Optimizations

.github/workflows/documentation.yml

@@ -6,7 +6,7 @@ on:
   push:
     branches:
       - main
-    tags: '*'
+    tags: "*"
   pull_request:
 
 jobs:
@@ -18,7 +18,7 @@ jobs:
       - uses: actions/checkout@v2
       - uses: julia-actions/setup-julia@latest
         with:
-          version: '1.7'
+          version: "1.7"
       - name: Install dependencies
         run: julia --project=docs/ -e 'using Pkg; Pkg.develop(PackageSpec(path=pwd())); Pkg.instantiate()'
       - name: Build and deploy

src/crossover.jl

@@ -78,10 +78,10 @@ Uniform crossover between parents `a` and `b`. Each gene
 of the chromosome is randomly selected from one of the parents.
 """
 function cross(::UniformRecombinator, a, b; rng=Random.GLOBAL_RNG)
-    child = copy(a)
+    child = similar(a)
 
     for i in eachindex(a)
-        @inbounds child[i] = rand(rng) < 0.5 ? b[i] : continue
+        @inbounds child[i] = rand(rng) < 0.5 ? b[i] : a[i]
     end
 
     return child

src/generators.jl

@@ -146,14 +146,16 @@ function unif_rand_particle_pop(n, lb, ub; rng=Random.GLOBAL_RNG)
 end
 
 """
-    normal_rand_particle_pop(n, μ, Σ; rng=Random.GLOBAL_RNG)
+    normal_rand_particle_pop(n, μ, Σ; y=Inf, rng=Random.GLOBAL_RNG)
 
 Generate a population of `n` [`Particle`](@ref) using a normal distribution with means `μ``
 and covariance `Σ`.
 
 `μ` expects a vector of length _l_ (i.e. number of dimensions) while `Σ` expects an _l x l_
 matrix of covariances.
 
+`y` is the evalutaion and current best value. `y` is set to `Inf` by default.
+
 # Examples
 
 ```julia
@@ -164,10 +166,8 @@ julia> normal_rand_particle_pop(3, [0, 0], [1 0; 0 1])
  Particle([0.5687241357408321, -0.7406267072113427], [0.0, 0.0], Inf, [0.5687241357408321, -0.7406267072113427], Inf)
 ```
 """
-function normal_rand_particle_pop(n, μ, Σ; rng=Random.GLOBAL_RNG)
-    # TODO: Add y as optional parameter but default to Inf?
+function normal_rand_particle_pop(n, μ, Σ; y=Inf, rng=Random.GLOBAL_RNG)
     D = MvNormal(μ, Σ)
-    y = Inf
     pop = Vector{Particle}(undef, n)
 
     for i in 1:n

src/logbook.jl

@@ -72,6 +72,8 @@ function compute!(notebooks::Vector{Logbook}, data::AbstractVector)
         record = namedtuple(fnames, callables)
         push!(logger.records, record)
     end
+
+    return nothing
 end
 
 
@@ -87,14 +89,16 @@ function summarise(logger::Logbook)
         data = [logger.records[j][i] for j in 1:n]
         printstyled("\n $(eachstat) \n"; bold=true)
         print("max: $(maximum(data)) \n" *
-            "avg: $(mean(data))\n" *
-            "median: $(median(data)) \n" *
-            "min: $(minimum(data))\n" *
-            "std: $(std(data))\n")
+              "avg: $(mean(data))\n" *
+              "median: $(median(data)) \n" *
+              "min: $(minimum(data))\n" *
+              "std: $(std(data))\n")
         plt = lineplot(data;
-                xlabel="it", ylabel=eachstat)
+            xlabel="it", ylabel=eachstat)
         print(plt)
     end
+
+    return nothing
 end
 
 function summarise(notebooks::Vector{Logbook})
@@ -104,13 +108,15 @@ function summarise(notebooks::Vector{Logbook})
             data = [eachnb.records[j][i] for j in 1:n]
             printstyled("\n $(eachstat) \n"; bold=true)
             print("max: $(maximum(data)) \n" *
-                "avg: $(mean(data))\n" *
-                "median: $(median(data)) \n" *
-                "min: $(minimum(data))\n" *
-                "std: $(std(data))\n")
+                  "avg: $(mean(data))\n" *
+                  "median: $(median(data)) \n" *
+                  "min: $(minimum(data))\n" *
+                  "std: $(std(data))\n")
             plt = lineplot(data;
-                    xlabel="it", ylabel=eachstat)
+                xlabel="it", ylabel=eachstat)
             print(plt)
         end
     end
+
+    return nothing
 end

src/mutation.jl

@@ -48,7 +48,7 @@ end
 Gaussian mutation with standard deviation `σ`, which should be a real number.
 """
 struct GaussianMutator <: ContinuousMutator
-    σ
+    σ::Real
 end
 
 """

src/selection.jl

@@ -19,7 +19,7 @@ abstract type ParentSelector <: Selector end
 Tournament parent selection with tournament size `T`.
 """
 struct TournamentSelector <: ParentSelector
-    T
+    T::Int
 end
 
 """

src/testfunctions.jl

@@ -74,8 +74,8 @@ f(x) = -a \\exp\\left(-b\\sqrt{\\frac{1}{d} \\sum_{i=1}^d x_i^2}\\right)
 """
 @inline function ackley(x::Vector{T} where {T<:Real}; a=20, b=0.2, c=2π)
     d = length(x)
-    return -a * exp(-b * sqrt(sum(x .^ 2) / d)) -
-           exp(sum(cos.(c * xi) for xi in x) / d) + a + exp(1)
+    return @fastmath -a * exp(-b * sqrt(sum(x .^ 2) / d)) -
+                     exp(sum(cos.(c * xi) for xi in x) / d) + a + exp(1)
 end
 
 """
@@ -87,7 +87,7 @@ f(x) = (x_1 + 2x_2 - 7)^2 + (2 x_1 + x_2 - 5)^2
 ```
 """
 @inline function booth(x::Vector{T} where {T<:Real})
-    return (x[1] + 2 * x[2] - 7)^2 + (2 * x[1] + x[2] - 5)^2
+    return @fastmath (x[1] + 2 * x[2] - 7)^2 + (2 * x[1] + x[2] - 5)^2
 end
 
 
@@ -105,8 +105,8 @@ f(x) = a(x_2 - bx_1^2 + cx_1 - r)^2 + s(1 - t)\\cos(x_1) + s
 ```
 """
 @inline function branin(x::Vector{T} where {T<:Real};
-            a=1, b=5.1 / (4π^2), c=5 / π, r=6, s=10, t=1 / (8π))
-    return a * (x[2] - b * x[1]^2 + c * x[1] - r)^2 + s * (1 - t) * cos(x[1]) + s
+    a=1, b=5.1 / (4π^2), c=5 / π, r=6, s=10, t=1 / (8π))
+    return @fastmath a * (x[2] - b * x[1]^2 + c * x[1] - r)^2 + s * (1 - t) * cos(x[1]) + s
 end
 
 
@@ -119,8 +119,8 @@ with optimiser ``\\mathbf{x}^* = (512, 404.231805)``.
 """
 function eggholder(x::Vector{T} where {T<:Real})
     n = length(x)
-    return -sum([(x[i+1]+47) * sin(sqrt(abs(x[i+1] + 47 + x[i]/2))) +
-        x[i] * sin(sqrt(abs(x[i] - (x[i+1] + 47)))) for i in 1:n-1])
+    return -sum([(x[i+1] + 47) * sin(sqrt(abs(x[i+1] + 47 + x[i] / 2))) +
+                 x[i] * sin(sqrt(abs(x[i] - (x[i+1] + 47)))) for i in 1:n-1])
 end
 
 
@@ -136,7 +136,7 @@ f(x) = -\\sum_{i=1}^{d}\\sin(x_i) \\sin^{2m}\\left(\\frac{ix_i^2}{\\pi}\\right)
 ```
 """
 @inline function michalewicz(x::Vector{T} where {T<:Real}; m=10)
-    return -sum(sin(v) * sin(i * v^2 / π)^(2m) for (i, v) in enumerate(x))
+    return @fastmath -sum(sin(v) * sin(i * v^2 / π)^(2m) for (i, v) in enumerate(x))
 end
 
 
@@ -149,9 +149,9 @@ optimiser ``\\mathbf{x}^* = `(-488.632577, 512)`.
 """
 @inline function rana(x::Vector{T} where {T<:Real})
     n = length(x)
-    return sum([x[i] * cos(sqrt(abs(x[i+1] + x[i] + 1))) * sin(sqrt(abs(x[i+1] - x[i] + 1))) +
-        (1 + x[i+1]) * sin(sqrt(abs(x[i+1] + x[i] + 1))) * cos(sqrt(abs(x[i+1] - x[i] + 1)))
-        for i in 1:n-1])
+    return @fastmath sum([x[i] * cos(sqrt(abs(x[i+1] + x[i] + 1))) * sin(sqrt(abs(x[i+1] - x[i] + 1))) +
+                          (1 + x[i+1]) * sin(sqrt(abs(x[i+1] + x[i] + 1))) * cos(sqrt(abs(x[i+1] - x[i] + 1)))
+                          for i in 1:n-1])
 end
 
 
@@ -173,7 +173,7 @@ f(x) = \\sum_{i=1}^{d-1} \\left[b(x_{i+1} - x_i^2)^2 + (x_i - 1)^2 \\right]
 """
 @inline function rosenbrock(x::Vector{T} where {T<:Real}; b=100)
     n = length(x)
-    return sum([b * (x[i+1] - x[i]^2)^2 + (x[i] - 1)^2 for i in 1:n-1])
+    return @fastmath sum([b * (x[i+1] - x[i]^2)^2 + (x[i] - 1)^2 for i in 1:n-1])
 end
 
 
@@ -188,5 +188,5 @@ f(x) = - \\exp(- (x_1 x_2 - a)^2 - (x_2 - a)^2 )
 ```
 """
 @inline function wheeler(x::Vector{T} where {T<:Real}; a=1.5)
-    return -exp(-(x[1] * x[2] - a)^2 - (x[2] - a)^2)
+    return @fastmath -exp(-(x[1] * x[2] - a)^2 - (x[2] - a)^2)
 end