charlie 0.6.0 → 0.7.0

data/History.txt

@@ -1,3 +1,17 @@
+== 0.7.0 / 2008-01-11
+* Added 1.9fixes.rb which removes all calls to super to avoid bug #16493. Now 100% 1.9-compatible.
+* N-point crossovers for lists.
+* Improved/fixed minor bug in PCross, PMutate.
+* Added PCrossN, PMutateN
+* Option check_every added for evolve_until.
+* Moved benchmark methods to GABenchmark, cleaned up code, added support for multiple stats.
+* Save & output overall best solution in benchmark. Also output version information.
+* Renamed PermutationMutator->TranspositionMutator
+* Added InsertionMutator, RotationMutator
+* Cleaned up examples, merged some, removed duplicates, added readme.
+* Changed order of unary/binary in TreeGenotype, removed defaults.
+* Extended tree mutator options.
+
 == 0.6.0 / 2007-12-26
 * Some fixes for some changes in Ruby 1.9
 * Edge recombination crossover added.

data/Manifest.txt

@@ -9,35 +9,26 @@ data/GENOTYPE
 data/MUTATION
 data/SELECTION
 data/template.html
-examples/bit.rb
-examples/function_opt_2peak.rb
-examples/function_opt_sombero.rb
-examples/gladiatorial_simple.rb
-examples/gladiatorial_sunburn.rb
-examples/gridwalk.rb
-examples/money.rb
-examples/output/flattened_sombero.html
-examples/output/flattened_sombero2_.html
-examples/output/fopt1_dblopt.html
-examples/output/hill10.html
-examples/output/hill2.csv
-examples/output/hill2.html
-examples/output/royalroad1_report.html
-examples/output/royalroad2_report.html
-examples/output/royalroadquick_report.html
+examples/EXAMPLES_README.txt
+examples/bitstring.rb
+examples/coevolution.rb
+examples/function_optimization.rb
+examples/output/bitstring_royalroad.html
+examples/output/function_optimization_sombrero.html
+examples/output/function_optimization_twopeak.csv
+examples/output/function_optimization_twopeak.html
+examples/output/string_weasel.html
 examples/output/tsp.html
-examples/output/weasel1_report.html
-examples/output/weasel2_report.html
-examples/royalroad.rb
-examples/royalroad2.rb
-examples/simple_climb_hill2.rb
+examples/permutation.rb
+examples/string.rb
 examples/tree.rb
 examples/tsp.rb
-examples/weasel.rb
 lib/charlie.rb
+lib/charlie/1.9fixes.rb
 lib/charlie/crossover.rb
 lib/charlie/etc/minireport.rb
 lib/charlie/etc/monkey.rb
+lib/charlie/gabenchmark.rb
 lib/charlie/genotype.rb
 lib/charlie/list/list_crossover.rb
 lib/charlie/list/list_genotype.rb

data/README.txt

@@ -1,4 +1,4 @@
-== Charlie version 0.6.0
+== Charlie version 0.7.0
 * http://rubyforge.org/projects/charlie/
 * http://charlie.rubyforge.org
 * mailto:sander.land+ruby@gmail.com
@@ -45,7 +45,7 @@ This example solves a TSP problem (also quiz #142):
  Population.new(TSP,20).evolve_on_console(50)
 
 This example finds a polynomial which approximates cos(x)
- class Cos < TreeGenotype([proc{3*rand-1.5},:x], [:+,:*,:-], [:-@])
+ class Cos < TreeGenotype([proc{3*rand-1.5},:x], [:-@], [:+,:*,:-])
    def fitness
     -[0,0.33,0.66,1].map{|x| (eval_genes(:x=>x) - Math.cos(x)).abs }.max
    end
@@ -58,7 +58,7 @@ This example finds a polynomial which approximates cos(x)
 
 (The MIT License)
 
-Copyright (c) 2007 Sander Land
+Copyright (c) 2007, 2008 Sander Land
 
 Permission is hereby granted, free of charge, to any person obtaining
 a copy of this software and associated documentation files (the

data/Rakefile

@@ -14,7 +14,7 @@ h = Hoe.new('charlie', Charlie::VERSION) do |p|
   p.email = 'sander.land+ruby@gmail.com'
 
   p.summary = 'A genetic algorithms library for Ruby.'
-  p.description =  [1,2,6,7].map{|n| p.paragraphs_of('README.txt', n)}.join("\n\n")
+  p.description =  [1,2,3,6,7].map{|n| p.paragraphs_of('README.txt', n)}.join("\n\n")
   
   p.changes = p.paragraphs_of('History.txt', 0..1).join("\n\n")
 

data/TODO.txt

@@ -1,30 +1,20 @@
 This is my todo list. Contributions and suggestions don't have to be restricted to these things in any way.
 
 === Small stuff
-* clean up PCross, PMutate
-* n-point crossover for lists
-* external elitism (or whatever this is called). apply elitism but do not allow the elite to reproduce.
-* move benchmark method(s) to separate class(?)
+* Arithmetic crossovers, Brood crossover? 
 === Bigger stuff
 * More builtin genotypes:
-   1. Tree-based, extending into genetic programming. (partially done, mutation options to be extended)
-   2. Graphs / adjacency matrix. -- or multidimensional arrays in general?
-   3. Neural networks, Perceptron
-* Method for population to do several shorter runs until convergence.
-* Method for population to do several shorter runs, save some the best in each to generate a new population, and evolve this. is this useful?
-* More selection algorithms.  Multiple populations w migration? niching?
+  * Tree-based, extending into genetic programming. mostly done. add option for ADFs?
+  * Graphs / adjacency matrix. Multidimensional arrays in general?
+  * Neural networks.
+* Method for population to do several shorter runs until convergence (cf. gridwalk example).
+* Method for population to do several shorter runs, save some the best in each to generate a new population, and evolve this. Is this useful?
+* More selection algorithms.  Multiple populations with migration? Niching?  gladiatorial/tournaments (+Prisoner's dilemma example)?
 * Extend benchmarking options:
-   1. Extend dsl to have options for manual test cases given by sel/cross/mut/etc options, or even free code blocks.
-   2. Option to compare several generation/population size settings. Especially both of them together while keeping the number of fitness evaluations roughly constant.
-   3. Detailed stats for use in comparing a small number of strategies? Including population stats for every generation and run, graphs?
-
-
-
-
-
-
-
-
+  * Extend dsl to have options for manual test cases given by sel/cross/mut/etc options, or even free code blocks.
+  * Option to compare several generation/population size settings. Especially both of them together while keeping the number of fitness evaluations roughly constant.
+  * Instead of (2), constraint on time, or # of fitness evals instead of generations. (?)
+  * Detailed stats for use in comparing a small number of strategies? Including population stats for every generation and run, graphs?
 
 
 

data/data/BENCHMARK

@@ -1,53 +1,55 @@
 == Benchmark documentation
-Population#benchmark can compare several selection, crossover and mutation methods and give a report comparing their
+GABenchmark.benchmark can compare several selection, crossover and mutation methods and give a report comparing their
 performance, both for convergence and speed.
 
-The call to the function is:
+An example for a call to the function is:
 
- Population.benchmark(genotype_class,html_output_file,csv_output_file=nil) {
+ GABenchmark.benchmark(YourGenotypeClass,'html_output.html','csv_output.csv') {
    selection RandomSelection, TournamentSelection(3)
    crossover SinglePointCrossover
-   mutator   ListMutator(
+   mutator   ListMutator(:single_point,:uniform[0.25]), ListMutator(:expected_n[2],:gaussian[0.2])
+   repeat 5
+   population_size 27
  }
 
-* +genotype_class+ is simply your genotype class.
-* +html_output_file+ is an output file where a large report with several tables of statistics will be written. Pass nil for no output file of this type.
-* +csv_output_file+ is an output file where the raw data (the maximum fitness value for each run) will be written. Pass nil or omit parameter for no output file of this type.
-
+* YourGenotypeClass is the class which has Genotype as an ancestor and which has fitness function.
+* <tt>html_output.html</tt> is an output file where a large report with several tables of statistics will be written. Pass nil for no output file of this type.
+* <tt>csv_output.csv</tt> is an output file where the raw data (the maximum fitness value for each run) will be written. Pass nil or omit parameter for no output file of this type.
 
 
 In the block you can call several functions to specify the parameters of the benchmark.
 This is sometimes refered to as a DSL. As inaccurate as that term may be, I'll still use it here.
 
 
-=== Benchmark DSL documentation  (also see StrategiesDSL)
-In the block you must call the following functions:
+=== Benchmark DSL documentation  (also see GABenchmark::StrategiesDSL)
+In the block you can call the following functions:
 
-==== selection(*s_args)
+==== <tt>selection(*s_args)</tt>
 Will cause each of the selection modules in +s_args+ to be tested.
-==== crossover(*c_args)
+==== <tt>crossover(*c_args)</tt>
 Will cause each of the crossover modules in +c_args+ to be tested.
-==== mutator(*m_args)
+==== <tt>mutator(*m_args)</tt> or <tt>mutation(*m_args)</tt>
 Will cause each of the mutation modules in +m_args+ to be tested.
 
-
-If you want to test the module that is already included without repeating it, just pass Module.new as one of the arguments (or Module.new{self.name='default'} for something more descriptive).
-
-
-You can also call the following functions to change some other settings:
-
-==== self.generations     =  g
+==== <tt>generations  g</tt>
 Changes the number of generations in each test to +g+. Default is 50.
 
-==== self.population_size =  s
+==== <tt>population_size  s</tt>
 Changes the population size in each test to +g+. Default is 20.
 
-==== self.repeat          =  r
+==== <tt>repeat  r</tt>
 Changes the number of times each test will be repeated to +r+. Default is 10, a higher number here will take longer to 
 complete, but will improve the accuracy of the results.
 
+==== <tt>track_stat{|best| best.fitness }</tt> or <tt>track_stats{|best| [best.fitness, best.some_other_stat] }</tt>
+* Changes the statistic you want to track, the individual with the highest fitness at the end of a run is yielded to the block.
+* Default is the fitness itself.
+* This can be an array, if you want to track multiple stats at once (for example: tracking both training and generalization error).
 
-
+==== Remarks 
+* If you want to test the module that is already included without repeating it, just pass Module.new as one of the arguments (or Module.new{self.name='default'} for something more descriptive). This is also the default, so <tt>GABenchmark.benchmark(klass){}</tt> will simply test the strategies already included.
+* The methods <tt>selection, crossover, mutator, generations, population_size, repeat</tt> are defined using metaprogramming as <tt>attr_dsl</tt>. Passing no arguments to the function will return the current value, passing more will assign the value. This means you  can do things like <tt>selection << TournamentSelection(3)</tt> in the block. 
+* You can also use calls like <tt>self.repeat = 10</tt>, which are aliases.
 
 
 

data/data/CROSSOVER

@@ -29,6 +29,9 @@ Standard single point crossover. Returns two children.
 === UniformCrossover
 Standard uniform crossover.Returns two children.
 
+=== NPointCrossover(n=2)
+N-point crossover. Returns two children.
+
 
 == Specialized Crossovers
 === For PermutationGenotype
@@ -50,5 +53,6 @@ Applies an arbitrary crossover and returns a random child.
 === PCross(p,crossover,othercrossover=NullCrossover)
 Applies an arbitrary crossover with probability p, and another crossover with probability 1-p.
 
-
+=== PCrossN(hash of module=>probability pairs)
+Version of PCross for more than two crossover operators. If sum(probability) < 1, NullCrossover will be used for the remaining probability.
  

data/data/GENOTYPE

@@ -1,5 +1,5 @@
 == Genotype documentation
-The genotype, aka genome, aka chromosomes is a simple representation of a solution to your problem.
+The genotype, aka genome is a simple representation of a solution to your problem.
 This is usually an array of numbers or bits.
 
 === Creating your own.
@@ -23,7 +23,7 @@ They are all generated dynamically by functions, and should be used like:
 
 
 ====   FloatListGenotype(n,range=0..1)   
-Genotype of +n+ floats in the range +range+.
+Genotype of +n+ floats initialized in the range +range+. Mutations can take the floats out of this range.
 
 Includes <tt>ListMutator()</tt> (with default parameters :expected_n[3], :uniform[ 0.25 ]) and <tt>SinglePointCrossover</tt> by default.
 
@@ -41,11 +41,11 @@ Includes <tt>ListMutator(:expected_n[2],:replace[*elements]), SinglePointCrossov
 These genotypes have their own crossover and mutation operators.
 
 ==== PermutationGenotype(n,elements=0...n)  (file)[link:files/lib/charlie/permutation/permutation_rb.html]
-Genotype for permutations. Includes PermutationMutator and PermutationCrossover by default.
+Genotype for permutations. Includes TranspositionMutator and PermutationCrossover by default.
 
-==== TreeGenotype(terminals = [proc{rand},:x], binary_ops = [:+,:*,:-,:/], unary_ops = nil)  (file)[link:files/lib/charlie/tree/tree_rb.html]
-* Genotype for trees. Useful for genetic programming. Includes TreeMutator and TreeCrossover by default.
-* Functions included: eval_genes(terminal_values_hash), size, module GPTreeHelper
+==== TreeGenotype(terminals, unary_ops, binary_ops)  (file)[link:files/lib/charlie/tree/tree_rb.html]
+* Genotype for trees. Useful for genetic programming. Includes PMutate(0.5,TreeReplaceMutator) and PCross(0.7,TreeCrossover) by default.
+* Functions included: eval_genes(terminal_values_hash), size, depth, module GPTreeHelper
 * Check examples/tree.rb for usage.
 
 

data/data/MUTATION

@@ -3,7 +3,7 @@ A mutation operator mutates a single instance in place.
 
 
 
-The mutation operator should be defined as a +mutate!+ method in your genotype class.
+The mutation operator should be defined as a <tt>mutate!</tt> method in your genotype class.
    class Example < Genotype
      def mutate!
        # apply mutation here.
@@ -31,12 +31,22 @@ Both of these parameters use Symbol#[] in the examples. <tt>:replace['a','b']</t
 
 == Specialized Mutators
 === For PermutationGenotype
-PermutationMutator is a transposition mutator for permutations.
-
-InversionMutator is another mutator for permutations. It inverses a part of the genes array.
-
-=== For TreeGenotype
-TreeMutator replaces a randomly chosen subtree with a new, randomly generated, subtree of depth <= 2.
+All of these will probably work on other array/string-based genotypes as well, but this is untested.
+
+* TranspositionMutator is a transposition mutator for permutations. It interchanges two elements and leaves the remaining elements in their original positions.
+* InversionMutator is another mutator for permutations. It inverses a part of the genes array.
+* InsertionMutator takes a random element of the permutation, and inserts it at a random position.
+* RotationMutator takes a random element, and rotates the permutation such that this is the first element, i.e. it effectively does nothing if the permutation represents a cycle.
+
+=== For TreeGenotype  (file)[link:files/lib/charlie/tree/tree_rb.html]
+* TreeReplaceMutator(depth,type) replaces a randomly chosen subtree with a new, randomly generated, subtree.
+* TreePruneMutator replaces a randomly chosen subtree with a randomly generated terminal.
+* TreeRemoveNodeMutator replaces a random node by one of its children.
+* TreeChopMutator replaces the root  by one of its children.
+* TreeInsertNodeMutator replaces a random node x by a new operator node having x as one of its children.
+* TreeTerminalMutator replaces a random terminal by a new one.
+* TreeNumTerminalMutator(mutate=:uniform[ 0.1 ]) mutates a random numeric terminal.
+* TreeEvalMutator(values_hash=Hash{0}) replaces a random terminal by the result of its evaluation.
 
 == Meta-mutators
 These functions take one or more mutator modules and generate a new mutator module.
@@ -44,5 +54,7 @@ These functions take one or more mutator modules and generate a new mutator modu
 === PMutate(p,mutator,othermutator=NullMutator)
 Applies an arbitrary mutator with probability p, and another mutator with probability 1-p.
 
+=== PMutateN(hash of module=>probability pairs)
+Version of PMutate for more than two mutators. If sum(probability) < 1, NullMutator will be used for the remaining probability.
 
  

data/data/SELECTION

@@ -6,7 +6,8 @@ The selection strategy should be defined as a +next_generation+ method in the me
      ...
      class << self
        def next_generation(population)
-         # select parents and yield(parent1,parent2) them to the crossover and mutation operator. Create an array from several of these calls to get the next generation.
+         # select parents and yield(parent1,parent2) them to the crossover and mutation operator. 
+         # Create an array from several of these calls to get the next generation.
        end
      end
      use RouletteSelection # or include some builtin operator

data/data/template.html

@@ -8,6 +8,7 @@
   table{
     background: #ccc;
     font-family: Verdana, Helvetica, sans-serif;
+    font-size: small;
   }
   th {
     background: #666;
@@ -18,7 +19,7 @@
   td {
     color: #333;
     background: #e2e2e2;
-    padding: 2px;
+    padding: 2px 3px;
     text-align: center;
   }
 </style>

data/examples/EXAMPLES_README.txt

@@ -0,0 +1,70 @@
+In this directory are several examples of how to use the library.
+This file serves as an index, which can be used to find an example for a specific function.
+
+====  function_optimization.rb  ====
+- ruby function_optimization.rb hill
+Description: Function optimization example in many dimensions, hill climbing.
+Uses: FloatListGenotype, TournamentSelection, UniformCrossover, ListMutator
+ 
+- ruby function_optimization.rb twopeak
+Description: Function optimization example in 1 dimension, avoiding a local maximum.
+Uses: FloatListGenotype (single float), NullCrossover, GABenchmark
+
+- ruby function_optimization.rb sombrero
+Description: Function optimization example in 2 dimensions, avoiding many maxima.
+Uses: FloatListGenotype, GABenchmark
+
+====  bitstring.rb  ====
+- ruby bitstring.rb 512
+Description: Very simple example that finds the binary representation of 512.
+Uses: BitStringGenotype
+ 
+- ruby bitstring.rb royalroad
+Description: Example for convergence with limited feedback from the fitness function.
+Uses: BitStringGenotype, TruncationSelection, UniformCrossover, ListMutator, cache_fitness, GABenchmark
+
+====  string.rb  ====
+- ruby string.rb weasel
+Description: Simple target string evolver.
+Uses: StringGenotype, BestOnlySelection, UniformCrossover, ListMutator, GABenchmark
+ 
+- ruby string.rb gridwalk
+Description: Example of using strings of symbols other than chars to evolve a route through a grid.
+Uses: StringGenotype, TournamentSelection, evolve_until
+
+==== tsp.rb  ====
+- ruby tsp.rb
+Description: Travelling salesperson problem.
+Uses: PermutationGenotype, PCross, PMutateN, evolve_until, TournamentSelection,
+      EdgeRecombinationCrossover, InversionMutator ,InsertionMutator, GABenchmark
+
+====  permutation.rb  ====
+- ruby permutation.rb
+Description: send+more=money problem with PermutationGenotype.
+Uses: PermutationGenotype
+
+====  coevolution.rb  ====
+- ruby coevolution.rb simple
+Description: Simple co-evolutionary example. Evolves a two-element array to be [large,small]
+Uses: GladiatorialSelection, FloatListGenotype, UniformCrossover, ListMutator
+
+- ruby coevolution.rb sunburn
+Description: More elaborate example of co-evolution. Evolves weapon systems on space ships (like Ashlock (2004), ch.4).
+Uses: GladiatorialSelection, StringGenotype, custom mutation/crossover, UniformCrossover
+
+====  tree.rb  ====
+- ruby tree.rb cos
+Description: Example for TreeGenotype/GP. Evolves a polynomial that approximates cos(x).
+Uses: TreeGenotype, ...
+
+- ruby tree.rb pors [n=31]
+Description: Plus One Recall Store example. Generate a number n using a minimal # of +,1,STO,RCL operations.
+Uses: TreeGenotype, passing proc{}'s to node values at evaluation, custom unary operators.
+
+- ruby tree.rb porsx [n=31]
+Description: Same idea as PORS, but generate a tree that evaluates to n*T where T can only be recalled once.
+
+- ruby tree.rb bloat
+Description: Just grows large trees.
+
+

data/examples/bitstring.rb

@@ -0,0 +1,72 @@
+begin
+  require '../lib/charlie'
+rescue LoadError
+  require 'rubygems'
+  require 'charlie'
+end
+
+if ARGV[0].nil?
+  puts "Several examples for bit string genotypes."
+  puts "Usage: ruby bitstring.rb 512|royalroad"
+
+
+elsif ARGV[0].downcase.gsub(/[^0-9]/,'') == '512'
+
+
+puts "Simple test with a bit string. Find the binary representation of 512."
+
+class Test < BitStringGenotype(10)
+  def fitness
+    -(genes.map(&:to_s).join.to_i(2) - 512).abs # or just to_s.to_i(2)
+  end
+end
+Population.new(Test).evolve_on_console
+
+
+
+elsif ARGV[0].downcase.gsub(/[^a-z]/,'') == 'royalroad'
+
+
+
+puts "The Royal Road problem."
+puts "Find a bit string of 64 ones, while only getting a fitness bonus for groups of 8 ones."
+puts "Useful as a test for convergence in situations where there is a lack of feedback."
+
+
+class RoyalRoad <  BitStringGenotype(64)  # Royal Road problem
+  def fitness
+    1 + genes.enum_slice(8).find_all{|e| e.all?{|x|x==1} }.size # +1 to avoid all fitness 0 for roulette
+  end
+  use TruncationSelection(0.3), UniformCrossover, ListMutator([:expected_n, 4],:flip)
+  cache_fitness
+end
+
+Population.new(RoyalRoad).evolve_on_console(100)
+
+
+
+output_file = 'output/bitstring_royalroad.html';
+
+puts "Running benchmark. Takes about 3 minutes on Ruby 1.9. Output in #{output_file}"
+
+puts "Press [Enter] to continue, or Ctrl-C to abort."
+STDIN.gets
+
+GABenchmark.benchmark(RoyalRoad,output_file){
+  selection TruncationSelection(0.3),
+            Elitism(ScaledRouletteSelection),
+            Elitism(RouletteSelection),
+            BestOnlySelection,
+            RouletteSelection
+
+  crossover NullCrossover, 
+            SinglePointCrossover,
+            UniformCrossover
+
+  mutator   *(1..6).map{|n| ListMutator(:expected_n[n],:flip) }
+  mutator << NullMutator
+  generations  100
+}
+
+
+end