ai4r 1.2 → 1.3

data/README.rdoc

@@ -1,26 +1,12 @@
 = Introduction
 
- AI4R is a collection of ruby algorithms implementations, covering several Artificial intelligence fields, 
- and simple practical examples using them. It implements:
-
-* Genetic algorithms (AI4R::GeneticAlgorithm::GeneticSearch)
-    
-* Neural networks (AI4R::NeuralNetwork::Backpropagation)
-    
-* ID3 Decision Trees (AI4R::Classifiers::ID3)
-
-* PRISM (J. Cendrowska, 1987) (AI4R::Classifiers::Prism)
-	
-* OneR (AKA One Attribute Rule, 1R) (AI4R::Classifiers::OneR)
-    
-* ZeroR (AI4R::Classifiers::ZeroR)
+This project aims to produce ruby implementations of 
+algorithms covering several Artificial intelligence fields.
 
 = Where can I find the lastest code and info on this project?
 
 http://ai4r.rubyforge.org
 
-http://ai4r.jadeferret.com
-
 = How to install
 
 1. Install the gem:
@@ -29,18 +15,19 @@ http://ai4r.jadeferret.com
 
 2. Include require statements in your code:
 
-	require "rubygems"
-	require "ai4r/classifiers/id3"en 
-	require "ai4r/classifiers/prism"
-	require "ai4r/classifiers/one_r"
-	require "ai4r/classifiers/zero_r"
-	require "ai4r/neural_network/backpropagation"
-	require "ai4r/genetic_algorithm/genetic_algorithm"
+  require "rubygems"
+  require "ai4r/classifiers/id3"en 
+  require "ai4r/classifiers/prism"
+  require "ai4r/classifiers/one_r"
+  require "ai4r/classifiers/zero_r"
+  require "ai4r/neural_network/backpropagation"
+  require "ai4r/genetic_algorithm/genetic_algorithm"
 
 = Feedback
 
-If you have some constructive comments about this project, please do send those 
-to me: Sergio Fierens (sergio(dot)fierens(at)gmail(dot)com) 
+If you have questions or constructive comments about this project, 
+please post them in the forum. If you do not want to make it public, 
+send it to me: Sergio Fierens (sergio(dot)fierens(at)gmail(dot)com) 
 
 = Warranty
 

data/examples/decision_trees/id3_example.rb

@@ -7,25 +7,22 @@
 # the Mozilla Public License version 1.1  as published by the 
 # Mozilla Foundation at http://www.mozilla.org/MPL/MPL-1.1.txt
 
-#require File.dirname(__FILE__) + '/../../lib/decision_tree/id3'
 require File.dirname(__FILE__) + '/../../lib/ai4r/classifiers/id3'
-require 'csv'
 
 # Load data from data_set.csv
-data_set = []
-CSV::Reader.parse(File.open("#{File.dirname(__FILE__)}/data_set.csv", 'r')) do |row|
-  data_set << row
-end
-data_labels = data_set.shift
+data_filename = "#{File.dirname(__FILE__)}/data_set.csv"
+data_set = Ai4r::Data::DataSet.new.load_data_and_labels_from_csv data_filename
 
 # Build ID3 tree
-id3 = Ai4r::Classifiers::ID3.new.build(data_set, data_labels)
+id3 = Ai4r::Classifiers::ID3.new.build(data_set)
 
 # Show rules
 puts "Discovered rules are:"
-puts id3.to_s
+puts id3.get_rules
+puts 
 
 # Try to predict some values
+puts "Prediction samples:"
 puts "['Moron Sur (GBA)','4','[86 m2 - 100 m2]'] => " + id3.eval(['Moron Sur (GBA)','4','[86 m2 - 100 m2]'])
 puts "['Moron Sur (GBA)','3','[101 m2 - 125 m2]'] => " + id3.eval(['Moron Sur (GBA)','3','[101 m2 - 125 m2]'])
 puts "['Recoleta (CABA)','3','[86 m2 - 100 m2]'] => " + id3.eval(['Recoleta (CABA)','3','[86 m2 - 100 m2]',])

data/examples/decision_trees/results.txt

@@ -23,6 +23,8 @@ elsif size=='[126 m2 - 160 m2]' and zone=='Moron Sur (GBA)' then price='[56K-75K
 elsif size=='[126 m2 - 160 m2]' and zone=='Recoleta (CABA)' then price='[200K-275K]'
 elsif size=='[126 m2 - 160 m2]' and zone=='Tigre (GBA)' then price='>275K'
 else raise 'There was not enough information during training to do a proper induction for this data element' end
+
+Prediction samples:
 ['Moron Sur (GBA)','4','[86 m2 - 100 m2]'] => [46K-55K]
 ['Moron Sur (GBA)','3','[101 m2 - 125 m2]'] => [76K-100K]
 ['Recoleta (CABA)','3','[86 m2 - 100 m2]'] => [126K-150K]

data/examples/genetic_algorithm/genetic_algorithm_example.rb

@@ -8,32 +8,30 @@
 # Mozilla Foundation at http://www.mozilla.org/MPL/MPL-1.1.txt
 
 require File.dirname(__FILE__) + '/../../lib/ai4r/genetic_algorithm/genetic_algorithm'
+require File.dirname(__FILE__) + '/../../lib/ai4r/data/data_set'
 require 'csv'
 
 # Load data from data_set.csv
-data_set = []
-CSV::Reader.parse(File.open("#{File.dirname(__FILE__)}/travel_cost.csv", 'r')) do |row|
-  data_set << row
-end
-data_labels = data_set.shift
-data_set.collect! do |column|
-  column.collect { |element| element.to_f}
-end
+data_filename = "#{File.dirname(__FILE__)}/travel_cost.csv"
+data_set = Ai4r::Data::DataSet.new.load_data_and_labels_from_csv data_filename
+data_set.data_items.collect! {|column| column.collect {|element| element.to_f}}
 
 Ai4r::GeneticAlgorithm::Chromosome.set_cost_matrix(data_set)
 
 puts "Some random selected tours costs: "
 3.times do
   c = Ai4r::GeneticAlgorithm::Chromosome.seed
-  puts "COST #{-1 * c.fitness} TOUR: #{c.data.collect{ |c| data_labels[c]} * ', '}"
+  puts "COST #{-1 * c.fitness} TOUR: "+
+   "#{c.data.collect{|c| data_set.data_labels[c]} * ', '}"
 end
 
 puts "Beginning genetic search, please wait... "
 search = Ai4r::GeneticAlgorithm::GeneticSearch.new(800, 100)
 result = search.run
-puts "BEST COST FOUND #{-1 * result.fitness} TOUR: #{result.data.collect{ |c| data_labels[c]} * ', '}"
+puts "COST #{-1 * result.fitness} TOUR: "+
+  "#{result.data.collect{|c| data_set.data_labels[c]} * ', '}"
 
- # $7611.99 TOUR: Moscow, Kiev, Warsaw, Hamburg, Berlin, Vienna, Munich, Milan, Rome, Barcelona, Madrid, Paris, Brussels, London, Dublin
- # $7659.81 TOUR: Moscow, Kiev, Warsaw, Vienna, Munich, Berlin, Hamburg, Brussels, Dublin, London, Paris, Milan, Rome, Barcelona, Madrid
- # $7596.74 TOUR: Moscow, Kiev, Warsaw, Berlin, Hamburg, Vienna, Munich, Milan, Rome, Barcelona, Madrid, Paris, Brussels, London Dublin
+# $7611.99 TOUR: Moscow, Kiev, Warsaw, Hamburg, Berlin, Vienna, Munich, Milan, Rome, Barcelona, Madrid, Paris, Brussels, London, Dublin
+# $7659.81 TOUR: Moscow, Kiev, Warsaw, Vienna, Munich, Berlin, Hamburg, Brussels, Dublin, London, Paris, Milan, Rome, Barcelona, Madrid
+# $7596.74 TOUR: Moscow, Kiev, Warsaw, Berlin, Hamburg, Vienna, Munich, Milan, Rome, Barcelona, Madrid, Paris, Brussels, London Dublin
  

data/examples/neural_network/xor_example.rb

@@ -0,0 +1,25 @@
+require File.dirname(__FILE__) + '/training_patterns'
+require File.dirname(__FILE__) + '/patterns_with_noise'
+require File.dirname(__FILE__) + '/patterns_with_base_noise'
+require File.dirname(__FILE__) + '/../../lib/ai4r/neural_network/backpropagation'
+
+examples = [
+    [[0, 0], [0, 1]],
+    [[0, 1], [1, 0]],
+    [[1, 0], [1, 0]],
+    [[1, 1], [0, 0]]
+]
+
+net = Ai4r::NeuralNetwork::Backpropagation.new([2, 1, 2, 1])
+
+i=0
+200.times {
+    examples.each do |ex|
+        2000.times {net.train(ex[0], [ex[1].first])}
+    end
+    puts(i=i+1)
+}
+
+examples.each do |ex|
+    print ex[0], ' => ', net.eval(ex[0]).inspect, ', should be ', ex[1].first, "\n"
+end       

data/lib/ai4r.rb

@@ -0,0 +1,10 @@
+require "ai4r/clusterers/clusterer"
+require "ai4r/clusterers/k_means"
+require "ai4r/clusterers/bisecting_k_means"
+require "ai4r/classifiers/classifier"
+require "ai4r/classifiers/id3"
+require "ai4r/classifiers/prism"
+require "ai4r/classifiers/one_r"
+require "ai4r/classifiers/zero_r"
+require "ai4r/neural_network/backpropagation"
+require "ai4r/genetic_algorithm/genetic_algorithm"

data/lib/ai4r/classifiers/classifier.rb

@@ -0,0 +1,46 @@
+# Author::    Sergio Fierens
+# License::   MPL 1.1
+# Project::   ai4r
+# Url::       http://ai4r.rubyforge.org/
+#
+# You can redistribute it and/or modify it under the terms of 
+# the Mozilla Public License version 1.1  as published by the 
+# Mozilla Foundation at http://www.mozilla.org/MPL/MPL-1.1.txt
+ 
+module Ai4r
+  module Classifiers
+    
+    # The only purpose of this class is to define a common API for classifiers.
+    # All methods in this class must be implemented in subclasses.
+    class Classifier
+
+      # Build a new classifier, using data examples found in data_set.
+      def build(data_set)
+        raise NotImplementedError
+      end
+      
+      # You can evaluate new data, predicting its class.
+      # e.g.
+      #   classifier.eval(['New York',  '<30', 'F'])  # => 'Y'
+      def eval(data)
+        raise NotImplementedError
+      end
+      
+      # This method returns the generated rules in ruby code.
+      # e.g.
+      #   
+      #   classifier.get_rules
+      #     # =>  marketing_target='Y'
+      #
+      # It is a nice way to inspect induction results, and also to execute them:  
+      #     marketing_target = nil
+      #     eval classifier.get_rules   
+      #     puts marketing_target
+      #       # =>  'Y'
+      def get_rules
+        raise NotImplementedError
+      end
+      
+    end
+  end
+end

data/lib/ai4r/classifiers/id3.rb

@@ -8,7 +8,8 @@
 # the Mozilla Public License version 1.1  as published by the 
 # Mozilla Foundation at http://www.mozilla.org/MPL/MPL-1.1.txt
 
-require File.dirname(__FILE__) + '/classifier_helper'
+require File.dirname(__FILE__) + '/../data/data_set'
+require File.dirname(__FILE__) + '/../classifiers/classifier'
 
 module Ai4r
   
@@ -44,9 +45,9 @@ module Ai4r
     #            ['Chicago',     '>80',      'F', 'Y']
     #          ]
     #   
-    #   id3 = DecisionTree::ID3.new(DATA_SET, DATA_LABELS)
+    #   id3 = Ai4r::Classifiers::ID3.new(DATA_SET, DATA_LABELS)
     #   
-    #   id3.to_s
+    #   id3.get_rules
     #     # =>  if age_range=='<30' then marketing_target='Y'
     #           elsif age_range=='[30-50)' and city=='Chicago' then marketing_target='Y'
     #           elsif age_range=='[30-50)' and city=='New York' then marketing_target='N'
@@ -69,14 +70,14 @@ module Ai4r
     #   end
     #   data_labels = data_set.shift
     #   
-    #   id3 = DecisionTree::ID3.new(data_set, data_labels)      
+    #   id3 = Ai4r::Classifiers::ID3.new(data_set, data_labels)      
     #   
     # = A nice tip for data evaluation
     # 
-    #   id3 = DecisionTree::ID3.new(DATA_SET, DATA_LABELS)
+    #   id3 = Ai4r::Classifiers::ID3.new(DATA_SET, DATA_LABELS)
     #     age_range = '<30'
     #     marketing_target = nil
-    #     eval id3.to_s   
+    #     eval id3.get_rules   
     #     puts marketing_target
     #       # =>  'Y'  
     # = More about ID3 and decision trees
@@ -87,50 +88,17 @@ module Ai4r
     # = About the project
     # Author::    Sergio Fierens
     # License::   MPL 1.1
-
-    class ID3
+    # Url::       http://ai4r.rubyforge.org/
+    class ID3 < Classifier
       
-      attr_reader :data_labels
-      include ClassifierHelper
+      attr_reader :data_set
       
-      # Create a new decision tree. If your data is classified with N attributed
-      # and M examples, then your data examples must have the following format:
-      # 
-      #     [   [ATT1_VAL1, ATT2_VAL1, ATT3_VAL1, ... , ATTN_VAL1,  CATEGORY_VAL1], 
-      #         [ATT1_VAL2, ATT2_VAL2, ATT3_VAL2, ... , ATTN_VAL2,  CATEGORY_VAL2], 
-      #         ...
-      #         [ATTM1_VALM, ATT2_VALM, ATT3_VALM, ... , ATTN_VALM,  CATEGORY_VALM], 
-      #     ]
-      #     
-      # e.g.
-      #     [   ['New York',  '<30',      'M', 'Y'],
-      #          ['Chicago',     '<30',      'M', 'Y'],
-      #          ['Chicago',     '<30',      'F', 'Y'],
-      #          ['New York',  '<30',      'M', 'Y'],
-      #          ['New York',  '<30',      'M', 'Y'],
-      #          ['Chicago',     '[30-50)',  'M', 'Y'],
-      #          ['New York',  '[30-50)',  'F', 'N'],
-      #          ['Chicago',     '[30-50)',  'F', 'Y'],
-      #          ['New York',  '[30-50)',  'F', 'N'],
-      #          ['Chicago',     '[50-80]', 'M', 'N'],
-      #          ['New York',  '[50-80]', 'F', 'N'],
-      #          ['New York',  '[50-80]', 'M', 'N'],
-      #          ['Chicago',     '[50-80]', 'M', 'N'],
-      #          ['New York',  '[50-80]', 'F', 'N'],
-      #          ['Chicago',     '>80',      'F', 'Y']
-      #        ]
-      #
-      # Data labels must have the following format:
-      #     [ 'city', 'age_range', 'gender', 'marketing_target'  ]
-      #
-      # If you do not provide labels for you data, the following labels will
-      # be created by default:
-      #     [ 'ATTRIBUTE_1', 'ATTRIBUTE_2', 'ATTRIBUTE_3', 'CATEGORY'  ]
-      #
-      def build(data_examples, data_labels=nil)
-        check_data_examples(data_examples)
-        @data_labels = (data_labels) ? data_labels : default_data_labels(data_examples)
-        preprocess_data(data_examples)
+      # Create a new ID3 classifier. You must provide a DataSet instance
+      # as parameter.
+      def build(data_set)
+        data_set.check_not_empty
+        @data_set = data_set
+        preprocess_data(@data_set.data_items)
         return self
       end
 
@@ -144,7 +112,7 @@ module Ai4r
       # This method returns the generated rules in ruby code.
       # e.g.
       #   
-      #   id3.to_s
+      #   id3.get_rules
       #     # =>  if age_range=='<30' then marketing_target='Y'
       #           elsif age_range=='[30-50)' and city=='Chicago' then marketing_target='Y'
       #           elsif age_range=='[30-50)' and city=='New York' then marketing_target='N'
@@ -155,10 +123,11 @@ module Ai4r
       # It is a nice way to inspect induction results, and also to execute them:  
       #     age_range = '<30'
       #     marketing_target = nil
-      #     eval id3.to_s   
+      #     eval id3.get_rules   
       #     puts marketing_target
       #       # =>  'Y'
-      def to_s
+      def get_rules
+        #return "Empty ID3 tree" if !@tree
         rules = @tree.get_rules
         rules = rules.collect do |rule|
             "#{rule[0..-2].join(' and ')} then #{rule.last}"
@@ -175,15 +144,15 @@ module Ai4r
       def build_node(data_examples, flag_att = [])
         return ErrorNode.new if data_examples.length == 0
         domain = domain(data_examples)   
-        return CategoryNode.new(@data_labels.last, domain.last[0]) if domain.last.length == 1
+        return CategoryNode.new(@data_set.data_labels.last, domain.last[0]) if domain.last.length == 1
         min_entropy_index = min_entropy_index(data_examples, domain, flag_att)
         flag_att << min_entropy_index
         split_data_examples = split_data_examples(data_examples, domain, min_entropy_index)
-        return CategoryNode.new(@data_labels.last, most_freq(data_examples, domain)) if split_data_examples.length == 1
+        return CategoryNode.new(@data_set.data_labels.last, most_freq(data_examples, domain)) if split_data_examples.length == 1
         nodes = split_data_examples.collect do |partial_data_examples|  
           build_node(partial_data_examples, flag_att)
         end
-        return EvaluationNode.new(@data_labels, min_entropy_index, domain[min_entropy_index], nodes)
+        return EvaluationNode.new(@data_set.data_labels, min_entropy_index, domain[min_entropy_index], nodes)
       end
 
       private 
@@ -247,7 +216,7 @@ module Ai4r
       def domain(data_examples)
         #return build_domains(data_examples)
         domain = []
-        @data_labels.length.times { domain << [] }
+        @data_set.data_labels.length.times { domain << [] }
         data_examples.each do |data|
           data.each_index do |i|
             domain[i] << data[i] if i<domain.length && !domain[i].include?(data[i])
@@ -297,7 +266,7 @@ module Ai4r
       LOG2 = Math.log(2)
     end
 
-    class EvaluationNode
+    class EvaluationNode #:nodoc: all
       
       attr_reader :index, :values, :nodes
       
@@ -330,7 +299,7 @@ module Ai4r
       
     end
 
-    class CategoryNode
+    class CategoryNode #:nodoc: all
       def initialize(label, value)
         @label = label
         @value = value
@@ -343,7 +312,7 @@ module Ai4r
       end
     end
 
-    class ErrorNode
+    class ErrorNode #:nodoc: all
       def value(data)
         raise "There was not enough information during training to do a proper induction for this data element."
       end

data/lib/ai4r/classifiers/one_r.rb

@@ -8,7 +8,8 @@
 # Mozilla Foundation at http://www.mozilla.org/MPL/MPL-1.1.txt
 
 require 'set'
-require File.dirname(__FILE__) + '/classifier_helper'
+require File.dirname(__FILE__) + '/../data/data_set'
+require File.dirname(__FILE__) + '/../classifiers/classifier'
 
 module Ai4r
   module Classifiers
@@ -19,58 +20,25 @@ module Ai4r
     # attribute to use to classify data that makes 
     # fewest prediction errors.
     # It generates rules based on a single attribute.
-    class OneR
+    class OneR < Classifier
       
-      attr_accessor :data_labels, :rule
-      include ClassifierHelper
+      attr_reader :data_set, :rule
 
-      # Build a new OneR classifier. If your data is classified with N attributed
-      # and M examples, then your data examples must have the following format:
-      # 
-      #     [   [ATT1_VAL1, ATT2_VAL1, ATT3_VAL1, ... , ATTN_VAL1,  CLASS_VAL1], 
-      #         [ATT1_VAL2, ATT2_VAL2, ATT3_VAL2, ... , ATTN_VAL2,  CLASS_VAL2], 
-      #         ...
-      #         [ATTM1_VALM, ATT2_VALM, ATT3_VALM, ... , ATTN_VALM, CLASS_VALM], 
-      #     ]
-      #     
-      # e.g.
-      #     [   ['New York',  '<30',      'M', 'Y'],
-      #          ['Chicago',     '<30',      'M', 'Y'],
-      #          ['Chicago',     '<30',      'F', 'Y'],
-      #          ['New York',  '<30',      'M', 'Y'],
-      #          ['New York',  '<30',      'M', 'Y'],
-      #          ['Chicago',     '[30-50)',  'M', 'Y'],
-      #          ['New York',  '[30-50)',  'F', 'N'],
-      #          ['Chicago',     '[30-50)',  'F', 'Y'],
-      #          ['New York',  '[30-50)',  'F', 'N'],
-      #          ['Chicago',     '[50-80]', 'M', 'N'],
-      #          ['New York',  '[50-80]', 'F', 'N'],
-      #          ['New York',  '[50-80]', 'M', 'N'],
-      #          ['Chicago',     '[50-80]', 'M', 'N'],
-      #          ['New York',  '[50-80]', 'F', 'N'],
-      #          ['Chicago',     '>80',      'F', 'Y']
-      #        ]
-      #
-      # Data labels must have the following format:
-      #     [ 'city', 'age_range', 'gender', 'marketing_target'  ]
-      #
-      # If you do not provide labels for you data, the following labels will
-      # be created by default:
-      #     [ 'attribute_1', 'attribute_2', 'attribute_3', 'class_value'  ]
-      #      
-      def build(data_examples, data_labels = nil)
-        check_data_examples(data_examples)
-        @data_labels = (data_labels) ? data_labels : default_data_labels(data_examples)
-        if (num_attributes(data_examples) == 1) 
-          @zero_r = ZeroR.new.build(data_examples, data_labels)
+      # Build a new OneR classifier. You must provide a DataSet instance
+      # as parameter.
+      def build(data_set)
+        data_set.check_not_empty
+        @data_set = data_set
+        if (data_set.num_attributes == 1) 
+          @zero_r = ZeroR.new.build(data_set)
           return self;
         else
           @zero_r = nil;
         end
-        domains = build_domains(data_examples)
+        domains = @data_set.build_domains
         @rule = nil
         domains[1...-1].each_index do |attr_index|
-          rule = build_rule(data_examples, attr_index, domains)
+          rule = build_rule(@data_set.data_items, attr_index, domains)
           @rule = rule if !@rule || rule[:correct] > @rule[:correct]
         end
         return self
@@ -88,7 +56,7 @@ module Ai4r
       # This method returns the generated rules in ruby code.
       # e.g.
       #   
-      #   classifier.to_s
+      #   classifier.get_rules
       #     # =>  if age_range == '<30' then marketing_target = 'Y'
       #           elsif age_range == '[30-50)' then marketing_target = 'N'
       #           elsif age_range == '[50-80]' then marketing_target = 'N'
@@ -96,14 +64,14 @@ module Ai4r
       #
       # It is a nice way to inspect induction results, and also to execute them:  
       #     marketing_target = nil
-      #     eval classifier.to_s   
+      #     eval classifier.get_rules   
       #     puts marketing_target
       #       # =>  'Y'
-      def to_s
-        return @zero_r.to_s if @zero_r
+      def get_rules
+        return @zero_r.get_rules if @zero_r
         sentences = []
-        attr_label = @data_labels[@rule[:attr_index]]
-        class_label = @data_labels.last
+        attr_label = @data_set.data_labels[@rule[:attr_index]]
+        class_label = @data_set.data_labels.last
         @rule[:rule].each_pair do |attr_value, class_value|
           sentences << "#{attr_label} == '#{attr_value}' then #{class_label} = '#{class_value}'"
         end
@@ -111,13 +79,6 @@ module Ai4r
       end
       
       protected
-      def build_domains(data_examples)
-        domains = Array.new(num_attributes(data_examples)) { Set.new }
-        data_examples.each do |data|
-          data.each_index {|attr_index| domains[attr_index] << data[attr_index]}
-        end
-        return domains
-      end
       
       def build_rule(data_examples, attr_index, domains)
         domain = domains[attr_index]