omnicat 0.2.2 → 0.3.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 65d6fbb05908315286db625537482191aa4067e4
+  data.tar.gz: c79f9470ca00165736ed3bc78b8dd9f6578b682a
+SHA512:
+  metadata.gz: 069cefc4f20e453dd59af4fb7ad6e46e0b9529bbb63b8dfe998ea3058a0f8f00c308210a3527fa901e422b20af481586a27fe0fd6eab3bd5133f0026e8cb3fea
+  data.tar.gz: 79a279215e8eb6984725c9fe302970c233dec11ef58095e8edefec68bf1596e5fd41f07c3a33826b3c26f3917491c8c067114bd83c05286abec68d498b6a0e33

data/lib/omnicat.rb

@@ -3,6 +3,7 @@ require File.dirname(__FILE__) + '/omnicat/configuration'
 require File.dirname(__FILE__) + '/omnicat/array'
 require File.dirname(__FILE__) + '/omnicat/base'
 require File.dirname(__FILE__) + '/omnicat/doc'
+require File.dirname(__FILE__) + '/omnicat/score'
 require File.dirname(__FILE__) + '/omnicat/result'
 require File.dirname(__FILE__) + '/omnicat/classifier'
 

data/lib/omnicat/classifier.rb

@@ -10,7 +10,7 @@ module OmniCat
     attr_accessor :strategy
 
     # delegate category methods
-    def_delegators :@strategy, :add_category, :add_categories
+    def_delegators :@strategy, :add_category, :add_categories, :category_size_limit
 
     # delegate training methods
     def_delegators :@strategy, :train, :train_batch, :untrain, :untrain_batch
@@ -26,30 +26,79 @@ module OmniCat
       @strategy = classifier
     end
 
+    # Changes classifier strategy and train new strategy if needed
+    #
     def strategy=(classifier)
       is_interchangeable?(classifier)
-      if @strategy && classifier.doc_count == 0
+      if @strategy && classifier.category_count == 0
         previous_strategy = @strategy
         @strategy = classifier
-        # pass previous strategy contents into the new one
-        previous_strategy.categories.each do |category_name, category|
-          @strategy.add_category(category_name)
-          category.docs.each do |_, doc|
-            doc.count.times do
-              @strategy.train(category_name, doc.content)
-            end
-          end
-        end
+        convert_categories_with_docs(previous_strategy)
       else
         @strategy = classifier
       end
     end
 
     private
+      # nodoc
+      def convert_categories_with_docs(previous_strategy)
+        if previous_strategy.categories.is_a?(Hash)
+          convert_categories_hash(previous_strategy.categories)
+        else
+          convert_categories_array(previous_strategy.categories)
+        end
+      end
+
+      # nodoc
+      def convert_categories_array(categories)
+        categories.each do |category|
+          convert_category(category)
+        end
+      end
+
+      # nodoc
+      def convert_categories_hash(categories)
+        categories.each do |_, category|
+          convert_category(category)
+        end
+      end
+
+      # nodoc
+      def convert_category(category)
+        @strategy.add_category(category.name)
+        if category.docs.is_a?(Hash)
+          convert_docs_hash(category.name, category.docs)
+        else
+          convert_docs_array(category.name, category.docs)
+        end
+      end
+
+      # nodoc
+      def convert_docs_array(category_name, docs)
+        docs.each do |doc|
+          convert_doc(category_name, doc)
+        end
+      end
+
+      # nodoc
+      def convert_docs_hash(category_name, docs)
+        docs.each do |_, doc|
+          convert_doc(category_name, doc)
+        end
+      end
+
+      # nodoc
+      def convert_doc(category_name, doc)
+        doc.count.times do
+          @strategy.train(category_name, doc.content)
+        end
+      end
+
+      # nodoc
       def is_interchangeable?(classifier)
-        if classifier.category_size_limit
+        unless classifier.category_size_limit == 0
           if @strategy.category_count > classifier.category_size_limit
-            raise StandardError, 
+            raise StandardError,
               'New classifier category size limit is less than the current classifier\'s category count.'
           end
         end

data/lib/omnicat/classifiers/strategy.rb

@@ -9,12 +9,12 @@ module OmniCat
     #
     # The class supplies abstract methods for possible text classifiers
     class Strategy < ::OmniCat::Base
-      attr_accessor :categories # ::OmniCat::Hash - Hash of categories
+      attr_accessor :categories # Hash - Hash of categories
       attr_accessor :category_count # Integer - Total category count
-      attr_accessor :category_size_limit # Integer - Max allowed category
-      attr_accessor :doc_count # Integer - Total token count
+      attr_accessor :category_size_limit # Integer - Max allowed category size
+      attr_accessor :doc_count # Integer - Total doc count
       attr_accessor :token_count # Integer - Total token count
-      attr_accessor :uniq_token_count # Integer - Total uniq token count
+      attr_accessor :unique_token_count # Integer - Total uniq token count
 
       def initialize(strategy_hash = {})
         @categories = {}
@@ -22,16 +22,16 @@ module OmniCat
         @category_size_limit = strategy_hash[:category_size_limit].to_i
         @doc_count = strategy_hash[:doc_count].to_i
         @token_count = strategy_hash[:token_count].to_i
-        @uniq_token_count = strategy_hash[:uniq_token_count].to_i
+        @unique_token_count = strategy_hash[:unique_token_count].to_i
       end
 
       # Abstract method for adding new classification category
       #
       # ==== Parameters
       #
-      # * +name+ - Name for category
+      # * +category_name+ - Name for category
       #
-      def add_category(name)
+      def add_category(category_name)
         not_implemented_error(__callee__)
       end
 
@@ -39,20 +39,20 @@ module OmniCat
       #
       # ==== Parameters
       #
-      # * +names+ - Array of categories
+      # * +category_names+ - Array of categories
       #
-      def add_categories(names)
-        names.each { |name| add_category(name) }
+      def add_categories(category_names)
+        category_names.each { |category_name| add_category(category_name) }
       end
 
       # Abstract method for training the desired category with a document
       #
       # ==== Parameters
       #
-      # * +category+ - Name of the category from added categories list
-      # * +doc+ - Document text
+      # * +category_name+ - Name of the category from added categories list
+      # * +doc_content+ - Document text
       #
-      def train(category_name, doc)
+      def train(category_name, doc_content)
         not_implemented_error(__callee__)
       end
 
@@ -60,21 +60,21 @@ module OmniCat
       #
       # ==== Parameters
       #
-      # * +category+ - Name of the category from added categories list
-      # * +docs+ - Array of documents
+      # * +category_name+ - Name of the category from added categories list
+      # * +doc_contents+ - Array of documents
       #
-      def train_batch(category, docs)
-        docs.each { |doc| train(category, doc) }
+      def train_batch(category_name, doc_contents)
+        doc_contents.each { |doc_content| train(category_name, doc_content) }
       end
 
       # Abstract method for untraining the desired category with a document
       #
       # ==== Parameters
       #
-      # * +category+ - Name of the category from added categories list
-      # * +doc+ - Document text
+      # * +category_name+ - Name of the category from added categories list
+      # * +doc_content+ - Document text
       #
-      def untrain(category_name, doc)
+      def untrain(category_name, doc_content)
         not_implemented_error(__callee__)
       end
 
@@ -82,24 +82,24 @@ module OmniCat
       #
       # ==== Parameters
       #
-      # * +category+ - Name of the category from added categories list
-      # * +docs+ - Array of documents
+      # * +category_name+ - Name of the category from added categories list
+      # * +doc_contents+ - Array of documents
       #
-      def untrain_batch(category, docs)
-        docs.each { |doc| untrain(category, doc) }
+      def untrain_batch(category_name, doc_contents)
+        doc_contents.each { |doc_content| untrain(category_name, doc_content) }
       end
 
       # Abstract method for classifying the given document
       #
       # ==== Parameters
       #
-      # * +doc+ - The document for classification
+      # * +doc_content+ - The document for classification
       #
       # ==== Returns
       #
       # * +result+ - OmniCat::Result object
       #
-      def classify(doc)
+      def classify(doc_content)
         not_implemented_error(__callee__)
       end
 
@@ -107,14 +107,14 @@ module OmniCat
       #
       # ==== Parameters
       #
-      # * +docs+ - Array of documents
+      # * +doc_contents+ - Array of documents
       #
       # ==== Returns
       #
       # * +result_set+ - Array of OmniCat::Result objects
       #
-      def classify_batch(docs)
-        docs.collect { |doc| classify(doc) }
+      def classify_batch(doc_contents)
+        doc_contents.collect { |doc_content| classify(doc_content) }
       end
 
       private
@@ -157,7 +157,7 @@ module OmniCat
             raise StandardError,
                   'At least 2 categories needed for classification process!'
             false
-          elsif doc_avability? == false
+          elsif doc_availability? == false
             raise StandardError,
                   'Each category must trained with at least one document!'
             false
@@ -167,7 +167,7 @@ module OmniCat
         end
 
         # nodoc
-        def doc_avability?
+        def doc_availability?
           @categories.each do |_, category|
             return false if category.doc_count == 0
           end

data/lib/omnicat/classifiers/strategy_internals/category.rb

@@ -4,11 +4,12 @@ module OmniCat
   module Classifiers
     module StrategyInternals
       class Category < ::OmniCat::Base
-        attr_accessor :doc_count, :docs, :tokens, :token_count
+        attr_accessor :doc_count, :docs, :name, :token_count, :tokens
 
         def initialize(category_hash = {})
           @doc_count = category_hash[:doc_count].to_i
           @docs = category_hash[:docs] || {}
+          @name = category_hash[:name] || nil
           @tokens = category_hash[:tokens] || {}
           @token_count = category_hash[:token_count].to_i
         end

data/lib/omnicat/configuration.rb

@@ -6,7 +6,7 @@ require 'logger'
 module OmniCat
   class Configuration
     include Singleton
-    attr_accessor :logger
+    attr_accessor :auto_train # [:off, :continues, :unique]
     attr_accessor :exclude_tokens, :logger, :token_patterns
 
     def self.default_logger
@@ -16,6 +16,7 @@ module OmniCat
     end
 
     @@defaults = {
+      auto_train: :off,
       exclude_tokens: ['a','about','across','after','all','almost','also','am','among','an','and','are','as','at','be','because','been','by','did','do','does','else','ever','every','for','from','get','got','had','has','have','he','her','hers','him','his','how','however','i','if','in','into','is','it','its','just','least','let','may','me','might','most','must','my','of','often','on','only','or','other','our','own','rather','said','say','says','she','should','since','so','some','than','that','the','their','them','then','there','these','they','this','tis','to','too','twas','us','wants','was','we','were','what','when','where','which','while','who','whom','will','with','would','yet','you','your'],
       logger: default_logger,
       token_patterns: {

data/lib/omnicat/doc.rb

@@ -1,12 +1,15 @@
 # encoding: UTF-8
+require 'digest'
 require File.dirname(__FILE__) + '/base'
 
 module OmniCat
   class Doc < ::OmniCat::Base
-    attr_reader :content, :count, :tokens
+    attr_reader :content, :content_md5, :count, :tokens
 
     def initialize(doc_hash = {})
+      @auto_classified = doc_hash[:auto_classified] || false
       @content = doc_hash[:content]
+      @content_md5 = doc_hash[:content_md5] || Digest::MD5.hexdigest("#{@content}")
       @count = (doc_hash[:count] || 1).to_i
       @tokens = tokenize_with_counts unless @tokens.is_a?(Hash)
     end

data/lib/omnicat/result.rb

@@ -2,12 +2,52 @@ require File.dirname(__FILE__) + '/base'
 
 module OmniCat
   class Result < ::OmniCat::Base
-    attr_accessor :category, :scores, :total_score
+    attr_accessor :scores
 
     def initialize
-      self.category = {}
-      self.scores = {}
-      self.total_score = 0
+      @top_score_key = nil
+      @scores = {}
+      @total_score = 0
     end
+
+    # Method for adding new score to result
+    #
+    # ==== Parameters
+    #
+    # * +score+ - OmniCat::Score
+    #
+    def add_score(score)
+      @total_score += score.value
+      @scores[score.key] = score
+      if @top_score_key.nil? || @scores[@top_score_key].value < score.value
+        @top_score_key = score.key
+      end
+    end
+
+    # Method for getting highest ranked score
+    #
+    # ==== Returns
+    #
+    # * +score+ - OmniCat::Score
+    #
+    def top_score
+      @scores[@top_score_key]
+    end
+
+    # Method for calculating percentages for scores
+    #
+    def calculate_percentages
+      @scores.each do |key, score|
+        @scores[key].percentage = percentage(score.value)
+      end
+    end
+
+    private
+      attr_reader :top_score_key, :total_score
+
+      # nodoc
+      def percentage(value)
+        (value * 100.0 / @total_score).round(0)
+      end
   end
 end

data/lib/omnicat/score.rb

@@ -0,0 +1,15 @@
+require File.dirname(__FILE__) + '/base'
+
+module OmniCat
+  class Score < ::OmniCat::Base
+    attr_accessor :key
+    attr_accessor :value
+    attr_accessor :percentage
+
+    def initialize(score_hash = {})
+      @key = score_hash[:key]
+      @value = score_hash[:value]
+      @percentage = score_hash[:percentage] || 0
+    end
+  end
+end

data/lib/omnicat/version.rb

@@ -1,3 +1,3 @@
 module OmniCat
-  VERSION = '0.2.2'
+  VERSION = '0.3.0'
 end

data/omnicat.gemspec

@@ -18,7 +18,7 @@ Gem::Specification.new do |spec|
   spec.test_files    = spec.files.grep(%r{^(test|spec|features)/})
   spec.require_paths = ['lib']
 
-  spec.add_dependency 'hashable', '~> 0.1.0'
+  spec.add_runtime_dependency 'hashable', '~> 0.1', '>= 0.1.2'
   spec.add_development_dependency 'bundler', '~> 1.3'
-  spec.add_development_dependency 'rake'
+  spec.add_development_dependency 'rake', '~> 10.1'
 end

data/test/unit/result_test.rb

@@ -0,0 +1,32 @@
+require File.expand_path(File.join(File.dirname(__FILE__), '..', 'test_helper'))
+
+class TestResult < Test::Unit::TestCase
+  def setup
+    @result = OmniCat::Result.new
+    @score1 = OmniCat::Score.new(key: 'pos', value: 0.43)
+    @score2 = OmniCat::Score.new(key: 'net', value: 0.76)
+    @score3 = OmniCat::Score.new(key: 'neg', value: 0.11)
+  end
+
+  def test_add_score
+    @result.add_score(@score1)
+    assert_equal(@score1, @result.scores[@score1.key])
+  end
+
+  def test_top_score
+    @result.add_score(@score1)
+    @result.add_score(@score2)
+    @result.add_score(@score3)
+    assert_equal(@score2, @result.top_score)
+  end
+
+  def test_percentage
+    @result.add_score(@score1)
+    @result.add_score(@score2)
+    @result.add_score(@score3)
+    @result.calculate_percentages
+    assert_equal(33, @score1.percentage)
+    assert_equal(58, @score2.percentage)
+    assert_equal(8, @score3.percentage)
+  end
+end

data/test/unit/score_test.rb

@@ -0,0 +1,11 @@
+require File.expand_path(File.join(File.dirname(__FILE__), '..', 'test_helper'))
+
+class TestScore < Test::Unit::TestCase
+  def setup
+    @score = OmniCat::Score.new(key: 'pos', value: 0.43)
+  end
+
+  def test_percentage
+    assert_equal(0, @score.percentage)
+  end
+end

metadata

@@ -1,64 +1,63 @@
 --- !ruby/object:Gem::Specification
 name: omnicat
 version: !ruby/object:Gem::Version
-  version: 0.2.2
-  prerelease: 
+  version: 0.3.0
 platform: ruby
 authors:
 - Mustafa Turan
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-07-11 00:00:00.000000000 Z
+date: 2014-02-19 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: hashable
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.1.0
+        version: '0.1'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.1.2
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ~>
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.1'
+    - - ">="
       - !ruby/object:Gem::Version
-        version: 0.1.0
+        version: 0.1.2
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.3'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.3'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '10.1'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '10.1'
 description: A generalized framework for text classifications.
 email:
 - mustafaturan.net@gmail.com
@@ -66,8 +65,8 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- .gitignore
-- .travis.yml
+- ".gitignore"
+- ".travis.yml"
 - CHANGELOG.txt
 - Gemfile
 - LICENSE.txt
@@ -82,6 +81,7 @@ files:
 - lib/omnicat/configuration.rb
 - lib/omnicat/doc.rb
 - lib/omnicat/result.rb
+- lib/omnicat/score.rb
 - lib/omnicat/version.rb
 - omnicat.gemspec
 - test/test_helper.rb
@@ -89,30 +89,31 @@ files:
 - test/unit/classifier_test.rb
 - test/unit/classifiers/strategy_test.rb
 - test/unit/doc_test.rb
+- test/unit/result_test.rb
+- test/unit/score_test.rb
 homepage: https://github.com/mustafaturan/omnicat
 licenses:
 - MIT
+metadata: {}
 post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.23
+rubygems_version: 2.2.0
 signing_key: 
-specification_version: 3
+specification_version: 4
 summary: A generalized framework for text classifications.
 test_files:
 - test/test_helper.rb
@@ -120,3 +121,5 @@ test_files:
 - test/unit/classifier_test.rb
 - test/unit/classifiers/strategy_test.rb
 - test/unit/doc_test.rb
+- test/unit/result_test.rb
+- test/unit/score_test.rb