feldtruby 0.4.1 → 0.4.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 0aea61457b6e6c7b8258675bed056af881761d67
-  data.tar.gz: b726de556b3d1c6e5471f88433bf0117847400e8
+  metadata.gz: 3a76753a45218658033069292166e1f2a6632958
+  data.tar.gz: f7885cd0eece07094aca1d27da9da38e9ddbf51f
 SHA512:
-  metadata.gz: dd59de72274a8ad60e97f0753e13449b77c1f4da793a78a20b16103e36f01132916f1f4616638a5430868877e14fd7ae4b87d8f6fc33df1072c85aab9fd273fa
-  data.tar.gz: 67cb9c8b514e6d4139842fb0750854687a952957f1a8650c33173fc068f93d189ffcb23821d3bed0f3cf0704eb29b570c6e4ac5d781ed9a4445a17cd463d5e58
+  metadata.gz: 96eef547a76262a695fc9480abf6b672b5536881953754ee5b8461de00fa08df05718c5b3e3e40894d4c4d65f527802cec3f11e695a4ca3e9d5b55ba24b2d9b8
+  data.tar.gz: db581f0014c8252980e263db25eacc6804723022d8b08130678a2cdf0f1711baf1118fbb26e710101e9225b2ff6ea00fb526be987017e2539d6f403acee8d95a

data/lib/feldtruby/optimize/objective.rb

@@ -41,7 +41,7 @@ class Objective
 
   attr_reader :global_min_values_per_goal, :global_max_values_per_goal
 
-  def initialize(qualityAggregator = WeightedSumAggregator.new, 
+  def initialize(qualityAggregator = MeanWeigthedGlobalRatios.new, #WeightedSumAggregator.new, 
     comparator = LowerAggregateQualityIsBetterComparator.new)
 
     # A quality aggregator maps the goal values of a candidate to a single number.
@@ -360,12 +360,13 @@ end
 #  P. J. Bentley and J. P. Wakefield, "Finding Acceptable Solutions in the 
 #  Pareto-Optimal Range using Multiobjective Genetic Algorithms", 1997
 #  http://eprints.hud.ac.uk/4052/1/PB_%26_JPW_1997_Finding_Acceptable_Solutions.htm
-# with the difference that that lower values indicate better quality.
+# with the difference that lower values indicate better quality and we use
+# mean instead of sum, and thus call it MWGR.
 # It is the weighted sum of the ratios to the best so far for each goal.
 # One of its benefits is that one need not sort individuals in relation to
 # their peers; the aggregate fitness value is fully determined by the individual
 # and the global min and max values for each objective.
-class Objective::SumOfWeigthedGlobalRatios < Objective::WeightedSumAggregator
+class Objective::MeanWeigthedGlobalRatios < Objective::WeightedSumAggregator
   def ratio(index, value, min, max)
     return 1000.0 if value == nil # We heavily penalize if one sub-quality could not be calculated. Max is otherwise 1.0.
     if objective.is_min_goal?(index)

data/lib/feldtruby/version.rb

@@ -1,3 +1,3 @@
 module FeldtRuby
-  VERSION = "0.4.1"
+  VERSION = "0.4.2"
 end

data/spikes/comparing_samplers_on_classic_optimization_functions/compare_samplers.rb

@@ -32,22 +32,6 @@ SamplerRadiuses = SamplerRadiuses1
 
 NumRepetitionsPerSampler = 5
 
-# Schwefel 2.22 function as stated in the JADE paper:
-#  http://150.214.190.154/EAMHCO/pdf/JADE.pdf
-class MinSchwefel2_22 < MinFunctionOfDimension
-  def objective_min_func(x)
-    t1 = x.inject(0.0) do |sum, xi|
-      sum + xi.abs
-    end
-
-    t2 = x.inject(0.0) do |mult, xi|
-      mult * xi.abs
-    end
-
-    t1 + t2
-  end
-end
-
 # Schwefel 1.2 function as stated in the JADE paper:
 #  http://150.214.190.154/EAMHCO/pdf/JADE.pdf
 class MinSchwefel1_2 < MinFunctionOfDimension

data/test/long_running/test_single_objective_optimization.rb

@@ -20,65 +20,55 @@ module MiniTest::Expectations
 end
 
 def best_from_de_on_objective(objective, dimensions, numSteps = 25_000, verbose = false)
-  objective.dimensions = dimensions
+  objective.dimensions = dimensions if objective.respond_to?(:dimensions=)
   ss = objective.search_space
   de = DEOptimizer.new(objective, ss, {:verbose => verbose, 
     :maxNumSteps => numSteps})
+
+  start_time = Time.now
   best = de.optimize().to_a
+  elapsed = Time.now - start_time
 
   val = objective.calc_func(best)
   val.must_be_close_to objective.minimum
   val.must_be :>, objective.minimum
 
-  return best, objective
+  return best, objective, elapsed
 end
 
 describe "Sphere function" do
   it 'can optimize the Sphere function in 3 dimensions' do
-    best, obj = best_from_de_on_objective MinSphere.new, 3, 12_000
+    best, obj, time = best_from_de_on_objective MinSphere.new, 3, 12_000
+    best.must_be_close_to_one_solution_of obj
+    time.must_be :<, 1.5
+  end
+
+  it 'can optimize the Sphere function in 10 dimensions' do
+    best, obj, time = best_from_de_on_objective MinSphere.new, 10, 60_000
     best.must_be_close_to_one_solution_of obj
+    time.must_be :<, 7.5
   end
 
-#  it 'can optimize the Sphere function in 10 dimensions' do
-#    best, obj = best_from_de_on_objective MinSphere.new, 10, 60_000
-#    best.must_be_close_to_one_solution_of obj
-#  end
-#
-#  it 'can optimize the Sphere function in 30 dimensions' do
-#    best, obj = best_from_de_on_objective MinSphere.new, 30, 220_000
-#    # We don't test closeness since it might take very long for 30D to get close on all dimensions.
-#  end
+  it 'can optimize the Sphere function in 30 dimensions' do
+    best, obj, time = best_from_de_on_objective MinSphere.new, 30, 220_000
+    time.must_be :<, 28.0
+    # We don't test closeness since it might take very long for 30D to get close on all dimensions.
+  end
 end
 
 describe "Levi13 function" do
   it 'can optimize the Levi13 function' do
-    objective = MinLevi13.new
-    ss = objective.search_space
-    de = DEOptimizer.new(objective, ss, {:verbose => false, 
-      :maxNumSteps => 7_500})
-    best = de.optimize().to_a
-
-    val = objective.calc_func(best)
-    val.must_be_close_to objective.minimum
-    val.must_be :>=, objective.minimum
-
-    best.must_be_close_to_one_solution_of objective, 0.01
+    best, obj, time = best_from_de_on_objective MinLevi13.new, nil, 7_500
+    best.must_be_close_to_one_solution_of obj
+    time.must_be :<, 1.0
   end
 end
 
 describe "Beale function" do
   it 'can optimize the Beale function' do
-    objective = MinBeale.new
-    ss = objective.search_space
-    de = DEOptimizer.new(objective, ss, {:verbose => false, 
-      :maxNumSteps => 7_500})
-    best = de.optimize().to_a
-
-    val = objective.calc_func(best)
-    val.must_be_close_to objective.minimum
-    val.must_be :>=, objective.minimum
-
-    best.must_be_close_to_one_solution_of objective, 0.01
+    best, obj, time = best_from_de_on_objective MinBeale.new, nil, 7_500
+    best.must_be_close_to_one_solution_of obj
+    time.must_be :<, 1.0
   end
 end
 
@@ -88,7 +78,7 @@ describe "Easom function" do
     ss = objective.search_space
     # Why can't we do this in 25_000 evals anymore? We did it before. Repeatedly. Very strange.
     de = DEOptimizer.new(objective, ss, {:verbose => false, 
-      :maxNumSteps => 35_000, :printFrequency => 0.0, 
+      :maxNumSteps => 40_000, :printFrequency => 0.0, 
       :samplerRadius => 5})
     best = de.optimize().to_a
 
@@ -104,7 +94,7 @@ describe "EggHolder function" do
   it 'can optimize the Eggholder function' do
     objective = MinEggHolder.new
     ss = objective.search_space
-    de = DEOptimizer.new(objective, ss, {:verbose => true, 
+    de = DEOptimizer.new(objective, ss, {:verbose => false, 
       :maxNumSteps => 25_000, :samplerRadius => 6})
     best = de.optimize().to_a
 
@@ -118,7 +108,8 @@ end
 
 describe "Schwefel 2.22 function" do
   it 'can optimize the Schwefel 2.22 function in 3 dimensions' do
-    best, obj = best_from_de_on_objective MinSchwefel2_22.new, 3, 12_000
+    best, obj, time = best_from_de_on_objective MinSchwefel2_22.new, 3, 12_000
     best.must_be_close_to_one_solution_of obj
+    time.must_be :<, 1.5
   end
 end

data/test/test_optimize_elite_archive.rb

@@ -16,7 +16,7 @@ end
 
 describe "EliteArchive" do
   before do
-    @o = TwoMinOneMax.new
+    @o = TwoMinOneMax.new(FeldtRuby::Optimize::Objective::WeightedSumAggregator.new)
     @a = FeldtRuby::Optimize::EliteArchive.new(@o, {
     :NumTopPerGoal => 2,
     :NumTopAggregate => 3})

data/test/test_optimize_objective.rb

@@ -12,7 +12,7 @@ end
 
 describe "a single minimizing objective" do
   before do
-    @o = SingleObjective1.new
+    @o = SingleObjective1.new(FeldtRuby::Optimize::Objective::WeightedSumAggregator.new)
   end
 
   it "has one goal" do
@@ -162,7 +162,7 @@ end
 
 describe "two sub-objectives" do
   before do
-    @o = TwoMinObjectives1.new
+    @o = TwoMinObjectives1.new(FeldtRuby::Optimize::Objective::WeightedSumAggregator.new)
   end
 
   it "has two aspects/sub-objectives" do
@@ -291,8 +291,8 @@ end
 
 describe "the objective itself and its updates" do
   before do
-    @o = SingleObjective1.new
-    @o2 = TwoMinObjectives1.new
+    @o = SingleObjective1.new(FeldtRuby::Optimize::Objective::WeightedSumAggregator.new)
+    @o2 = TwoMinObjectives1.new(FeldtRuby::Optimize::Objective::WeightedSumAggregator.new)
     @c = [1,2,3]
   end
 
@@ -383,7 +383,7 @@ end
 
 describe "calculating quality when there is one max goal" do
   before do
-    @o = OneMinOneMaxObjective1.new
+    @o = OneMinOneMaxObjective1.new(FeldtRuby::Optimize::Objective::WeightedSumAggregator.new)
   end
 
   it "inverts the max value so that it grows downwards" do
@@ -395,43 +395,86 @@ end
 
 describe "calculating quality with weights" do
   before do
-    @o = OneMinOneMaxObjective1.new
+    @o = OneMinOneMaxObjective1.new(FeldtRuby::Optimize::Objective::WeightedSumAggregator.new)
     @o.weights = {:objective_min_distance_between => 2, :objective_max_sum => 3}
   end
 
-  it "uses the weights when calculating quality" do
+  it "uses the weights when calculating quality, even after weights change" do
     i1 = [1,2,3]
-    @o.weights = {:objective_min_distance_between => 2, :objective_max_sum => 3}
     q1 = @o.quality_of(i1)
-    q1.value.must_equal( 2*((2-1) + (3-2)) + (-3)*(1+2+3) )
     q1.sub_qualities.must_equal [2.0, 6.0]
+    q1.value.must_equal( 2*((2-1) + (3-2)) + (-3)*(1+2+3) )
     q1.candidate.must_equal i1
     q1.objective.must_equal @o
 
     @o.weights = {:objective_min_distance_between => 5, :objective_max_sum => 30}
     q2 = @o.quality_of(i1)
-    q2.value.must_equal( 5*((2-1) + (3-2)) + (-30)*(1+2+3) )
     q2.sub_qualities.must_equal [2.0, 6.0]
+    q2.value.must_equal( 5*((2-1) + (3-2)) + (-30)*(1+2+3) )
+
+    i2 = [1,2,5]
+    q3 = @o.quality_of i2
+    q3.sub_qualities.must_equal [4.0, 8.0]
+    q3.value.must_equal( 5*4.0 + (-30)*8.0 )
+
+    @o.weights = {:objective_min_distance_between => -10, :objective_max_sum => 1}
+    q3.value.must_equal( (-10)*4.0 + (-1)*8.0 )
+    q2.value.must_equal( (-10)*2.0 + (-1)*6.0 )
+    q1.value.must_equal( (-10)*2.0 + (-1)*6.0 )
+
+    @o.weights = {:objective_min_distance_between => -4, :objective_max_sum => -2}
+    q3.value.must_equal( (-4)*4.0 + (2)*8.0 ) # 0.0
+    q2.value.must_equal( (-4)*2.0 + (2)*6.0 ) # 4.0
+    q1.value.must_equal( (-4)*2.0 + (2)*6.0 )
+
+    i3 = [1,2,3,4,5,6]
+    q4 = @o.quality_of i3
+    q4.sub_qualities.must_equal [5.0, 21.0]
+    q4.value.must_equal( (-4)*5.0 + (2)*21.0 ) # 22.0
+
+    @o.rank_candidates([i1, i2, i3]).must_equal [i2, i1, i3]
+
+    @o.weights = {:objective_min_distance_between => 1, :objective_max_sum => 100}
+    @o.rank_candidates([i1, i2, i3]).must_equal [i3, i2, i1]
+
+    @o.weights = {:objective_min_distance_between => 100, :objective_max_sum => 1}
+    @o.rank_candidates([i1, i2, i3]).must_equal [i1, i2, i3]
   end
 end
 
 describe "Using MWGR for range-independent aggregate fitness calc" do
   before do
-    @qa = FeldtRuby::Optimize::Objective::SumOfWeigthedGlobalRatios.new
+    @qa = FeldtRuby::Optimize::Objective::MeanWeigthedGlobalRatios.new
     @o = OneMinOneMaxObjective1.new(@qa)
   end
 
   it 'works for a simple scenario' do
-    q1 = @o.quality_of([1,2]) # [1, 3] => 0.0
+    i1 = [1,2]
+    q1 = @o.quality_of(i1) # [1, 3] => 0.0
     q1.value.must_equal 0.0 # First eval must give perfect score since scales are tight...
 
-    q2 = @o.quality_of([1,3]) # [2, 4] => 0.5
+    i2 = [1,3]
+    q2 = @o.quality_of(i2) # [2, 4] => 0.5
     q2.value.must_equal 0.5 # Perfect on one (max sum) and worst on other (min distance) so (0+1.0)/2
     q1.value.must_equal 0.5 # Perfect on one (min distance) and worst on other (max sum) so (1.0+0.0)/2
 
-    q3 = @o.quality_of([1,4]) # [3, 5] => (0+1.0)/2
+    i3 = [1,4]
+    q3 = @o.quality_of(i3) # [3, 5] => (0+1.0)/2
     q3.value.must_equal 0.5 # Perfect on one (max sum) and worst on other (min distance) so (0+1.0)/2
     q2.value.must_equal( ((2.0-1.0)/(3.0-1.0) + ((5.0-4.0)/(5.0-3.0)))/2.0 )
     q1.value.must_equal( ((1.0-1.0)/(3.0-1.0) + ((5.0-3.0)/(5.0-3.0)))/2.0 )
+
+    i4 = [1,2,3]
+    q4 = @o.quality_of(i4) # [2, 6]
+    q4.value.must_equal( ((2.0-1.0)/(3.0-1.0) + ((6.0-6.0)/(6.0-3.0)))/2.0 )
+    q3.value.must_equal( ((3.0-1.0)/(3.0-1.0) + ((6.0-5.0)/(6.0-3.0)))/2.0 )
+    q2.value.must_equal( ((2.0-1.0)/(3.0-1.0) + ((6.0-4.0)/(6.0-3.0)))/2.0 )
+    q1.value.must_equal( ((1.0-1.0)/(3.0-1.0) + ((6.0-3.0)/(6.0-3.0)))/2.0 )
+
+    @o.rank_candidates([i1,i3,i4]).must_equal [i4,i1,i3]
+    @o.rank_candidates([i3,i4,i2]).must_equal [i4,i2,i3]
+    @o.rank_candidates([i3,i4]).must_equal [i4,i3]
+    @o.rank_candidates([i3,i1]).must_equal [i1,i3]
+    @o.rank_candidates([i2,i3]).must_equal [i2,i3]
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: feldtruby
 version: !ruby/object:Gem::Version
-  version: 0.4.1
+  version: 0.4.2
 platform: ruby
 authors:
 - Robert Feldt
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-04-06 00:00:00.000000000 Z
+date: 2013-04-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rinruby