rrbayes 0.0.2 → 1.0.0

data/Rakefile

@@ -10,7 +10,7 @@ begin
     gem.email = "hungryblank@gmail.com"
     gem.homepage = "http://github.com/hungryblank/rrbayes"
     gem.authors = ["hungryblank"]
-    gem.add_dependency "redis", ">= 0"
+    gem.add_dependency "redis", ">= 2.0.0"
     gem.add_development_dependency "riot", ">= 0"
     # gem is a Gem::Specification... see http://www.rubygems.org/read/chapter/20 for additional settings
   end

data/VERSION

@@ -1 +1 @@
-0.0.2
+1.0.0

data/lib/rrbayes/category.rb

@@ -14,18 +14,16 @@ module Rrbayes
     end
 
     def learn(frequency_map)
-      @db.pipelined do |pipe|
+      @db.pipelined do
         frequency_map.each do |attribute, evidences|
-          pipe.incrby(attribute_key(attribute), evidences)
-          pipe.incrby(evidences_key, evidences)
+          store_evidences(attribute, evidences)
         end
-        pipe.incr(documents_key)
-        pipe.incr(DOCUMENTS_SCOPE)
+        increment_documents
       end
     end
 
     def attributes_score(attributes)
-      total = total_evidences.to_f
+      total = evidences_total.to_f
       attributes.map do |attribute, evidences|
         (evidences_for(attribute) || 0.1).to_f / total
       end.inject(0) do |score, attr_likelyhood|
@@ -34,17 +32,27 @@ module Rrbayes
     end
 
     def evidences_for(attribute)
-      @db.get(attribute_key(attribute))
+      @db.hget(name, attribute)
     end
 
-    def total_evidences
+    def evidences_total
       @db.get(evidences_key)
     end
 
+    def documents_total
+      @db.get(documents_key)
+    end
+
     private
 
-    def attribute_key(attribute)
-      key_for(name, attribute)
+    def store_evidences(attribute, evidences)
+      @db.hincrby(name, attribute, evidences)
+      @db.incrby(evidences_key, evidences)
+    end
+
+    def increment_documents
+      @db.incr(documents_key)
+      @db.incr(DOCUMENTS_SCOPE)
     end
 
     def evidences_key

data/lib/rrbayes/rrbayes.rb

@@ -15,7 +15,6 @@ module Rrbayes
     #
     def initialize(options = {}, redis_options = {})
       @db = Redis.new(redis_options)
-      @db.connect_to_server
       @categories = find_categories(options).map { |c| Category.new(c, self) }
     end
 
@@ -37,7 +36,7 @@ module Rrbayes
     #  => 'spam'
     #
     def classify(frequency_map)
-      @categories.map { |c| [c.name, c.attributes_score(frequency_map)] }.sort_by { |c| -c[1] }.first[0]
+      @categories.sort_by { |c| -c.attributes_score(frequency_map) }.first.name
     end
 
     #Returns the category objects with the name provided in the argument
@@ -48,7 +47,7 @@ module Rrbayes
     private
 
     def load_categories
-      @db.set_members('categories')
+      @db.smembers('categories')
     end
 
     def find_categories(options)

data/rrbayes.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = %q{rrbayes}
-  s.version = "0.0.2"
+  s.version = "1.0.0"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["hungryblank"]
-  s.date = %q{2009-12-22}
+  s.date = %q{2010-05-26}
   s.description = %q{Implementation of a naive Bayes classifier with a Redis backend}
   s.email = %q{hungryblank@gmail.com}
   s.extra_rdoc_files = [
@@ -27,7 +27,7 @@ Gem::Specification.new do |s|
      "lib/rrbayes/category.rb",
      "lib/rrbayes/rrbayes.rb",
      "rrbayes.gemspec",
-     "test/rrbayes_test.rb",
+     "test/classifier_test.rb",
      "test/teststrap.rb"
   ]
   s.homepage = %q{http://github.com/hungryblank/rrbayes}
@@ -37,7 +37,7 @@ Gem::Specification.new do |s|
   s.summary = %q{Rrbayes: Ruby Redis Bayes}
   s.test_files = [
     "test/teststrap.rb",
-     "test/rrbayes_test.rb"
+     "test/classifier_test.rb"
   ]
 
   if s.respond_to? :specification_version then
@@ -45,14 +45,14 @@ Gem::Specification.new do |s|
     s.specification_version = 3
 
     if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then
-      s.add_runtime_dependency(%q<redis>, [">= 0"])
+      s.add_runtime_dependency(%q<redis>, [">= 2.0.0"])
       s.add_development_dependency(%q<riot>, [">= 0"])
     else
-      s.add_dependency(%q<redis>, [">= 0"])
+      s.add_dependency(%q<redis>, [">= 2.0.0"])
       s.add_dependency(%q<riot>, [">= 0"])
     end
   else
-    s.add_dependency(%q<redis>, [">= 0"])
+    s.add_dependency(%q<redis>, [">= 2.0.0"])
     s.add_dependency(%q<riot>, [">= 0"])
   end
 end

data/test/classifier_test.rb

@@ -0,0 +1,48 @@
+require 'teststrap'
+
+include Rrbayes
+
+DB_NUM = 0
+
+context "a new spam/ham classifier" do
+
+  setup do
+    db = Redis.new :db => DB_NUM
+    db.flushdb
+    @classifier = Classifier.new({:categories => %w(spam ham)}, :db => DB_NUM)
+  end
+
+  asserts("categories") { topic.db.smembers('categories') }.equals %w(spam ham)
+
+  asserts("loaded categories") { topic.send(:load_categories) }.equals %w(spam ham)
+
+  asserts("recognized categories") { Classifier.new({}, :db => DB_NUM).categories.map { |c| c.name } }.equals %w(spam ham)
+
+
+  asserts("initialize with same categories") { Classifier.new({:categories => %w(spam ham)}, :db => DB_NUM).categories.map { |c| c.name } == %w(spam ham) }
+
+  asserts("category mismatch") { Classifier.new({:categories => %w(bad good)}, :db => DB_NUM) }.raises(LoadingError)
+
+  context "trained with spam and ham" do
+
+    setup do
+      spam_frequencies = {'enlarge' => 1, 'your' => 2 ,'viagra' => 3}
+      @classifier.learn(spam_frequencies, :as => 'spam')
+
+      ham_frequencies = {'dear' => 1, 'Jon' => 2 ,'how' => 3}
+      @classifier.learn(ham_frequencies, :as => 'ham')
+    end
+
+    asserts("evidences number for spam") { @classifier.category('spam').evidences_for('viagra') }.equals '3'
+
+    asserts("category documents total") { @classifier.category('spam').documents_total }.equals '1'
+
+    asserts("evidences total") { @classifier.db['documents'] }.equals '2'
+
+    should("classify spam") { @classifier.classify({'viagra' => 1, 'cheap' => 1}) == 'spam' }
+
+    should("classify ham") { @classifier.classify({'dear' => 1, 'molly' => 2}) == 'ham' }
+
+  end
+
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: rrbayes
 version: !ruby/object:Gem::Version 
-  version: 0.0.2
+  version: 1.0.0
 platform: ruby
 authors: 
 - hungryblank
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-12-22 00:00:00 -08:00
+date: 2010-05-26 00:00:00 -07:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -20,7 +20,7 @@ dependencies:
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
-        version: "0"
+        version: 2.0.0
     version: 
 - !ruby/object:Gem::Dependency 
   name: riot
@@ -52,7 +52,7 @@ files:
 - lib/rrbayes/category.rb
 - lib/rrbayes/rrbayes.rb
 - rrbayes.gemspec
-- test/rrbayes_test.rb
+- test/classifier_test.rb
 - test/teststrap.rb
 has_rdoc: true
 homepage: http://github.com/hungryblank/rrbayes
@@ -84,4 +84,4 @@ specification_version: 3
 summary: "Rrbayes: Ruby Redis Bayes"
 test_files: 
 - test/teststrap.rb
-- test/rrbayes_test.rb
+- test/classifier_test.rb

data/test/rrbayes_test.rb

@@ -1,52 +0,0 @@
-require 'teststrap'
-
-include Rrbayes
-
-DB_NUM = 0
-
-context "a new classifier" do
-
-  setup do
-    db = Redis.new :db => DB_NUM
-    db.connect_to_server
-    db.flushdb
-    @classifier = Classifier.new({:categories => %w(spam ham)}, :db => DB_NUM)
-  end
-
-  should("persist categories") { topic.db.set_members('categories') }.equals %w(spam ham)
-
-  should("load categories") { topic.send(:load_categories) }.equals %w(spam ham)
-
-  should("initialize with no categories") { Classifier.new({}, :db => DB_NUM).categories.map { |c| c.name } }.equals %w(spam ham)
-
-  should("initialize with same categories") { Classifier.new({:categories => %w(spam ham)}, :db => DB_NUM).categories.map { |c| c.name } }.equals %w(spam ham)
-
-  should("detect category mismatch") { Classifier.new({:categories => %w(bad good)}, :db => DB_NUM) }.raises(LoadingError)
-
-  context "in training" do
-
-    setup do
-      spam_frequencies = {'enlarge' => 1, 'your' => 2 ,'viagra' => 3}
-      @classifier.learn(spam_frequencies, :as => 'spam')
-
-      ham_frequencies = {'dear' => 1, 'Jon' => 2 ,'how' => 3}
-      @classifier.learn(ham_frequencies, :as => 'ham')
-    end
-
-    should("store evidences number") { @classifier.db['spam:viagra'] }.equals '3'
-
-    should("store category evidences total") { @classifier.db['documents:spam'] }.equals '1'
-
-    should("store evidences total") { @classifier.db['documents'] }.equals '2'
-
-    context "for a while" do
-
-      should("classify_spam") { @classifier.classify({'viagra' => 1, 'cheap' => 1}) }.equals 'spam'
-
-      should("classify_ham") { @classifier.classify({'dear' => 1, 'molly' => 2}) }.equals 'ham'
-
-    end
-
-  end
-
-end