """
-@onlyreal softsign(x) = x / (one(x) + abs(x))
+softsign(x::Real) = x / (one(x) + abs(x))
 
 
 """
     softplus(x) = log(exp(x) + 1)
 
 See [Deep Sparse Rectifier Neural Networks](http://proceedings.mlr.press/v15/glorot11a/glorot11a.pdf).
 """
-@onlyreal softplus(x) = log1p(exp(x))
+softplus(x::Real) = log1p(exp(x))
+
+# Provide an informative error message if activation functions are called with an array
+for f in (:σ, :σ_stable, :logσ, :relu, :leakyrelu, :elu, :gelu, :swish, :selu, :softsign, :softplus)
+  @eval $(f)(x::AbstractArray, args...) =
+    error("Use explicit invocations such as `", $(string(f)), ".(x)` to apply activation functions to tensors!")
+end