distance_measures 0.0.1 → 0.0.2

data/.gitignore

@@ -17,5 +17,7 @@ tmtags
 coverage
 rdoc
 pkg
+tmp/*
+benchmarks/*
 
 ## PROJECT::SPECIFIC

data/Rakefile

@@ -1,16 +1,18 @@
 require 'rubygems'
 require 'rake'
+require 'rake/extensiontask'
 
 begin
   require 'jeweler'
   Jeweler::Tasks.new do |gem|
     gem.name = "distance_measures"
-    gem.summary = %Q{A bundle of distance measures}
-    gem.description = %Q{A bundle of distance measures}
+    gem.summary = %Q{A bundle of distance measures with C extensions for the slow bits}
+    gem.description = %Q{A bundle of distance measures with C extensions for the slow bits}
     gem.email = "reddavis@gmail.com"
     gem.homepage = "http://github.com/reddavis/distance_measure"
     gem.authors = ["reddavis"]
     gem.add_development_dependency "rspec", ">= 1.2.9"
+    gem.extensions = FileList['ext/**/extconf.rb']
     # gem is a Gem::Specification... see http://www.rubygems.org/read/chapter/20 for additional settings
   end
   Jeweler::GemcutterTasks.new
@@ -30,8 +32,17 @@ Spec::Rake::SpecTask.new(:rcov) do |spec|
   spec.rcov = true
 end
 
-task :spec => :check_dependencies
+# Euclidean Distance
+Rake::ExtensionTask.new('euclidean_distance') do |ext|
+  ext.lib_dir = File.join('lib', 'distance_measures')
+end
+
+# Core
+Rake::ExtensionTask.new('core') do |ext|
+  ext.lib_dir = File.join('lib', 'distance_measures')
+end
 
+task :spec => :check_dependencies
 task :default => :spec
 
 require 'rake/rdoctask'
@@ -43,3 +54,5 @@ Rake::RDocTask.new do |rdoc|
   rdoc.rdoc_files.include('README*')
   rdoc.rdoc_files.include('lib/**/*.rb')
 end
+
+Rake::Task[:spec].prerequisites << :compile

data/VERSION

@@ -1 +1 @@
-0.0.1
+0.0.2

data/distance_measures.gemspec

@@ -5,13 +5,14 @@
 
 Gem::Specification.new do |s|
   s.name = %q{distance_measures}
-  s.version = "0.0.1"
+  s.version = "0.0.2"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["reddavis"]
-  s.date = %q{2010-01-26}
-  s.description = %q{A bundle of distance measures}
+  s.date = %q{2010-07-31}
+  s.description = %q{A bundle of distance measures with C extensions for the slow bits}
   s.email = %q{reddavis@gmail.com}
+  s.extensions = ["ext/core/extconf.rb", "ext/euclidean_distance/extconf.rb"]
   s.extra_rdoc_files = [
     "LICENSE",
      "README.rdoc"
@@ -24,10 +25,14 @@ Gem::Specification.new do |s|
      "Rakefile",
      "VERSION",
      "distance_measures.gemspec",
+     "ext/core/core.c",
+     "ext/core/extconf.rb",
+     "ext/euclidean_distance/euclidean_distance.c",
+     "ext/euclidean_distance/extconf.rb",
      "lib/distance_measures.rb",
-     "lib/distance_measures/core.rb",
+     "lib/distance_measures/core.bundle",
      "lib/distance_measures/cosine_similarity.rb",
-     "lib/distance_measures/euclidean_distance.rb",
+     "lib/distance_measures/euclidean_distance.bundle",
      "lib/distance_measures/jaccard.rb",
      "lib/distance_measures/tanimoto_coefficient.rb",
      "spec/distance_measures_spec.rb",
@@ -37,8 +42,8 @@ Gem::Specification.new do |s|
   s.homepage = %q{http://github.com/reddavis/distance_measure}
   s.rdoc_options = ["--charset=UTF-8"]
   s.require_paths = ["lib"]
-  s.rubygems_version = %q{1.3.5}
-  s.summary = %q{A bundle of distance measures}
+  s.rubygems_version = %q{1.3.6}
+  s.summary = %q{A bundle of distance measures with C extensions for the slow bits}
   s.test_files = [
     "spec/distance_measures_spec.rb",
      "spec/spec_helper.rb"

data/ext/core/core.c

@@ -0,0 +1,174 @@
+#include <ruby.h>
+#include <math.h>
+
+// Prototypes
+long c_array_size(VALUE array);
+// END
+
+/*
+
+def dot_product(other)
+  sum = 0.0
+  self.each_with_index do |n, index|
+    sum += n * other[index]
+  end
+
+  sum
+end
+
+*/
+static VALUE rb_dot_product(VALUE self, VALUE other_array) {
+  double sum = 0;
+
+  //TODO: check they're the same size
+  long array_size = c_array_size(self);
+  int index;
+
+  for(index = 0; index <= array_size; index++) {
+    double x, y;
+
+    x = NUM2DBL(RARRAY(self)->ptr[index]);
+    y = NUM2DBL(RARRAY(other_array)->ptr[index]);
+
+    sum += x * y;
+  }
+
+  return rb_float_new(sum);
+}
+
+/*
+
+def sum_of_squares
+  inject(0) {|sum, n| sum + n ** 2}
+end
+
+*/
+static VALUE rb_sum_of_squares(VALUE self) {
+  double sum = 0;
+  long array_size = c_array_size(self);
+  int index;
+
+  for(index = 0; index <= array_size; index++) {
+    double x;
+
+    x = NUM2DBL(RARRAY(self)->ptr[index]);
+
+    sum += pow(x, 2);
+  }
+
+  return rb_float_new(sum);
+}
+
+/*
+
+def euclidean_normalize
+  sum = 0.0
+  self.each do |n|
+    sum += n ** 2
+  end
+
+  Math.sqrt(sum)
+end
+
+*/
+static VALUE rb_euclidean_normalize(VALUE self) {
+  double sum = 0;
+  long array_size = c_array_size(self);
+  int index;
+
+  for(index = 0; index <= array_size; index++) {
+    double x;
+
+    x = NUM2DBL(RARRAY(self)->ptr[index]);
+
+    sum += pow(x, 2);
+  }
+
+  return rb_float_new(sqrt(sum));
+}
+
+/*
+
+def binary_union_with(other)
+  unions = []
+  self.each_with_index do |n, index|
+    if n == 1 || other[index] == 1
+      unions << 1
+    else
+      unions << 0
+    end
+  end
+
+  unions
+end
+
+*/
+static VALUE rb_binary_union_with(VALUE self, VALUE other_array) {
+  //TODO: check arrays are same size
+  long array_size = c_array_size(self);
+  int index;
+  VALUE results = rb_ary_new();
+
+  for(index = 0; index <= array_size; index++) {
+    int self_attribute = NUM2INT(RARRAY(self)->ptr[index]);
+    int other_array_attribute = NUM2INT(RARRAY(other_array)->ptr[index]);
+
+    if(self_attribute == 1 || other_array_attribute == 1) {
+      rb_ary_push(results, rb_int_new(1));
+    } else {
+      rb_ary_push(results, rb_int_new(0));
+    }
+  }
+
+  return results;
+}
+
+/*
+
+def binary_intersection_with(other)
+  intersects = []
+  self.each_with_index do |n, index|
+    if n == 1 && other[index] == 1
+      intersects << 1
+    else
+      intersects << 0
+    end
+  end
+
+  intersects
+end
+
+*/
+static VALUE rb_binary_intersection_with(VALUE self, VALUE other_array) {
+  //TODO: check arrays are same size
+  long array_size = c_array_size(self);
+  int index;
+  VALUE results = rb_ary_new();
+
+  for(index = 0; index <= array_size; index++) {
+    int self_attribute = NUM2INT(RARRAY(self)->ptr[index]);
+    int other_array_attribute = NUM2INT(RARRAY(other_array)->ptr[index]);
+
+    if(self_attribute == 1 && other_array_attribute == 1) {
+      rb_ary_push(results, rb_int_new(1));
+    } else {
+      rb_ary_push(results, rb_int_new(0));
+    }
+  }
+
+  return results;
+}
+
+// return the size of a Ruby array - 1
+long c_array_size(VALUE array) {
+  return (RARRAY(array)->len - 1);
+}
+
+void Init_core() {
+  VALUE distance_measures = rb_define_module("DistanceMeasures");
+  rb_define_method(distance_measures, "dot_product", rb_dot_product, 1);
+  rb_define_method(distance_measures, "sum_of_squares", rb_sum_of_squares, 0);
+  rb_define_method(distance_measures, "euclidean_normalize", rb_euclidean_normalize, 0);
+  rb_define_method(distance_measures, "binary_union_with", rb_binary_union_with, 1);
+  rb_define_method(distance_measures, "binary_intersection_with", rb_binary_intersection_with, 1);
+}

data/ext/core/extconf.rb

@@ -0,0 +1,2 @@
+require 'mkmf'
+create_makefile('core/core')

data/ext/euclidean_distance/euclidean_distance.c

@@ -0,0 +1,37 @@
+#include <ruby.h>
+#include <math.h>
+
+/*
+
+def euclidean_distance(other)
+  sum = 0.0
+  self.each_index do |i|
+    sum += (self[i] - other[i])**2
+  end
+  Math.sqrt(sum)
+end
+
+*/
+static VALUE rb_euclidean_distance(VALUE self, VALUE other_array) {
+  double value = 0.0;
+
+  //TODO: check they're the same size
+  long vector_length = (RARRAY(self)->len - 1);
+  int index;
+
+  for(index = 0; index <= vector_length; index++) {
+    double x, y;
+
+    x = NUM2DBL(RARRAY(self)->ptr[index]);
+    y = NUM2DBL(RARRAY(other_array)->ptr[index]);
+
+    value += pow(x - y, 2);
+  }
+
+  return rb_float_new(sqrt(value));
+}
+
+void Init_euclidean_distance() {
+  VALUE distance_measures = rb_define_module("DistanceMeasures");
+  rb_define_method(distance_measures, "euclidean_distance", rb_euclidean_distance, 1);
+}

data/ext/euclidean_distance/extconf.rb

@@ -0,0 +1,2 @@
+require 'mkmf'
+create_makefile('euclidean_distance/euclidean_distance')

data/lib/distance_measures/cosine_similarity.rb

@@ -1,9 +1,8 @@
-# http://en.wikipedia.org/wiki/Cosine_similarity
 module DistanceMeasures
   def cosine_similarity(other)
     dot_product = self.dot_product(other)
     normalization = self.euclidean_normalize * other.euclidean_normalize
-        
+
     handle_nan(dot_product / normalization)
   end
 end

data/lib/distance_measures.rb

@@ -8,4 +8,22 @@ require 'distance_measures/jaccard'
 
 class Array
   include DistanceMeasures
+
+  # http://en.wikipedia.org/wiki/Intersection_(set_theory)
+  def intersection_with(other)
+    (self & other)
+  end
+
+  # http://en.wikipedia.org/wiki/Union_(set_theory)
+  def union_with(other)
+    (self + other).uniq
+  end
+
+  private
+
+  # Checks if we're dealing with NaN's and will return 0.0 unless
+  # handle NaN's is set to false
+  def handle_nan(result)
+    result.nan? ? 0.0 : result
+  end
 end

data/spec/distance_measures_spec.rb

@@ -1,101 +1,117 @@
 require File.expand_path(File.dirname(__FILE__) + '/spec_helper')
 
 describe "DistanceMeasures" do
-  
+
   describe "Euclidean Distance" do
-    it "should return 1" do
+    it "should return 0.0" do
       array.euclidean_distance(array).should == 0.0
     end
+
+    it "should return 4.0" do
+      [5].euclidean_distance([1]).should == 4.0
+    end
   end
-  
+
   describe "Cosine Similarity" do
     it "should return 1.0" do
-      array.cosine_similarity(array).should > 0.99
+      array.cosine_similarity(array).should.to_s == "1.0" # WTF
     end
-    
+
     it "should handle NaN's" do
       [0.0, 0.0].cosine_similarity([0.0, 0.0]).nan?.should be_false
     end
   end
-  
+
   describe "Tanimoto Coefficient" do
     it "should return 1.0" do
       array.tanimoto_coefficient(array).should == 1.0
     end
-    
+
     it "should handle NaN's" do
       [0.0, 0.0].tanimoto_coefficient([0.0, 0.0]).nan?.should be_false
     end
   end
-  
+
   describe "Sum of Squares" do
     it "should return 50" do
       array.sum_of_squares.should == 50
     end
   end
-  
+
   describe "Jaccard" do
     describe "Jaccard Distance" do
       it "should return" do
         array_2.jaccard_distance(array_3).should == (1 - 3.0/7.0)
       end
     end
-    
+
     describe "Jaccard Index" do
       it "should return" do
         array_2.jaccard_index(array_3).should == 3.0/7.0
       end
     end
-    
+
     describe "Binary Jaccard Index" do
       it "should return 1/4" do
         [1,1,1,1].binary_jaccard_index([0,1,0,0]).should == 1/4.0
       end
     end
   end
-  
+
   describe "Binary Jaccard Distance" do
     it "should return 0.75" do
       [1,1,1,1].binary_jaccard_distance([0,1,0,0]).should == 1 - (1/4.0)
     end
   end
-  
+
   describe "Intersection" do
     it "should return [7,4,1]" do
       array_2.intersection_with(array_3).should == [7,4,1]
     end
   end
-  
+
   describe "Union" do
     it "should return " do
-      array_2.union_with(array_3).should == [7,3,2,4,1,9,5] 
+      array_2.union_with(array_3).should == [7,3,2,4,1,9,5]
     end
   end
-  
+
   describe "Binary Intersection" do
     it "should return [0,1,0,0]" do
       [1,1,1,1].binary_intersection_with([0,1,0,0]).should == [0,1,0,0]
     end
   end
-  
+
   describe "Binary Union" do
     it "should return [1,1,1,0]" do
       [1,1,1,0].binary_union_with([0,0,0,0]).should == [1,1,1,0]
     end
   end
-  
+
+  describe "Dot Product" do
+    it "should return 50" do
+      [5, 5].dot_product([5, 5]).should == 50.0
+    end
+  end
+
+  describe "Euclidean normalize" do
+    it "should" do
+      [10].euclidean_normalize.should == 10
+    end
+  end
+
   private
-    
+
   def array
     [5, 5]
   end
-  
+
   def array_2
     [7, 3, 2, 4, 1]
   end
-  
+
   def array_3
     [4,1,9,7,5]
   end
-    
+
 end

metadata

@@ -1,7 +1,12 @@
 --- !ruby/object:Gem::Specification 
 name: distance_measures
 version: !ruby/object:Gem::Version 
-  version: 0.0.1
+  prerelease: false
+  segments: 
+  - 0
+  - 0
+  - 2
+  version: 0.0.2
 platform: ruby
 authors: 
 - reddavis
@@ -9,25 +14,30 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-01-26 00:00:00 +00:00
+date: 2010-07-31 00:00:00 +01:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
   name: rspec
-  type: :development
-  version_requirement: 
-  version_requirements: !ruby/object:Gem::Requirement 
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
+        segments: 
+        - 1
+        - 2
+        - 9
         version: 1.2.9
-    version: 
-description: A bundle of distance measures
+  type: :development
+  version_requirements: *id001
+description: A bundle of distance measures with C extensions for the slow bits
 email: reddavis@gmail.com
 executables: []
 
-extensions: []
-
+extensions: 
+- ext/core/extconf.rb
+- ext/euclidean_distance/extconf.rb
 extra_rdoc_files: 
 - LICENSE
 - README.rdoc
@@ -39,10 +49,14 @@ files:
 - Rakefile
 - VERSION
 - distance_measures.gemspec
+- ext/core/core.c
+- ext/core/extconf.rb
+- ext/euclidean_distance/euclidean_distance.c
+- ext/euclidean_distance/extconf.rb
 - lib/distance_measures.rb
-- lib/distance_measures/core.rb
+- lib/distance_measures/core.bundle
 - lib/distance_measures/cosine_similarity.rb
-- lib/distance_measures/euclidean_distance.rb
+- lib/distance_measures/euclidean_distance.bundle
 - lib/distance_measures/jaccard.rb
 - lib/distance_measures/tanimoto_coefficient.rb
 - spec/distance_measures_spec.rb
@@ -61,21 +75,23 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
+      segments: 
+      - 0
       version: "0"
-  version: 
 required_rubygems_version: !ruby/object:Gem::Requirement 
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
+      segments: 
+      - 0
       version: "0"
-  version: 
 requirements: []
 
 rubyforge_project: 
-rubygems_version: 1.3.5
+rubygems_version: 1.3.6
 signing_key: 
 specification_version: 3
-summary: A bundle of distance measures
+summary: A bundle of distance measures with C extensions for the slow bits
 test_files: 
 - spec/distance_measures_spec.rb
 - spec/spec_helper.rb

data/lib/distance_measures/core.rb

@@ -1,68 +0,0 @@
-module DistanceMeasures
-  def dot_product(other)
-    sum = 0.0
-    self.each_with_index do |n, index|
-      sum += n * other[index]
-    end
-    
-    sum
-  end
-  
-  def euclidean_normalize
-    sum = 0.0
-    self.each do |n|
-      sum += n ** 2
-    end
-    
-    Math.sqrt(sum)
-  end
-  
-  def sum_of_squares
-    inject(0) {|sum, n| sum + n ** 2}
-  end
-  
-  # http://en.wikipedia.org/wiki/Intersection_(set_theory)
-  def intersection_with(other)
-    (self & other)
-  end
-  
-  # http://en.wikipedia.org/wiki/Union_(set_theory)
-  def union_with(other)
-    (self + other).uniq
-  end
-  
-  # 1's & 0's
-  def binary_intersection_with(other)
-    intersects = []
-    self.each_with_index do |n, index|
-      if n == 1 && other[index] == 1
-        intersects << 1
-      else
-        intersects << 0
-      end
-    end
-    
-    intersects
-  end
-  
-  def binary_union_with(other)
-    unions = []
-    self.each_with_index do |n, index|
-      if n == 1 || other[index] == 1
-        unions << 1
-      else
-        unions << 0
-      end
-    end
-    
-    unions
-  end
-  
-  private
-  
-  # Checks if we're dealing with NaN's and will return 0.0 unless 
-  # handle NaN's is set to false
-  def handle_nan(result)
-    result.nan? ? 0.0 : result
-  end
-end

data/lib/distance_measures/euclidean_distance.rb

@@ -1,10 +0,0 @@
-# http://en.wikipedia.org/wiki/Euclidean_distance
-module DistanceMeasures
-  def euclidean_distance(other)
-    sum = 0.0
-    self.each_index do |i|
-      sum += (self[i] - other[i])**2
-    end
-    Math.sqrt(sum)
-  end
-end