ai4ruby 1.11

data/test/classifiers/prism_test.rb

@@ -0,0 +1,85 @@
+require 'test/unit'
+require File.dirname(__FILE__) + '/../../lib/ai4r/classifiers/prism'
+
+
+class PrismTest < Test::Unit::TestCase
+  
+  include Ai4r::Classifiers
+  include Ai4r::Data
+
+  @@data_examples = [   ['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 = [ 'city', 'age_range', 'gender', 'marketing_target'  ]
+  
+  def test_build
+    assert_raise(ArgumentError) { Prism.new.build(DataSet.new) } 
+    classifier = Prism.new.build(DataSet.new(:data_items=>@@data_examples))
+    assert_not_nil(classifier.data_set.data_labels)
+    assert_not_nil(classifier.rules)
+    assert_equal("attribute_1", classifier.data_set.data_labels.first)
+    assert_equal("class_value", classifier.data_set.data_labels.last)
+    classifier = Prism.new.build(DataSet.new(:data_items => @@data_examples, 
+        :data_labels => @@data_labels))
+    assert_not_nil(classifier.data_set.data_labels)
+    assert_not_nil(classifier.rules)
+    assert_equal("city", classifier.data_set.data_labels.first)
+    assert_equal("marketing_target", classifier.data_set.data_labels.last)
+    assert !classifier.rules.empty?
+
+    Prism.send(:public, *Prism.protected_instance_methods)
+    Prism.send(:public, *Prism.private_instance_methods)
+  end
+  
+  def test_eval
+    classifier = Prism.new.build(DataSet.new(:data_items=>@@data_examples))
+    @@data_examples.each do |data|
+      assert_equal(data.last, classifier.eval(data[0...-1]))
+    end
+  end
+  
+  def test_get_rules
+    classifier = Prism.new.build(DataSet.new(:data_items => @@data_examples, 
+        :data_labels => @@data_labels))
+    marketing_target = nil
+    age_range = nil
+    city = 'Chicago'
+    eval(classifier.get_rules) 
+    age_range = '<30'
+    eval(classifier.get_rules) 
+    assert_equal("Y", marketing_target)
+    age_range = '[30-50)'
+    eval(classifier.get_rules) 
+    assert_equal("Y", marketing_target)
+    age_range = '[30-50)'
+    city = 'New York'
+    eval(classifier.get_rules) 
+    assert_equal("N", marketing_target)
+    age_range = '[50-80]'
+    eval(classifier.get_rules) 
+    assert_equal("N", marketing_target)   
+  end
+    
+  def test_matches_conditions
+    classifier = Prism.new.build(DataSet.new(:data_labels => @@data_labels,
+      :data_items => @@data_examples))
+
+    assert classifier.matches_conditions(['New York', '<30', 'M', 'Y'], {"age_range" => "<30"})
+    assert !classifier.matches_conditions(['New York', '<30', 'M', 'Y'], {"age_range" => "[50-80]"})
+  end
+end
+

data/test/classifiers/zero_r_test.rb

@@ -0,0 +1,49 @@
+require 'test/unit'
+require File.dirname(__FILE__) + '/../../lib/ai4r/classifiers/zero_r'
+require File.dirname(__FILE__) + '/../../lib/ai4r/data/data_set'
+
+class ZeroRTest < Test::Unit::TestCase
+  
+  include Ai4r::Classifiers
+  include Ai4r::Data
+  
+  @@data_examples = [   ['New York',  '<30',      'M', 'Y'],
+                ['Chicago',     '<30',      'M', 'Y'],
+                ['New York',  '<30',      'M', 'Y'],
+                ['New York',  '[30-50)',  'F', 'N'],
+                ['Chicago',     '[30-50)',  'F', 'Y'],
+                ['New York',  '[30-50)',  'F', 'N'],
+                ['Chicago',     '[50-80]', 'M', 'N'],
+              ]
+
+  @@data_labels = [ 'city', 'age_range', 'gender', 'marketing_target'  ]
+  
+  def test_build
+    assert_raise(ArgumentError) { ZeroR.new.build(DataSet.new) } 
+    classifier = ZeroR.new.build(DataSet.new(:data_items => @@data_examples))
+    assert_equal("Y", classifier.class_value)
+    assert_equal("attribute_1", classifier.data_set.data_labels.first)
+    assert_equal("class_value", classifier.data_set.data_labels.last)
+    classifier = ZeroR.new.build(DataSet.new(:data_items => @@data_examples, 
+        :data_labels => @@data_labels))
+    assert_equal("Y", classifier.class_value)
+    assert_equal("city", classifier.data_set.data_labels.first)
+    assert_equal("marketing_target", classifier.data_set.data_labels.last)
+  end
+  
+  def test_eval
+    classifier = ZeroR.new.build(DataSet.new(:data_items => @@data_examples))
+    assert_equal('Y', classifier.eval(@@data_examples.first) )
+    assert_equal('Y', classifier.eval(@@data_examples.last) )
+  end
+  
+  def test_get_rules
+    classifier = ZeroR.new.build(DataSet.new(:data_items => @@data_examples,
+      :data_labels => @@data_labels))
+    marketing_target = nil
+    eval(classifier.get_rules) 
+    assert_equal('Y', marketing_target)
+  end
+  
+end
+

data/test/clusterers/average_linkage_test.rb

@@ -0,0 +1,51 @@
+# Author::    Sergio Fierens (implementation)
+# 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
+
+require 'test/unit'
+require File.dirname(__FILE__) + '/../../lib/ai4r/clusterers/average_linkage'
+
+class Ai4r::Clusterers::AverageLinkage < Ai4r::Clusterers::SingleLinkage
+  attr_accessor :data_set, :number_of_clusters, :clusters, :distance_matrix
+end
+
+class AverageLinkageTest < Test::Unit::TestCase
+  
+  include Ai4r::Clusterers
+  include Ai4r::Data
+
+  @@data = [  [10, 3], [3, 10], [2, 8], [2, 5], [3, 8], [10, 3],
+              [1, 3], [8, 1], [2, 9], [2, 5], [3, 3], [9, 4]]
+            
+  @@expected_distance_matrix = [ 
+        [98.0], 
+        [89.0, 5.0], 
+        [68.0, 26.0, 9.0], 
+        [74.0, 4.0, 1.0, 10.0], 
+        [0.0, 98.0, 89.0, 68.0, 74.0], 
+        [81.0, 53.0, 26.0, 5.0, 29.0, 81.0], 
+        [8.0, 106.0, 85.0, 52.0, 74.0, 8.0, 53.0], 
+        [100.0, 2.0, 1.0, 16.0, 2.0, 100.0, 37.0, 100.0], 
+        [68.0, 26.0, 9.0, 0.0, 10.0, 68.0, 5.0, 52.0, 16.0], 
+        [49.0, 49.0, 26.0, 5.0, 25.0, 49.0, 4.0, 29.0, 37.0, 5.0], 
+        [2.0, 72.0, 65.0, 50.0, 52.0, 2.0, 65.0, 10.0, 74.0, 50.0, 37.0]]
+      
+   def setup
+     Ai4r::Clusterers::AverageLinkage.send(:public, 
+       *Ai4r::Clusterers::AverageLinkage.protected_instance_methods)  
+   end
+  
+   def test_linkage_distance
+    clusterer = Ai4r::Clusterers::AverageLinkage.new
+    clusterer.distance_matrix = @@expected_distance_matrix
+    assert_equal 93.5, clusterer.linkage_distance(0,1,2)
+    assert_equal 37.5, clusterer.linkage_distance(4,2,5)
+  end
+
+end
+

data/test/clusterers/bisecting_k_means_test.rb

@@ -0,0 +1,66 @@
+# Author::    Sergio Fierens (implementation)
+# 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
+
+require 'test/unit'
+require File.dirname(__FILE__) + '/../../lib/ai4r/clusterers/bisecting_k_means'
+
+class BisectingKMeansTest < Test::Unit::TestCase
+  
+  include Ai4r::Clusterers
+  include Ai4r::Data
+
+  @@data = [  [10, 3], [3, 10], [2, 8], [2, 5], [3, 8], [10, 3],
+              [1, 3], [8, 1], [2, 9], [2, 5], [3, 3], [9, 4]]
+  
+  def test_build_without_refine
+    build(false)
+  end
+  
+  def test_build_with_refine
+    build(true)
+  end
+  
+  protected
+  def build(refine)
+    data_set = DataSet.new(:data_items => @@data, :data_labels => ["X", "Y"])
+    clusterer = BisectingKMeans.new
+    clusterer.set_parameters :refine => refine
+    clusterer.build(data_set, 4)
+    #draw_map(clusterer)
+    # Verify that all 4 clusters are created
+    assert_equal 4, clusterer.clusters.length
+    assert_equal 4, clusterer.centroids.length
+    # The addition of all instances of every cluster must be equal than
+    # the number of data points
+    total_length = 0
+    clusterer.clusters.each do |cluster| 
+      total_length += cluster.data_items.length
+    end
+    assert_equal @@data.length, total_length
+    # Data inside clusters must be the same as orifinal data
+    clusterer.clusters.each do |cluster| 
+     cluster.data_items.each do |data_item|
+       assert @@data.include?(data_item)
+     end
+    end
+  end
+  
+  private
+  def draw_map(clusterer)
+    map = Array.new(11) {Array.new(11, 0)}
+    clusterer.clusters.each_index do |i|
+      clusterer.clusters[i].data_items.each do |point|
+        map[point.first][point.last]=(i+1)
+      end
+    end
+    map.each { |row| puts row.inspect}
+  end
+  
+end
+

data/test/clusterers/centroid_linkage_test.rb

@@ -0,0 +1,53 @@
+# Author::    Sergio Fierens (implementation)
+# 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
+
+require 'test/unit'
+require File.dirname(__FILE__) + '/../../lib/ai4r/clusterers/centroid_linkage'
+
+class Ai4r::Clusterers::CentroidLinkage
+  attr_accessor :data_set, :number_of_clusters, :clusters, :distance_matrix, :index_clusters
+end
+
+class Ai4r::Clusterers::CentroidLinkageTest < Test::Unit::TestCase
+  
+  include Ai4r::Clusterers
+  include Ai4r::Data
+
+  @@data = [  [10, 3], [3, 10], [2, 8], [2, 5], [3, 8], [10, 3],
+              [1, 3], [8, 1], [2, 9], [2, 5], [3, 3], [9, 4]]
+            
+  @@expected_distance_matrix = [ 
+        [98.0], 
+        [89.0, 5.0], 
+        [68.0, 26.0, 9.0], 
+        [74.0, 4.0, 1.0, 10.0], 
+        [0.0, 98.0, 89.0, 68.0, 74.0], 
+        [81.0, 53.0, 26.0, 5.0, 29.0, 81.0], 
+        [8.0, 106.0, 85.0, 52.0, 74.0, 8.0, 53.0], 
+        [100.0, 2.0, 1.0, 16.0, 2.0, 100.0, 37.0, 100.0], 
+        [68.0, 26.0, 9.0, 0.0, 10.0, 68.0, 5.0, 52.0, 16.0], 
+        [49.0, 49.0, 26.0, 5.0, 25.0, 49.0, 4.0, 29.0, 37.0, 5.0], 
+        [2.0, 72.0, 65.0, 50.0, 52.0, 2.0, 65.0, 10.0, 74.0, 50.0, 37.0]]
+  
+   def setup
+     Ai4r::Clusterers::CentroidLinkage.send(:public, 
+       *Ai4r::Clusterers::CentroidLinkage.protected_instance_methods)  
+   end
+  
+   def test_linkage_distance
+    clusterer = Ai4r::Clusterers::CentroidLinkage.new
+    clusterer.data_set = DataSet.new :data_items => @@data
+    clusterer.index_clusters = clusterer.create_initial_index_clusters
+    clusterer.distance_matrix = @@expected_distance_matrix
+    assert_equal 92.25, clusterer.linkage_distance(0,1,2)
+    assert_equal 15.25, clusterer.linkage_distance(4,2,5)
+  end
+
+end
+

data/test/clusterers/complete_linkage_test.rb

@@ -0,0 +1,57 @@
+# Author::    Sergio Fierens (implementation)
+# 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
+
+require 'test/unit'
+require File.dirname(__FILE__) + '/../../lib/ai4r/clusterers/complete_linkage'
+
+class Ai4r::Clusterers::CompleteLinkage
+  attr_accessor :data_set, :number_of_clusters, :clusters, :distance_matrix
+end
+
+class CompleteLinkageTest < Test::Unit::TestCase
+  
+  include Ai4r::Clusterers
+  include Ai4r::Data
+
+  @@data = [  [10, 3], [3, 10], [2, 8], [2, 5], [3, 8], [10, 3],
+              [1, 3], [8, 1], [2, 9], [2, 5], [3, 3], [9, 4]]
+            
+  @@expected_distance_matrix = [ 
+        [98.0], 
+        [89.0, 5.0], 
+        [68.0, 26.0, 9.0], 
+        [74.0, 4.0, 1.0, 10.0], 
+        [0.0, 98.0, 89.0, 68.0, 74.0], 
+        [81.0, 53.0, 26.0, 5.0, 29.0, 81.0], 
+        [8.0, 106.0, 85.0, 52.0, 74.0, 8.0, 53.0], 
+        [100.0, 2.0, 1.0, 16.0, 2.0, 100.0, 37.0, 100.0], 
+        [68.0, 26.0, 9.0, 0.0, 10.0, 68.0, 5.0, 52.0, 16.0], 
+        [49.0, 49.0, 26.0, 5.0, 25.0, 49.0, 4.0, 29.0, 37.0, 5.0], 
+        [2.0, 72.0, 65.0, 50.0, 52.0, 2.0, 65.0, 10.0, 74.0, 50.0, 37.0]]
+      
+   def setup
+     Ai4r::Clusterers::CompleteLinkage.send(:public, 
+       *Ai4r::Clusterers::CompleteLinkage.protected_instance_methods)  
+   end
+  
+   def test_linkage_distance
+    clusterer = Ai4r::Clusterers::CompleteLinkage.new
+    clusterer.distance_matrix = @@expected_distance_matrix
+    assert_equal 98, clusterer.linkage_distance(0,1,2)
+    assert_equal 74, clusterer.linkage_distance(4,2,5)
+  end
+
+  def test_distance_between_item_and_cluster
+    clusterer = CompleteLinkage.new
+    assert_equal 32.0, clusterer.distance_between_item_and_cluster([1,2], 
+      DataSet.new(:data_items => [[3,4],[5,6]]))
+  end
+  
+end
+

data/test/clusterers/diana_test.rb

@@ -0,0 +1,69 @@
+# Author::    Sergio Fierens (implementation)
+# 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
+
+require 'test/unit'
+require File.dirname(__FILE__) + '/../../lib/ai4r/clusterers/diana'
+
+class Ai4r::Clusterers::Diana
+  attr_accessor :data_set, :number_of_clusters, :clusters
+end
+
+class DianaTest < Test::Unit::TestCase
+
+  @@data = [  [10, 3], [3, 10], [2, 8], [2, 5], [3, 8], [10, 3],
+              [1, 3], [8, 1], [2, 9], [2, 5], [3, 3], [9, 4]]
+              
+  include Ai4r::Clusterers
+  include Ai4r::Data
+
+  def setup
+    Diana.send(:public, *Diana.protected_instance_methods)
+    @data_set = DataSet.new(:data_items => @@data.clone)
+  end
+          
+  def test_build
+    clusterer = Diana.new.build(@data_set, 4)
+    assert_equal 4, clusterer.clusters.length
+  end
+  
+  def test_eval
+    clusterer = Diana.new.build(@data_set, 4)
+    assert_equal 0, clusterer.eval([0, 0])
+    assert_equal 1, clusterer.eval([9, 3])
+    assert_equal 2, clusterer.eval([0,10])
+    assert_equal 3, clusterer.eval([8, 0])
+  end
+  
+  def test_cluster_diameter
+    clusterer = Diana.new
+    assert_equal 106, clusterer.cluster_diameter(@data_set)
+    assert_equal 98, clusterer.cluster_diameter(@data_set[0..2])
+  end
+  
+  def test_max_diameter_cluster
+    clusterer = Diana.new
+    assert_equal 0, clusterer.max_diameter_cluster([@data_set, @data_set[0..2]])
+    assert_equal 1, clusterer.max_diameter_cluster([@data_set[0..2], @data_set])
+  end 
+
+  def test_init_splinter_cluster
+    clusterer = Diana.new 
+    assert_equal [10,3], clusterer.
+      init_splinter_cluster(@data_set).data_items[0]
+  end
+  
+  def test_max_distance_difference
+    clusterer = Diana.new 
+    data_set_a = @data_set[1..-1]
+    data_set_b = @data_set[0]
+    assert_equal [63.7, 4], clusterer.
+      max_distance_difference(data_set_a, data_set_b)
+  end
+  
+end

data/test/clusterers/k_means_test.rb

@@ -0,0 +1,100 @@
+# Author::    Sergio Fierens (implementation)
+# 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
+
+require 'test/unit'
+require File.dirname(__FILE__) + '/../../lib/ai4r/clusterers/k_means'
+
+class KMeansTest < Test::Unit::TestCase
+  
+  include Ai4r::Clusterers
+  include Ai4r::Data
+
+  @@data = [  [10, 3], [3, 10], [2, 8], [2, 5], [3, 8], [10, 3],
+              [1, 3], [8, 1], [2, 9], [2, 5], [3, 3], [9, 4]]
+  
+  def test_build
+    data_set = DataSet.new(:data_items => @@data, :data_labels => ["X", "Y"])
+    clusterer = KMeans.new.build(data_set, 4)
+    #draw_map(clusterer)
+    # Verify that all 4 clusters are created
+    assert_equal 4, clusterer.clusters.length
+    assert_equal 4, clusterer.centroids.length
+    # The addition of all instances of every cluster must be equal than
+    # the number of data points
+    total_length = 0
+    clusterer.clusters.each do |cluster| 
+      total_length += cluster.data_items.length
+    end
+    assert_equal @@data.length, total_length
+    # Data inside clusters must be the same as orifinal data
+    clusterer.clusters.each do |cluster| 
+     cluster.data_items.each do |data_item|
+       assert @@data.include?(data_item)
+     end
+    end
+  end
+  
+  def test_eval
+    data_set = DataSet.new(:data_items => @@data, :data_labels => ["X", "Y"])
+    clusterer = KMeans.new.build(data_set, 4)
+    item = [10,0]
+    cluster_index = clusterer.eval(item)
+    # Must return a valid cluster index [0-3]
+    assert cluster_index >= 0 && cluster_index < 4
+    # Distance to cluster centroid must be less than distance to any other
+    # centroid
+    min_distance = clusterer.distance(clusterer.centroids[cluster_index], item)
+    clusterer.centroids.each do |centroid|
+      assert clusterer.distance(centroid, item) >= min_distance
+    end
+  end
+  
+  def test_distance
+    clusterer = KMeans.new
+    # By default, distance returns the eucledian distance to the power of 2
+    assert_equal 2385, clusterer.distance(
+      [1, 10, "Chicago", 2], 
+      [10, 10, "London", 50])
+    # Test new distance definition
+    manhattan_distance = lambda do |a, b| 
+        dist = 0.0
+        a.each_index do |index|
+          if a[index].is_a?(Numeric) && b[index].is_a?(Numeric)
+            dist = dist + (a[index]-b[index]).abs
+          end
+        end
+        dist
+    end
+    clusterer.set_parameters({:distance_function => manhattan_distance})
+    assert_equal 57, clusterer.distance(
+      [1, 10, "Chicago", 2], 
+      [10, 10, "London", 50])
+  end
+
+  def test_max_iterations
+    data_set = DataSet.new(:data_items => @@data, :data_labels => ["X", "Y"])
+    clusterer = KMeans.new.
+      set_parameters({:max_iterations=>1}).
+      build(data_set, 4)
+    assert_equal 1, clusterer.iterations
+  end
+  
+  private
+  def draw_map(clusterer)
+    map = Array.new(11) {Array.new(11, 0)}
+    clusterer.clusters.each_index do |i|
+      clusterer.clusters[i].data_items.each do |point|
+        map[point.first][point.last]=(i+1)
+      end
+    end
+    map.each { |row| puts row.inspect}
+  end
+  
+end
+