Merge 271c728 into 3351bb6

src/Optim.jl

@@ -93,22 +93,6 @@ module Optim
     include("golden_section.jl")
     include("brent.jl")
 
-    const methods = Dict{Symbol, Type}(
-        :gradient_descent => GradientDescent,
-        :momentum_gradient_descent => MomentumGradientDescent,
-        :cg => ConjugateGradient,
-        :bfgs => BFGS,
-        :l_bfgs => LBFGS,
-        :newton => Newton,
-        :newton_tr => NewtonTrustRegion,
-        :nelder_mead => NelderMead,
-        :simulated_annealing => SimulatedAnnealing,
-        :particle_swarm => ParticleSwarm,
-        :brent => Brent,
-        :golden_section => GoldenSection,
-        :accelerated_gradient_descent => AcceleratedGradientDescent,
-        :fminbox => Fminbox)
-
     # Backward compatibility
     include("deprecate.jl")
 
@@ -117,6 +101,7 @@ module Optim
 
     # Examples for testing
     include(joinpath("problems", "unconstrained.jl"))
+    include(joinpath("problems", "univariate.jl"))
 
     cgdescent(args...) = error("API has changed. Please use cg.")
 end

src/problems/univariate.jl

@@ -0,0 +1,51 @@
+module UnivariateProblems
+
+    ### Sources
+    ###
+    ### [1]  http://infinity77.net/global_optimization/test_functions_1d.html
+
+    immutable UnivariateProblem
+        name::AbstractString
+        f::Function
+        bounds::Vector{Float64}
+        minimizers::Vector{Float64}
+        minima::Vector{Float64}
+    end
+
+    examples = Dict{AbstractString, UnivariateProblem}()
+
+    f(x) = 2x^2+3x+1
+
+    examples["Polynomial"] = UnivariateProblem("Polynomial",
+                                                  f,
+                                                  [-2.0, 1.0],
+                                                  [-0.75,],
+                                                  [f(-0.75),])
+
+    # Problem 04 from [1]
+    p04(x) = -(16x^2-24x+5)exp(-x)
+
+    examples["Problem04"] = UnivariateProblem("Problem04",
+                                                p04,
+                                                [1.9, 3.9],
+                                                [2.868034,],
+                                                [p04(2.868034),])
+
+    # Problem 13 from [1]
+    p13(x) = -x^(2/3)-(1-x^2)^(1/3)
+
+    examples["Problem13"] = UnivariateProblem("Problem13",
+                                                p13,
+                                                [0.001, 0.99],
+                                                [1./sqrt(2.0),],
+                                                [p13(1./sqrt(2.0)),])
+
+    # Problem 18 from [1]
+    p18(x) = x <= 3.0 ? (x-2.0)^2 : 2.0*log(x-2.0)+1.0
+
+    examples["Problem18"] = UnivariateProblem("Problem18",
+                                                p18,
+                                                [0.0, 6.0],
+                                                [2.0,],
+                                                [p18(2.0),])
+end

test/accelerated_gradient_descent.jl

@@ -1,16 +1,17 @@
-using Optim
+# TODO expand tests here
+let
+    f(x) = x[1]^4
+    function g!(x, storage)
+        storage[1] = 4 * x[1]^3
+        return
+    end
 
-f(x) = x[1]^4
-function g!(x, storage)
-    storage[1] = 4 * x[1]^3
-    return
-end
-
-d = DifferentiableFunction(f, g!)
+    d = DifferentiableFunction(f, g!)
 
-initial_x = [1.0]
+    initial_x = [1.0]
 
-Optim.optimize(d, initial_x,
-               method = AcceleratedGradientDescent()
-               show_trace = true,
-               iterations = 10)
+    Optim.optimize(d, initial_x,
+                   method = AcceleratedGradientDescent()
+                   show_trace = true,
+                   iterations = 10)
+end