scoruby 0.2.2 → 0.2.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 29d319117ac389447c955bcd448235865fd72104
-  data.tar.gz: 851feb27437019ef236b7ab8130c52800f3239a4
+  metadata.gz: 59d91f99c8a04a124993b71950d425f6b1b89353
+  data.tar.gz: ee709d362a7699749561a7d1cb60f7b7aa40c902
 SHA512:
-  metadata.gz: e13cff01a0cf6800aa30c522dbb27bee0cfa931f55c7aa466811f7a46aff6fc865d9d9cdcacd2ed462b334cd6e38078c6e92a5235287cadd62383ae8abf1ff22
-  data.tar.gz: b8664649eeef47a290f0c8547b486896d0044b3803580639468d41f8f1098e27c4de8b5f0df06a8807463e39c8d827e101fac4457de1ef3dcee947887404fe37
+  metadata.gz: 4d433e761e5fc203d298ecec3fb5019e024ae6cd753296f08b304fc0dd790017f5ff0e8621a34995811d6db21c475f3e91a41d079722c5c1ec888d151740c9c6
+  data.tar.gz: c347e88a7cf8e5345be89160f6f6f6f07ff2bb7dfed93401f02700e898886b3205e0ced850ba1fad247779f4b0f16c712b751aca46be15ac0d75b6fe3f1cc9d1

data/.gitignore

@@ -0,0 +1,10 @@
+.idea
+*.log
+
+test_gbm.rb
+
+test_gbm.pmml
+
+*.gem
+
+sample.pmml

data/Gemfile.lock

@@ -0,0 +1,67 @@
+PATH
+  remote: .
+  specs:
+    scoruby (0.2.2)
+      nokogiri (~> 1.7)
+
+GEM
+  remote: https://rubygems.org/
+  specs:
+    coderay (1.1.0)
+    coveralls (0.8.20)
+      json (>= 1.8, < 3)
+      simplecov (~> 0.14.1)
+      term-ansicolor (~> 1.3)
+      thor (~> 0.19.4)
+      tins (~> 1.6)
+    diff-lcs (1.2.5)
+    docile (1.1.5)
+    json (2.1.0)
+    method_source (0.8.2)
+    mini_portile2 (2.1.0)
+    nokogiri (1.7.0.1)
+      mini_portile2 (~> 2.1.0)
+    pry (0.10.3)
+      coderay (~> 1.1.0)
+      method_source (~> 0.8.1)
+      slop (~> 3.4)
+    rake (12.0.0)
+    rspec (3.5.0)
+      rspec-core (~> 3.5.0)
+      rspec-expectations (~> 3.5.0)
+      rspec-mocks (~> 3.5.0)
+    rspec-core (3.5.4)
+      rspec-support (~> 3.5.0)
+    rspec-expectations (3.5.0)
+      diff-lcs (>= 1.2.0, < 2.0)
+      rspec-support (~> 3.5.0)
+    rspec-mocks (3.5.0)
+      diff-lcs (>= 1.2.0, < 2.0)
+      rspec-support (~> 3.5.0)
+    rspec-support (3.5.0)
+    ruby-prof (0.16.2)
+    simplecov (0.14.1)
+      docile (~> 1.1.0)
+      json (>= 1.8, < 3)
+      simplecov-html (~> 0.10.0)
+    simplecov-html (0.10.0)
+    slop (3.6.0)
+    term-ansicolor (1.6.0)
+      tins (~> 1.0)
+    thor (0.19.4)
+    tins (1.13.2)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  bundler (~> 1.10)
+  coveralls
+  pry (~> 0.10)
+  rake (~> 12.0)
+  rspec (~> 3.5)
+  ruby-prof
+  scoruby!
+
+BUNDLED WITH
+   1.11.2

data/lib/scoruby/decision.rb

@@ -0,0 +1,17 @@
+module Scoruby
+  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
+end

data/lib/scoruby/features.rb

@@ -0,0 +1,18 @@
+module Scoruby
+  class Features
+
+    attr_reader :formatted
+
+    def initialize(features)
+      @formatted = format_booleans(features)
+    end
+
+    def format_booleans(features)
+      features.map {|k, v|
+        features[k] = 'f' if v == false
+        features[k] = 't' if v == true
+      }
+      features
+    end
+  end
+end

data/lib/scoruby/models/decision_tree.rb

@@ -0,0 +1,46 @@
+require 'node'
+
+module Scoruby
+  module Models
+    class DecisionTree
+
+      attr_reader :root
+
+      def initialize(tree_xml)
+        @id   = tree_xml.attribute('id')
+        @root = Node.new(tree_xml.xpath('TreeModel/Node'))
+      end
+
+      def decide(features)
+        curr = @root
+        while curr.children[0]
+          prev = curr
+          curr = step(curr, features)
+          return if didnt_step?(curr, prev)
+        end
+
+        curr.decision
+      end
+
+      private
+
+      def step(curr, features)
+        curr = step_on_true(curr, features, 0)
+        curr = step_on_true(curr, features, 1)
+        curr = step_on_true(curr, features, 2)
+        curr
+      end
+
+      def step_on_true(curr, features, num)
+        return curr.children[num] if curr.children && curr.children[num] && curr.children[num].true?(features)
+        curr
+      end
+
+      def didnt_step?(curr, prev)
+        return false if (prev.pred != curr.pred)
+        Scoruby.logger.error "Null tree: #{@id}, bad feature: #{curr.children[0].pred.field }"
+        true
+      end
+    end
+  end
+end

data/lib/scoruby/models/gbm.rb

@@ -0,0 +1,31 @@
+require 'models/decision_tree'
+require 'features'
+
+module Scoruby
+  module Models
+    class Gbm
+      GBM_FOREST_XPATH = '//Segmentation[@multipleModelMethod="sum"]/Segment'
+      CONST_XPATH      = '//Target/@rescaleConstant'
+
+      def initialize(xml)
+        @decision_trees = xml.xpath(GBM_FOREST_XPATH).collect {|xml_tree|
+          DecisionTree.new(xml_tree)
+        }
+        @const          = Float(xml.xpath(CONST_XPATH).to_s)
+      end
+
+      def tree_count
+        @decision_trees.count
+      end
+
+      def score(features)
+        formatted_features = Features.new(features).formatted
+        x                  = @decision_trees.map {|dt|
+          score = dt.decide(formatted_features).score
+          score.to_s.to_f
+        }.reduce(:+) + @const
+        Math.exp(x) / (1 + Math.exp(x))
+      end
+    end
+  end
+end

data/lib/scoruby/models/random_forest.rb

@@ -0,0 +1,26 @@
+module Scoruby
+  module Models
+    class RandomForest
+      RF_FOREST_XPATH = 'PMML/MiningModel/Segmentation/Segment'
+
+      def initialize(xml)
+        xml_trees       = xml.xpath(RF_FOREST_XPATH)
+        @decision_trees = xml_trees.collect {|xml_tree|
+          DecisionTree.new(xml_tree)
+        }
+      end
+
+      def decisions_count(features)
+        formatted_features = Features.new(features).formatted
+        decisions          = @decision_trees.collect {|decision_tree|
+          decision_tree.decide(formatted_features).score
+        }
+        decisions.inject(Hash.new(0)) {|h, e| h[e] += 1; h}
+      end
+
+      def predict(features)
+        decisions_count(features).max_by {|_, v| v}[0]
+      end
+    end
+  end
+end

data/lib/scoruby/models_factory.rb

@@ -0,0 +1,31 @@
+require 'models/decision_tree'
+require 'models/gbm'
+require 'models/random_forest'
+
+module Scoruby
+  class ModelsFactory
+    RANDOM_FOREST_MODEL       = 'randomForest_Model'
+    GBM_INDICATION            = '//OutputField[@name="scaledGbmValue"]'
+    MODEL_NOT_SUPPORTED_ERROR = 'model not supported'
+
+    def self.factory_for(xml)
+      return Models::RandomForest.new(xml) if random_forest?(xml)
+      return Models::Gbm.new(xml) if gbm?(xml)
+      return Models::DecisionTree.new(xml.child) if decision_tree?(xml)
+
+      raise MODEL_NOT_SUPPORTED_ERROR
+    end
+
+    def self.decision_tree?(xml)
+      !xml.xpath('PMML/TreeModel').empty?
+    end
+
+    def self.random_forest?(xml)
+      xml.xpath('PMML/MiningModel/@modelName').to_s == RANDOM_FOREST_MODEL
+    end
+
+    def self.gbm?(xml)
+      !xml.xpath(GBM_INDICATION).empty?
+    end
+  end
+end

data/lib/scoruby/node.rb

@@ -0,0 +1,38 @@
+require 'predicate_factory'
+require 'decision'
+
+module Scoruby
+  class Node
+
+    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     = PredicateFactory.for(pred_xml)
+      @children = []
+
+      return if children.count == 1
+
+      @children << Node.new(children[1]) if children[1]
+      @children << Node.new(children[2]) if children[2]
+      @children << Node.new(children[3]) if children[3]
+    end
+
+    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
+end

data/lib/scoruby/predicate_factory.rb

@@ -0,0 +1,19 @@
+require 'predicates/compound_predicate'
+require 'predicates/simple_predicate'
+require 'predicates/simple_set_predicate'
+require 'predicates/true_predicate'
+require 'predicates/false_predicate'
+
+module Scoruby
+  class PredicateFactory
+
+    def self.for(pred_xml)
+      return Predicates::SimplePredicate.new(pred_xml) if pred_xml.name == 'SimplePredicate'
+      return Predicates::SimpleSetPredicate.new(pred_xml) if pred_xml.name == 'SimpleSetPredicate'
+      return Predicates::CompoundPredicate.new(pred_xml) if pred_xml.name == 'CompoundPredicate'
+      return Predicates::TruePredicate.new if pred_xml.name == 'True'
+      return Predicates::FalsePredicate.new if pred_xml.name == 'False'
+    end
+  end
+end
+

data/lib/scoruby/predicates/compound_predicate.rb

@@ -0,0 +1,44 @@
+module Scoruby
+  module Predicates
+    class CompoundPredicate
+
+      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).flatten.compact
+      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
+
+      def is_missing?(features)
+        @field.any? {|f| !features.keys.include?(f)}
+      end
+
+      private
+
+      def surrogate?(features)
+        return @predicates[1].true?(features) if @predicates[0].is_missing?(features)
+        @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
+  end
+end

data/lib/scoruby/predicates/false_predicate.rb

@@ -0,0 +1,17 @@
+module Scoruby
+  module Predicates
+    class FalsePredicate
+      def field
+        nil
+      end
+
+      def true?(_)
+        false
+      end
+
+      def is_missing?(_)
+        false
+      end
+    end
+  end
+end

data/lib/scoruby/predicates/simple_predicate.rb

@@ -0,0 +1,47 @@
+module Scoruby
+  module Predicates
+    class SimplePredicate
+
+      GREATER_THAN     = 'greaterThan'
+      LESS_THAN        = 'lessThan'
+      LESS_OR_EQUAL    = 'lessOrEqual'
+      GREATER_OR_EQUAL = 'greaterOrEqual'
+      MATH_OPS         = [GREATER_THAN, LESS_THAN, LESS_OR_EQUAL, GREATER_OR_EQUAL]
+      EQUAL            = 'equal'
+      IS_MISSING       = 'isMissing'
+
+      attr_reader :field
+
+      def initialize(pred_xml)
+        attributes = pred_xml.attributes
+
+        @field    = attributes['field'].value.to_sym
+        @operator = attributes['operator'].value
+        return if @operator == IS_MISSING
+        @value = attributes['value'].value
+      end
+
+      def true?(features)
+        return num_true?(features) if MATH_OPS.include?(@operator)
+        return features[@field] == @value if @operator == EQUAL
+        features[field].nil? || !features.has_key?(field) if @operator == IS_MISSING
+      end
+
+      def is_missing?(features)
+        !features.keys.include?(@field)
+      end
+
+      private
+
+      def num_true?(features)
+        return false unless features[@field]
+        curr_value = Float(features[@field])
+        value      = Float(@value)
+        return curr_value > value if @operator == GREATER_THAN
+        return curr_value < value if @operator == LESS_THAN
+        return curr_value <= value if @operator == LESS_OR_EQUAL
+        curr_value >= value if @operator == GREATER_OR_EQUAL
+      end
+    end
+  end
+end

data/lib/scoruby/predicates/simple_set_predicate.rb

@@ -0,0 +1,33 @@
+module Scoruby
+  module Predicates
+    class SimpleSetPredicate
+
+      IS_IN = 'isIn'
+
+      attr_reader :field
+
+      def initialize(pred_xml)
+        attributes = pred_xml.attributes
+        @field     = attributes['field'].value.to_sym
+        @array     = single_or_quoted_words(pred_xml.children[0].content)
+        @operator  = attributes['booleanOperator'].value
+      end
+
+      def true?(features)
+        @array.include? features[@field] if @operator == IS_IN
+      end
+
+      def is_missing?(features)
+        !features.keys.include?(@field)
+      end
+
+      private
+
+      def single_or_quoted_words(string)
+        string.split(/\s(?=(?:[^"]|"[^"]*")*$)/).
+          reject(&:empty?).
+          map {|w| w.tr('"', '')}
+      end
+    end
+  end
+end

data/lib/scoruby/predicates/true_predicate.rb

@@ -0,0 +1,17 @@
+module Scoruby
+  module Predicates
+    class TruePredicate
+      def field
+        nil
+      end
+
+      def true?(_)
+        true
+      end
+
+      def is_missing?(_)
+        false
+      end
+    end
+  end
+end

data/lib/scoruby/version.rb

@@ -1,3 +1,3 @@
 module Scoruby
-  VERSION = '0.2.2'
+  VERSION = '0.2.3'
 end

data/lib/scoruby.rb

@@ -1,8 +1,7 @@
 require 'scoruby/version'
-require 'models_factory'
+require 'scoruby/models_factory'
 require 'nokogiri'
 require 'logger'
-require 'pry'
 
 module Scoruby
   class << self

data/scoruby.gemspec

@@ -16,7 +16,7 @@ Gem::Specification.new do |spec|
   spec.files         = `git ls-files -z`.split("\x0").reject { |f| f.match(%r{^(test|spec|features)/}) }
   spec.bindir        = "exe"
   spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
-  spec.require_paths = ["lib", "lib/random_forest", "lib/gbm"]
+  spec.require_paths = ["lib", "lib/scoruby", "lib/scoruby/models/random_forest", "lib/scoruby/models/gbm"]
 
   spec.add_development_dependency "bundler", "~> 1.10"
   spec.add_development_dependency "rake", "~> 12.0"

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: scoruby
 version: !ruby/object:Gem::Version
-  version: 0.2.2
+  version: 0.2.3
 platform: ruby
 authors:
 - Asaf Schers
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2017-07-09 00:00:00.000000000 Z
+date: 2017-07-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -115,29 +115,31 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- ".gitignore"
 - ".rspec"
 - ".travis.yml"
 - CODE_OF_CONDUCT.md
 - Gemfile
+- Gemfile.lock
 - LICENSE.txt
 - README.md
 - Rakefile
 - bin/console
 - bin/setup
-- lib/decision.rb
-- lib/features.rb
-- lib/models/decision_tree.rb
-- lib/models/gbm.rb
-- lib/models/random_forest.rb
-- lib/models_factory.rb
-- lib/node.rb
-- lib/predicate_factory.rb
-- lib/predicates/compound_predicate.rb
-- lib/predicates/false_predicate.rb
-- lib/predicates/simple_predicate.rb
-- lib/predicates/simple_set_predicate.rb
-- lib/predicates/true_predicate.rb
 - lib/scoruby.rb
+- lib/scoruby/decision.rb
+- lib/scoruby/features.rb
+- lib/scoruby/models/decision_tree.rb
+- lib/scoruby/models/gbm.rb
+- lib/scoruby/models/random_forest.rb
+- lib/scoruby/models_factory.rb
+- lib/scoruby/node.rb
+- lib/scoruby/predicate_factory.rb
+- lib/scoruby/predicates/compound_predicate.rb
+- lib/scoruby/predicates/false_predicate.rb
+- lib/scoruby/predicates/simple_predicate.rb
+- lib/scoruby/predicates/simple_set_predicate.rb
+- lib/scoruby/predicates/true_predicate.rb
 - lib/scoruby/version.rb
 - scoruby.gemspec
 homepage: https://github.com/asafschers/scoruby
@@ -148,8 +150,9 @@ post_install_message:
 rdoc_options: []
 require_paths:
 - lib
-- lib/random_forest
-- lib/gbm
+- lib/scoruby
+- lib/scoruby/models/random_forest
+- lib/scoruby/models/gbm
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="

data/lib/decision.rb

@@ -1,15 +0,0 @@
-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/features.rb

@@ -1,16 +0,0 @@
-class Features
-
-  attr_reader :formatted
-
-  def initialize(features)
-    @formatted = format_booleans(features)
-  end
-
-  def format_booleans(features)
-    features.map { |k, v|
-      features[k] = 'f' if v == false
-      features[k] = 't' if v == true
-    }
-    features
-  end
-end

data/lib/models/decision_tree.rb

@@ -1,42 +0,0 @@
-require 'node'
-
-class DecisionTree
-
-  attr_reader :root
-
-  def initialize(tree_xml)
-    @id = tree_xml.attribute('id')
-    @root = Node.new(tree_xml.xpath('TreeModel/Node'))
-  end
-
-  def decide(features)
-    curr = @root
-    while curr.children[0]
-      prev = curr
-      curr = step(curr, features)
-      return if didnt_step?(curr, prev)
-    end
-
-    curr.decision
-  end
-
-  private
-
-  def step(curr, features)
-    curr = step_on_true(curr, features, 0)
-    curr = step_on_true(curr, features, 1)
-    curr = step_on_true(curr, features, 2)
-    curr
-  end
-
-  def step_on_true(curr, features, num)
-    return curr.children[num] if curr.children && curr.children[num] && curr.children[num].true?(features)
-    curr
-  end
-
-  def didnt_step?(curr, prev)
-    return false if (prev.pred != curr.pred)
-    Scoruby.logger.error "Null tree: #{@id}, bad feature: #{curr.children[0].pred.field }"
-    true
-  end
-end

data/lib/models/gbm.rb

@@ -1,29 +0,0 @@
-require 'models/decision_tree'
-require 'features'
-
-class Gbm
-  GBM_FOREST_XPATH = '//Segmentation[@multipleModelMethod="sum"]/Segment'
-  CONST_XPATH      = '//Target/@rescaleConstant'
-
-  def initialize(xml)
-    @decision_trees = xml.xpath(GBM_FOREST_XPATH).collect{ |xml_tree|
-      DecisionTree.new(xml_tree)
-    }
-    @const = Float(xml.xpath(CONST_XPATH).to_s)
-  end
-
-  def tree_count
-    @decision_trees.count
-  end
-
-  def score(features)
-    formatted_features = Features.new(features).formatted
-    x = @decision_trees.map { |dt|
-      score = dt.decide(formatted_features).score
-      score.to_s.to_f
-    }.reduce(:+) + @const
-    Math.exp(x) / (1 + Math.exp(x))
-  end
-
-end
-

data/lib/models/random_forest.rb

@@ -1,25 +0,0 @@
-require 'models/decision_tree'
-
-class RandomForest
-  RF_FOREST_XPATH = 'PMML/MiningModel/Segmentation/Segment'
-
-  def initialize(xml)
-    xml_trees = xml.xpath(RF_FOREST_XPATH)
-    @decision_trees = xml_trees.collect{ |xml_tree|
-      DecisionTree.new(xml_tree)
-    }
-  end
-
-  def decisions_count(features)
-    formatted_features = Features.new(features).formatted
-    decisions = @decision_trees.collect { |decision_tree|
-      decision_tree.decide(formatted_features).score
-    }
-    decisions.inject(Hash.new(0)) { |h, e| h[e] += 1 ; h }
-  end
-
-  def predict(features)
-    decisions_count(features).max_by {|_, v|  v }[0]
-  end
-
-end

data/lib/models_factory.rb

@@ -1,28 +0,0 @@
-require 'models/random_forest'
-require 'models/gbm'
-
-class ModelsFactory
-  RANDOM_FOREST_MODEL = 'randomForest_Model'
-  GBM_INDICATION = '//OutputField[@name="scaledGbmValue"]'
-  MODEL_NOT_SUPPORTED_ERROR = 'model not supported'
-
-  def self.factory_for(xml)
-    return RandomForest.new(xml) if random_forest?(xml)
-    return Gbm.new(xml) if gbm?(xml)
-    return DecisionTree.new(xml.child) if decision_tree?(xml)
-
-    raise MODEL_NOT_SUPPORTED_ERROR
-  end
-
-  def self.decision_tree?(xml)
-    !xml.xpath('PMML/TreeModel').empty?
-  end
-
-  def self.random_forest?(xml)
-    xml.xpath('PMML/MiningModel/@modelName').to_s == RANDOM_FOREST_MODEL
-  end
-
-  def self.gbm?(xml)
-    !xml.xpath(GBM_INDICATION).empty?
-  end
-end

data/lib/node.rb

@@ -1,36 +0,0 @@
-require 'predicate_factory'
-require 'decision'
-
-class Node
-
-  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 = PredicateFactory.for(pred_xml)
-    @children = []
-
-    return if children.count == 1
-
-    @children << Node.new(children[1]) if children[1]
-    @children << Node.new(children[2]) if children[2]
-    @children << Node.new(children[3]) if children[3]
-  end
-
-  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/predicate_factory.rb

@@ -1,18 +0,0 @@
-require 'predicates/compound_predicate'
-require 'predicates/simple_predicate'
-require 'predicates/simple_set_predicate'
-require 'predicates/true_predicate'
-require 'predicates/false_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
-
-

data/lib/predicates/compound_predicate.rb

@@ -1,40 +0,0 @@
-class CompoundPredicate
-
-  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).flatten.compact
-  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
-
-  def is_missing?(features)
-    @field.any? { |f| !features.keys.include?(f) }
-  end
-
-  private
-
-  def surrogate?(features)
-    return @predicates[1].true?(features) if @predicates[0].is_missing?(features)
-    @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/predicates/false_predicate.rb

@@ -1,13 +0,0 @@
-class FalsePredicate
-  def field
-    nil
-  end
-
-  def true?(_)
-    false
-  end
-
-  def is_missing?(_)
-    false
-  end
-end

data/lib/predicates/simple_predicate.rb

@@ -1,43 +0,0 @@
-class SimplePredicate
-
-  GREATER_THAN = 'greaterThan'
-  LESS_THAN = 'lessThan'
-  LESS_OR_EQUAL = 'lessOrEqual'
-  GREATER_OR_EQUAL = 'greaterOrEqual'
-  MATH_OPS = [GREATER_THAN, LESS_THAN, LESS_OR_EQUAL, GREATER_OR_EQUAL]
-  EQUAL = 'equal'
-  IS_MISSING = 'isMissing'
-
-  attr_reader :field
-
-  def initialize(pred_xml)
-    attributes = pred_xml.attributes
-
-    @field = attributes['field'].value.to_sym
-    @operator = attributes['operator'].value
-    return if @operator == IS_MISSING
-    @value = attributes['value'].value 
-  end
-
-  def true?(features)
-    return num_true?(features) if MATH_OPS.include?(@operator)
-    return features[@field] == @value if @operator == EQUAL
-    features[field].nil? || !features.has_key?(field) if @operator == IS_MISSING
-  end
-
-  def is_missing?(features)
-    !features.keys.include?(@field)
-  end
-
-  private
-
-  def num_true?(features)
-    return false unless features[@field]
-    curr_value = Float(features[@field])
-    value = Float(@value)
-    return curr_value > value if @operator == GREATER_THAN
-    return curr_value < value if @operator == LESS_THAN
-    return curr_value <= value if @operator == LESS_OR_EQUAL
-    curr_value >= value if @operator == GREATER_OR_EQUAL
-  end
-end

data/lib/predicates/simple_set_predicate.rb

@@ -1,29 +0,0 @@
-class SimpleSetPredicate
-
-  IS_IN = 'isIn'
-
-  attr_reader :field
-
-  def initialize(pred_xml)
-    attributes = pred_xml.attributes
-    @field = attributes['field'].value.to_sym
-    @array = single_or_quoted_words(pred_xml.children[0].content)
-    @operator = attributes['booleanOperator'].value
-  end
-
-  def true?(features)
-    @array.include? features[@field] if @operator == IS_IN
-  end
-
-  def is_missing?(features)
-    !features.keys.include?(@field)
-  end
-
-  private
-
-  def single_or_quoted_words(string)
-    string.split(/\s(?=(?:[^"]|"[^"]*")*$)/).
-        reject(&:empty?).
-        map { |w| w.tr('"','')}
-  end
-end

data/lib/predicates/true_predicate.rb

@@ -1,13 +0,0 @@
-class TruePredicate
-  def field
-    nil
-  end
-
-  def true?(_)
-    true
-  end
-
-  def is_missing?(_)
-    false
-  end
-end