measurable 0.0.3 → 0.0.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: c923bf9e9bd70c37d84330fcbb9d883f72344b04
-  data.tar.gz: bea042df8b59927f38b7ace662f7f8bb41f3f33a
+  metadata.gz: 24f0ca4dbb60cda53bab68a614a171df7e337434
+  data.tar.gz: 8675c8a2e723203f287ce4dac3a6e6237fe2b675
 SHA512:
-  metadata.gz: 5ce3eaec6a905c087b6538baf92172e0099b149eaf206b6af456645d6bc6e9b3b3975566de4c44a8dc84ef0c7dbc6cd11a53ce40d7f8de1b404540b37fa12c52
-  data.tar.gz: a9d1eb70f7c8b1e13878f23b3001a0eabfbbbe6fb59bf9b56704b6f3e8672ea6cf0954a92d8556219eb6a8ab6f79c7489dede6b82c33230bab821b56ec63ef71
+  metadata.gz: ff4de5c4fbbe64592a16e7980182a76fa7e3960931d401004f9cebd8e439ea17acf0e16b8a465131b80e832fd560fd97f5fd6e1054f43678ded44d730a4e90c3
+  data.tar.gz: 62396f9fb4208745628848a5447872bb86e407d2b8ec34b15acaae6ac193f8a2c1a63aa68ef69046d5d6080a693c4cfc38bc9f419c697445eb251bb861cc9af4

data/.rspec

@@ -1 +1,2 @@
---color
+--color
+--require spec_helper

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    measurable (0.0.3)
+    measurable (0.0.4)
 
 GEM
   remote: http://rubygems.org/

data/Rakefile

@@ -1,5 +1,6 @@
 require 'rake'
 require 'bundler/gem_tasks'
+require "rspec/core/rake_task"
 
 # Setup the necessary gems, specified in the gemspec.
 require 'bundler'
@@ -11,6 +12,9 @@ rescue Bundler::BundlerError => e
   exit e.status_code
 end
 
+# Run all the specs.
+RSpec::Core::RakeTask.new(:spec)
+
 # Compile task.
 # Rake::ExtensionTask.new do |ext|
 #     ext.name = 'measurable'          

data/lib/measurable.rb

@@ -1,55 +1,47 @@
 require 'measurable/version.rb'
 
 # Distance measures.
-reqiore 'measurable/euclidean'
+require 'measurable/euclidean'
 require 'measurable/cosine'
-require 'measurable/tanimoto_coefficient'
+require 'measurable/tanimoto'
 require 'measurable/jaccard'
 require 'measurable/haversine'
+require 'measurable/maxmin'
 
 module Measurable
   # PI = 3.1415926535
   RAD_PER_DEG = 0.017453293  #  PI/180
-
-  # http://en.wikipedia.org/wiki/Intersection_(set_theory)
-  def intersection(u, v)
-    (u & v)
-  end
-
-  # http://en.wikipedia.org/wiki/Union_(set_theory)
-  def union(u, v)
-    (u + v).uniq
-  end
-
-  def binary_union(u, v)
-    unions = []
-    u.each_with_index do |n, index|
-      if n == 1 || v[index] == 1
-        unions << 1
-      else
-        unions << 0
+  class << self
+    def binary_union(u, v)
+      unions = []
+      u.each_with_index do |n, index|
+        if n == 1 || v[index] == 1
+          unions << 1
+        else
+          unions << 0
+        end
       end
-    end
 
-    unions
-  end
+      unions
+    end
 
-  def binary_intersection(u, v)
-    intersects = []
-    u.each_with_index do |n, index|
-      if n == 1 && v[index] == 1
-        intersects << 1
-      else
-        intersects << 0
+    def binary_intersection(u, v)
+      intersects = []
+      u.each_with_index do |n, index|
+        if n == 1 && v[index] == 1
+          intersects << 1
+        else
+          intersects << 0
+        end
       end
-    end
 
-    intersects
-  end
+      intersects
+    end
 
-  # 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
+    # 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
 end

data/lib/measurable/cosine.rb

@@ -1,8 +1,10 @@
 module Measurable
-  def cosine(u, v)
-    dot_product = dot(u, v)
-    normalization = self.euclidean_normalize * other.euclidean_normalize
+  class << self
+    def cosine(u, v)
+      dot_product = dot(u, v)
+      normalization = self.euclidean_normalize * other.euclidean_normalize
 
-    handle_nan(dot_product / normalization)
+      handle_nan(dot_product / normalization)
+    end
   end
 end

data/lib/measurable/euclidean.rb

@@ -1,17 +1,40 @@
 module Measurable
-  def euclidean(u, v)
-    sum = 0.0
+  class << self
+    # Add documentation here!
+    def euclidean(u, v = nil)
+      # If the second argument is nil, the method should return the norm of
+      # vector u. For this, we need the distance between u and the origin.
+      if v.nil?
+        v = Array.new(u.size, 0)
+      end
+      
+      # We could make it work with vector of different sizes because of #zip
+      # but it's unreliable. It's better to just throw an exception.
+      # TODO: Change this to a more specific, custom-made exception.
+      raise ArgumentError if u.size != v.size
+      
+      sum = u.zip(v).reduce(0.0) do |acc, ary|
+        acc += (ary[0] - ary[-1])**2
+      end
     
-    u.zip(v).each do |ary|
-      sum += (ary.first - ary.last)**2
+      Math.sqrt(sum)
     end
-    
-    Math.sqrt(sum)
-  end
   
-  def euclidean_squared(u, v)
-    u.zip(v).reduce(0.0) do |acc, ary|
-      acc += (ary.first - ary.last)**2
+    def euclidean_squared(u, v = nil)
+      # If the second argument is nil, the method should return the norm of
+      # vector u. For this, we need the distance between u and the origin.
+      if v.nil?
+        v = Array.new(u.size, 0)
+      end
+      
+      # We could make it work with vector of different sizes because of #zip
+      # but it's unreliable. It's better to just throw an exception.
+      # TODO: Change this to a more specific, custom-made exception.
+      raise ArgumentError if u.size != v.size
+      
+      u.zip(v).reduce(0.0) do |acc, ary|
+        acc += (ary[0] - ary[-1])**2
+      end
     end
   end
 end

data/lib/measurable/haversine.rb

@@ -23,22 +23,24 @@ module Measurable
     :meters => R_KM * 1000
   }
 
-  def haversine(u, v, um = :meters)
-    dlon = u[1] - v[1]
-    dlat = u[0] - v[0]
+  class << self
+    def haversine(u, v, um = :meters)
+      dlon = u[1] - v[1]
+      dlat = u[0] - v[0]
 
-    dlon_rad = dlon * RAD_PER_DEG 
-    dlat_rad = dlat * RAD_PER_DEG
+      dlon_rad = dlon * RAD_PER_DEG 
+      dlat_rad = dlat * RAD_PER_DEG
 
-    lat1_rad = v[0] * RAD_PER_DEG
-    lon1_rad = v[1] * RAD_PER_DEG
+      lat1_rad = v[0] * RAD_PER_DEG
+      lon1_rad = v[1] * RAD_PER_DEG
 
-    lat2_rad = u[0] * RAD_PER_DEG
-    lon2_rad = u[1] * RAD_PER_DEG
+      lat2_rad = u[0] * RAD_PER_DEG
+      lon2_rad = u[1] * RAD_PER_DEG
 
-    a = (Math.sin(dlat_rad/2))**2 + Math.cos(lat1_rad) * Math.cos(lat2_rad) * (Math.sin(dlon_rad/2))**2
-    c = 2 * Math.atan2( Math.sqrt(a), Math.sqrt(1-a))
+      a = (Math.sin(dlat_rad/2))**2 + Math.cos(lat1_rad) * Math.cos(lat2_rad) * (Math.sin(dlon_rad/2))**2
+      c = 2 * Math.atan2( Math.sqrt(a), Math.sqrt(1-a))
 
-    R[um] * c
+      R[um] * c
+    end
   end
 end

data/lib/measurable/jaccard.rb

@@ -1,26 +1,26 @@
 # http://en.wikipedia.org/wiki/Jaccard_coefficient
 module Measurable
+  class << self
+    def jaccard(u, v)
+      1 - jaccard_index(u, v)
+    end
   
-  def jaccard(u, v)
-    1 - jaccard_index(u, v)
-  end
-  
-  def jaccard_index(u, v)
-    union = (u + v).uniq.size.to_f
-    i = intersection(u, v).size.to_f
+    def jaccard_index(u, v)
+      union = (u | v).size.to_f
+      intersection = (u & v).size.to_f
         
-    i / union
-  end
+      intersection / union
+    end
   
-  def binary_jaccard(u, v)
-    1 - binary_jaccard_index(u, v)
-  end
+    def binary_jaccard(u, v)
+      1 - binary_jaccard_index(u, v)
+    end
   
-  def binary_jaccard_index(u, v)
-    i = binary_intersection(u, v).delete_if {|x| x == 0}.size.to_f
-    union = binary_union(u, v).delete_if {|x| x == 0}.size.to_f
+    def binary_jaccard_index(u, v)
+      intersection = binary_intersection(u, v).delete_if {|x| x == 0}.size.to_f
+      union = binary_union(u, v).delete_if {|x| x == 0}.size.to_f
     
-    i / union
+      intersection / union
+    end
   end
-  
 end

data/lib/measurable/maxmin.rb

@@ -0,0 +1,13 @@
+module Measurable
+  class << self
+    def maxmin(u, v)
+      sum_min, sum_max = u.zip(v).reduce([0.0, 0.0]) do |acc, attributes|
+        acc[0] += attributes.min
+        acc[-1] += attributes.max
+        acc
+      end
+    
+      sum_min / sum_max
+    end
+  end
+end

data/lib/measurable/tanimoto.rb

@@ -1,9 +1,11 @@
 # http://en.wikipedia.org/wiki/Jaccard_index#Tanimoto_coefficient_.28extended_Jaccard_coefficient.29
 module Measurable
-  def tanimoto(u, v)
-    dot = dot(u, v).to_f
-    result = dot / (u.sum_of_squares + v.sum_of_squares - dot).to_f
+  class << self
+    def tanimoto(u, v)
+      dot = dot(u, v).to_f
+      result = dot / (u.sum_of_squares + v.sum_of_squares - dot).to_f
     
-    handle_nan(result)
+      handle_nan(result)
+    end
   end
 end

data/lib/measurable/version.rb

@@ -1,3 +1,3 @@
 module Measurable
-  VERSION = "0.0.3"
+  VERSION = "0.0.4"
 end

data/spec/measurable_spec.rb

@@ -1,35 +1,77 @@
 describe Measurable do
 
-  let(:u) { [1, 3, 16] }
-  let(:v) { [1, 4, 16] }
-  let(:w) { [4, 5, 6] }
-
-  describe "Euclidean distance" do
-    it "accepts two arguments" do
-      expect { Measurable::euclidean(:u) }.to raise_error(ArgumentError)
-      expect { Measurable::euclidean(:u, :v) }.to_not raise_error(ArgumentError)
-      expect { Measurable::euclidean(:u, :v, :w) }.to raise_error(ArgumentError)
+  describe "Binary union" do
+    
+  end
+
+  describe "Binary intersection" do
+    
+  end
+
+  describe "Euclidean" do
+    
+    before :all do
+      @u = [1, 3, 16]
+      @v = [1, 4, 16]
+      @w = [4, 5, 6]
     end
     
-    it "accepts one argument and returns the vector's norm"
+    context "Distance" do
+      it "accepts two arguments" do
+        expect { Measurable.euclidean(@u, @v) }.to_not raise_error
+        expect { Measurable.euclidean(@u, @v, @w) }.to raise_error(ArgumentError)
+      end
+    
+      it "accepts one argument and returns the vector's norm" do
+        # Remember that 3^2 + 4^2 = 5^2.
+        Measurable.euclidean([3, 4]).should == 5
+      end
             
-    it "should be symmetric"
+      it "should be symmetric" do
+        Measurable.euclidean(@u, @v).should == Measurable.euclidean(@v, @u)
+      end
 
-    it "should return the correct value" do
-      Measurable::euclidean(:u, :u).should == 0
-      euclidean(:u, :v).should == 1
+      it "should return the correct value" do
+        Measurable.euclidean(@u, @u).should == 0
+        Measurable.euclidean(@u, @v).should == 1
+      end
+    
+      it "shouldn't work with vectors of different length" do
+        expect { Measurable.euclidean(@u, [2, 2, 2, 2]) }.to raise_error
+      end
     end
     
-    it "shouldn't work with vectors of different length" do
-      expect { Measurable::euclidean(:u, [2, 2, 2, 2]) }.to raise_error(DiffLengthError)
+    context "Squared Distance" do
+      it "accepts two arguments" do
+        expect { Measurable.euclidean_squared(@u, @v) }.to_not raise_error
+        expect { Measurable.euclidean_squared(@u, @v, @w) }.to raise_error(ArgumentError)
+      end
+    
+      it "accepts one argument and returns the vector's norm" do
+        # Remember that 3^2 + 4^2 = 5^2.
+        Measurable.euclidean_squared([3, 4]).should == 25
+      end
+            
+      it "should be symmetric" do
+        x = Measurable.euclidean_squared(@u, @v)
+        y = Measurable.euclidean_squared(@v, @u)
+        
+        x.should == y
+      end
+
+      it "should return the correct value" do
+        Measurable.euclidean_squared(@u, @u).should == 0
+        Measurable.euclidean_squared(@u, @v).should == 1
+      end
+    
+      it "shouldn't work with vectors of different length" do
+        expect { Measurable.euclidean_squared(@u, [2, 2, 2, 2]) }.to raise_error
+      end    
     end
+    
   end
 
-  describe "Binary union" do
-
-  describe "Binary intersection" do
-
-  describe "Cosine similarity measure" do
+  describe "Cosine distance" do
     it "accepts two arguments"
     
     it "accepts one argument and returns the vector's norm"
@@ -55,7 +97,55 @@ describe Measurable do
     it "shouldn't work with vectors of different length"
   end
   
-  describe "Max-min similarity measure" do
+  describe "Tanimoto distance" do
+    it "accepts two arguments"
+    
+    it "accepts one argument and returns the vector's norm"
+        
+    it "should be symmetric"
+
+    it "should return the correct value"
+    
+    it "shouldn't work with vectors of different length"    
+  end
+
+  describe "Haversine distance" do
+    it "accepts two arguments"
+    
+    it "accepts one argument and returns the vector's norm"
+        
+    it "should be symmetric"
+
+    it "should return the correct value"
+    
+    it "shouldn't work with vectors of different length"
+  end
+  
+  describe "Jaccard distance" do
+    it "accepts two arguments"
+    
+    it "accepts one argument and returns the vector's norm"
+        
+    it "should be symmetric"
+
+    it "should return the correct value"
+    
+    it "shouldn't work with vectors of different length"
+  end
+  
+  describe "Binary Jaccard distance" do
+    it "accepts two arguments"
+    
+    it "accepts one argument and returns the vector's norm"
+        
+    it "should be symmetric"
+
+    it "should return the correct value"
+    
+    it "shouldn't work with vectors of different length"
+  end
+  
+  describe "Max-min distance" do
     it "accepts two arguments"
     
     it "accepts one argument and returns the vector's norm"

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: measurable
 version: !ruby/object:Gem::Version
-  version: 0.0.3
+  version: 0.0.4
 platform: ruby
 authors:
 - Carlos Agarie
@@ -70,6 +70,7 @@ files:
 - lib/measurable/euclidean.rb
 - lib/measurable/haversine.rb
 - lib/measurable/jaccard.rb
+- lib/measurable/maxmin.rb
 - lib/measurable/tanimoto.rb
 - lib/measurable/version.rb
 - measurable.gemspec