scoruby 0.2.0 → 0.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: f1a4d1fcce322b50113aa532d34ea7397e17dfa7
-  data.tar.gz: e39e4e849ac6c98c14625372e1e17219b7ae2999
+  metadata.gz: e090c4502c064f07dc851a28873a6e60bbd7a301
+  data.tar.gz: 3ef721448831698663309efa187103971ebe8ce1
 SHA512:
-  metadata.gz: bdd1e14f38a6ab1675c54e7b795bc56f7bab845f6b05f3f8725c1b439737e23471dec7a3f94b60d982f8dbdef349a892d4c045c2f48ff45e0b0c06057599e86e
-  data.tar.gz: ec3ae16a4b3fd4b6299f89c925718ad9583fca0d7d27dc45661edc5ea8597a183869a29552aeaa74ddb01af92f421144a89cf1adedf93527e0d0811855c3c0f0
+  metadata.gz: da34ae5f3501a5f73cf28c43bd4d736f07fda38988c0f4c4504c8418a4f9e990e9a97eb94afbd2af405001252100c3b96ab43c1a61f102083b19f5725aa90714
+  data.tar.gz: e038681f58084d1c80998d4d6075969917072c6e45b52c67427311e9b304e85a9f9d8d560e99d20be3e0045d89ce2afcdee0bd37509dc9b446caff998b06fa30

data/lib/random_forest/compound_predicate.rb

@@ -0,0 +1,37 @@
+class CompoundPredicate
+  # TODO: spec
+
+  attr_reader :field
+
+  def initialize(pred_xml)
+    attributes = pred_xml.attributes
+    children = pred_xml.children
+
+    @boolean_operator = attributes['booleanOperator'].value
+    @predicates = []
+    @predicates << PredicateFactory.for(children[0])
+    @predicates << PredicateFactory.for(children[1])
+    @field = @predicates.map(&:field)
+  end
+
+  def true?(features)
+    return surrogate?(features) if @boolean_operator == 'surrogate'
+    return or?(features) if @boolean_operator == 'or'
+    and?(features) if @boolean_operator == 'and'
+  end
+
+  private
+
+  def surrogate?(features)
+    # TODO: return 1 if 0 is missing
+    @predicates[0].true?(features)
+  end
+
+  def or?(features)
+    @predicates.any? { |p| p.true?(features) }
+  end
+
+  def and?(features)
+    @predicates.all? { |p| p.true?(features) }
+  end
+end

data/lib/random_forest/decision.rb

@@ -0,0 +1,15 @@
+class Decision
+
+  attr_reader :score, :score_distribution
+
+  def initialize(score, score_distributions)
+    @score = score
+    return if score_distributions.empty?
+
+    @score_distribution = {}
+    score_distributions.each { |score_distribution|
+      attributes = score_distribution.attributes
+      @score_distribution[attributes['value'].to_s]  = attributes['probability'].to_s
+    }
+  end
+end

data/lib/random_forest/decision_tree.rb

@@ -11,7 +11,7 @@ class DecisionTree
 
   def decide(features)
     curr = @root
-    while curr.decision == ''
+    while curr.children[0]
       prev = curr
       curr = step(curr, features)
       return if didnt_step?(curr, prev)

data/lib/random_forest/gbm.rb

@@ -19,7 +19,7 @@ class Gbm
   def score(features)
     formatted_features = Features.new(features).formatted
     x = @decision_trees.map { |dt|
-      score = dt.decide(formatted_features)
+      score = dt.decide(formatted_features).score
       score.to_s.to_f
     }.reduce(:+) + @const
     Math.exp(x) / (1 + Math.exp(x))

data/lib/random_forest/node.rb

@@ -1,17 +1,21 @@
-require 'simple_predicate'
-require 'simple_set_predicate'
+require 'predicate_factory'
+require 'decision'
 
 class Node
 
-  attr_reader :decision, :pred, :children 
+  attr_reader :decision, :pred, :children
 
   def initialize(xml)
     children = xml.children
+
+    @decision = Decision.new(xml.attribute('score').to_s,
+                             children.select { |c| c.name == 'ScoreDistribution' } )
+
+    children = remove_nodes(children)
+
     pred_xml = children[0]
-    @pred = SimplePredicate.new(pred_xml) if pred_xml.name == 'SimplePredicate'
-    @pred = SimpleSetPredicate.new(pred_xml) if pred_xml.name == 'SimpleSetPredicate'
+    @pred = PredicateFactory.for(pred_xml)
     @children = []
-    @decision = xml.attribute('score').to_s
 
     return if children.count == 1
 
@@ -23,4 +27,10 @@ class Node
   def true?(features)
     @pred.nil? || @pred.true?(features)
   end
+
+  private
+
+  def remove_nodes(children)
+    children.reject { |c| %w(Extension ScoreDistribution).include? c.name  }
+  end
 end

data/lib/random_forest/predicate_factory.rb

@@ -0,0 +1,35 @@
+require 'compound_predicate'
+require 'simple_predicate'
+require 'simple_set_predicate'
+
+class PredicateFactory
+
+  def self.for(pred_xml)
+    return SimplePredicate.new(pred_xml) if pred_xml.name == 'SimplePredicate'
+    return SimpleSetPredicate.new(pred_xml) if pred_xml.name == 'SimpleSetPredicate'
+    return CompoundPredicate.new(pred_xml) if pred_xml.name == 'CompoundPredicate'
+    return TruePredicate.new if pred_xml.name == 'True'
+    return FalsePredicate.new if pred_xml.name == 'False'
+  end
+end
+
+class TruePredicate
+  def field
+    nil
+  end
+
+  def true?(_)
+    true
+  end
+end
+
+class FalsePredicate
+  def field
+    nil
+  end
+
+  def true?(_)
+    false
+  end
+end
+

data/lib/random_forest/random_forest.rb

@@ -13,7 +13,7 @@ class RandomForest
   def decisions_count(features)
     formatted_features = Features.new(features).formatted
     decisions = @decision_trees.collect { |decision_tree|
-      decision_tree.decide(formatted_features)
+      decision_tree.decide(formatted_features).score
     }
     decisions.inject(Hash.new(0)) { |h, e| h[e] += 1 ; h }
   end

data/lib/scoruby/version.rb

@@ -1,3 +1,3 @@
 module Scoruby
-  VERSION = '0.2.0'
+  VERSION = '0.2.1'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: scoruby
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.2.1
 platform: ruby
 authors:
 - Asaf Schers
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2017-07-01 00:00:00.000000000 Z
+date: 2017-07-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -124,10 +124,13 @@ files:
 - Rakefile
 - bin/console
 - bin/setup
+- lib/random_forest/compound_predicate.rb
+- lib/random_forest/decision.rb
 - lib/random_forest/decision_tree.rb
 - lib/random_forest/features.rb
 - lib/random_forest/gbm.rb
 - lib/random_forest/node.rb
+- lib/random_forest/predicate_factory.rb
 - lib/random_forest/random_forest.rb
 - lib/random_forest/simple_predicate.rb
 - lib/random_forest/simple_set_predicate.rb