classifier 1.0 → 1.1

data/Rakefile

@@ -5,7 +5,7 @@ require 'rake/rdoctask'
 require 'rake/gempackagetask'
 require 'rake/contrib/rubyforgepublisher'
 
-PKG_VERSION = "1.0.0"
+PKG_VERSION = "1.1"
 
 PKG_FILES = FileList[
     "lib/**/*", "bin/*", "test/**/*", "[A-Z]*", "Rakefile"
@@ -38,7 +38,7 @@ spec = Gem::Specification.new do |s|
   #### Basic information.
 
   s.name = 'classifier'
-  s.version = "1.0"
+  s.version = PKG_VERSION
   s.summary = <<-EOF
    A general classifier module to allow Bayesian and other types of classifications.
   EOF
@@ -74,4 +74,4 @@ end
 desc "Publish to RubyForge"
 task :rubyforge do
     Rake::RubyForgePublisher.new('classifier', 'cardmagic').upload
-end
+end

data/lib/classifier/bayes.rb

@@ -12,7 +12,7 @@ class Bayes
 	end
 
 	def classify(text)
-		(classifications(text).sort { |a, b| b[1] <=> a[1] })[0][0]
+		(classifications(text).sort_by { |a| -a[1] })[0][0]
 	end
 	
 	def classifications(text)
@@ -29,9 +29,9 @@ class Bayes
 	end
 
 	def method_missing(name, *args)
-		category = name.to_s.gsub(/train_([\w]+)/, '\1').capitalize.intern
+		category = name.to_s.gsub(/train_([\w]+)/, '\1').gsub("_"," ").capitalize.intern
 		if @categories.has_key? category
-			args.each {|text| add_words category, text}
+			args.each {|text| train category, text}
 		elsif name.to_s =~ /train_([\w]+)/
 			raise StandardError, "No such category: #{category}"
 		else
@@ -39,9 +39,8 @@ class Bayes
 		end
 	end
 
-	private
-
-	def add_words(category, text)
+	def train(category, text)
+		category = category.to_s.gsub("_"," ").capitalize.intern
 		text.word_hash.each do |word, count|
 			@categories[category][word]	||=	0
 			@categories[category][word]	 +=	count
@@ -50,4 +49,4 @@ class Bayes
 	end
 end
 
-end
+end

data/lib/classifier/string_extensions/porter_stemmer.rb

@@ -1,3 +1,11 @@
+# = Author
+# 
+# Greg Fast, gdf@speakeasy.net
+# 
+# = Copyright
+# 
+# Copyright 2005 Greg Fast <gdf@speakeasy.net>
+
 module Classifier
 	
 module Stemmable
@@ -185,4 +193,4 @@ module Stemmable
 
 end
 
-end
+end

data/lib/classifier/string_extensions/word_hash.rb

@@ -12,13 +12,15 @@ module WordHash
 	def word_hash
 		d = Hash.new
 		corpus = without_punctuation
-		(corpus.split + gsub(/[\w]/,"").split).each do |word|
+		(corpus.split + gsub(/[\w+]/,"").split).each do |word|
 			key = word.downcase.stem.intern
-			d[key] ||= 0
-			d[key] += 1
+			if !(word =~ /[\w+]/) || word.length > 2
+				d[key] ||= 0
+				d[key] += 1
+			end
 		end
 		return d
 	end
 end
 
-end
+end

data/test/{bayes/string_extensions_test.rb → string_extensions/word_hash_test.rb}

@@ -1,8 +1,8 @@
 require File.dirname(__FILE__) + '/../test_helper'
 class StringExtensionsTest < Test::Unit::TestCase
 	def test_word_hash
-		hash = {:some=>1, :good=>1, :hope=>1, :word=>1, :you=>1, :here=>1, :love=>1, :i=>1, :ar=>1, :them=>1, :"."=>1, :"!"=>1}
+		hash = {:some=>1, :good=>1, :hope=>1, :word=>1, :you=>1, :here=>1, :love=>1, :ar=>1, :them=>1, :"."=>1, :"!"=>1}
 		
 		assert_equal hash, "here are some good words. I hope you love them!".word_hash
 	end
-end
+end

metadata

@@ -3,8 +3,8 @@ rubygems_version: 0.8.6
 specification_version: 1
 name: classifier
 version: !ruby/object:Gem::Version 
-  version: "1.0"
-date: 2005-04-10
+  version: "1.1"
+date: 2005-04-11
 summary: A general classifier module to allow Bayesian and other types of classifications.
 require_paths: 
   - lib
@@ -36,9 +36,10 @@ files:
   - lib/classifier/string_extensions/word_hash.rb
   - bin/bayes.rb
   - test/bayes
+  - test/string_extensions
   - test/test_helper.rb
   - test/bayes/bayesian_test.rb
-  - test/bayes/string_extensions_test.rb
+  - test/string_extensions/word_hash_test.rb
   - Rakefile
   - README
 test_files: []