measurable 0.0.5 → 0.0.11

data/lib/measurable/tanimoto.rb

@@ -1,32 +1,51 @@
+require 'measurable/jaccard'
+
 module Measurable
+  module Tanimoto
+
+    # call-seq:
+    #     tanimoto(u, v) -> Float
+    #
+    # Tanimoto distance is a coefficient explicitly chosen such as to allow for
+    # two dissimilar specimens to be similar to a third one. This breaks the
+    # triangle inequality, thus this isn't a metric.
+    #
+    # More information and references on this are needed. It's left here mostly
+    # as a piece of curiosity.
+    #
+    # See: # http://en.wikipedia.org/wiki/Jaccard_index#Tanimoto.27s_Definitions_of_Similarity_and_Distance
+    #
+    # Arguments:
+    # - +u+ -> An array of Numeric objects.
+    # - +v+ -> An array of Numeric objects.
+    # Returns:
+    # - A measure of the similarity between +u+ and +v+.
+    # Raises:
+    # - +ArgumentError+ -> The sizes of +u+ and +v+ don't match.
+    def tanimoto(u, v)
+      # TODO: Change this to a more specific, custom-made exception.
+      raise ArgumentError if u.size != v.size
 
-  # Tanimoto similarity is the same as Jaccard similarity.
-  alias :tanimoto_similarity :jaccard
+      -Math.log2(jaccard_index(u, v))
+    end
 
-  # call-seq:
-  #     tanimoto(u, v) -> Float
-  #
-  # Tanimoto distance is a coefficient explicitly chosen such as to allow for
-  # two dissimilar specimens to be similar to a third one. This breaks the
-  # triangle inequality, thus this isn't a metric.
-  #
-  # More information and references on this are needed. It's left here mostly
-  # as a piece of curiosity.
-  #
-  # See: # http://en.wikipedia.org/wiki/Jaccard_index#Tanimoto.27s_Definitions_of_Similarity_and_Distance
-  #
-  # * *Arguments* :
-  #   - +u+ -> An array of Numeric objects.
-  #   - +v+ -> An array of Numeric objects.
-  # * *Returns* :
-  #   - A measure of the similarity between +u+ and +v+.
-  # * *Raises* :
-  #   - +ArgumentError+ -> The sizes of +u+ and +v+ doesn't match.
-  #
-  def tanimoto(u, v)
-    # TODO: Change this to a more specific, custom-made exception.
-    raise ArgumentError if u.size != v.size
+    def self.extended(base) # :nodoc:
+      # Tanimoto similarity is the same as Jaccard similarity.
+      base.instance_eval do
+        extend Measurable::Jaccard
+        alias :tanimoto_similarity :jaccard
+      end
+      super
+    end
 
-    -Math.log2(jaccard_index(u, v))
+    def self.included(base) # :nodoc:
+      base.class_eval do
+        include Measurable::Jaccard
+        alias :tanimoto_similarity :jaccard
+      end
+      super
+    end
   end
-end
+
+  extend Measurable::Tanimoto
+end

data/lib/measurable/version.rb

@@ -1,3 +1,3 @@
 module Measurable
-  VERSION = "0.0.5" # :nodoc:
-end
+  VERSION = "0.0.11" # :nodoc:
+end

data/measurable.gemspec

@@ -4,12 +4,13 @@ require 'measurable/version'
 require 'date'
 
 Gem::Specification.new do |gem|
-  gem.name = "measurable"  
+  gem.name = "measurable"
   gem.version = Measurable::VERSION
   gem.date = Date.today.to_s
+  gem.license = "MIT"
   gem.summary = %Q{A Ruby gem with a lot of distance measures for your projects.}
   gem.description = %Q{A Ruby gem with a lot of distance measures for your projects.}
-  
+
   gem.authors = ["Carlos Agarie"]
   gem.email = "carlos.agarie@gmail.com"
   gem.homepage = "http://github.com/agarie/measurable"
@@ -23,6 +24,7 @@ Gem::Specification.new do |gem|
   gem.required_ruby_version = '>= 1.9.3'
 
   gem.add_development_dependency 'bundler'
-  gem.add_development_dependency 'rake', '~> 0.9'
-  gem.add_development_dependency 'rspec', '~> 2.9.0'
+  gem.add_development_dependency 'rake', '>= 12.3.3'
+  gem.add_development_dependency 'rdoc', '>= 6.0.0'
+  gem.add_development_dependency 'rspec', '~> 3.2'
 end

data/spec/chebyshev_spec.rb

@@ -0,0 +1,48 @@
+describe "Chebyshev distance" do
+
+  before :all do
+    @u = [1.4, 2.5, 5.8]
+    @v = [2.2, 3.6, 2.7]
+    @w = [4.1, 5.7, 1.2]
+  end
+
+  it "accepts two arguments" do
+    expect { Measurable.chebyshev(@u, @v) }.to_not raise_error
+    expect { Measurable.chebyshev(@u, @v, @w) }.to raise_error(ArgumentError)
+  end
+
+  it "should be symmetric" do
+    x = Measurable.chebyshev(@u, @v)
+    y = Measurable.chebyshev(@v, @u)
+
+    expect(x).to be_within(TOLERANCE).of(y)
+  end
+
+  it "should return the correct value" do
+    x = Measurable.chebyshev(@u, @v)
+    expect(x).to be_within(TOLERANCE).of(3.1)
+  end
+
+  it "shouldn't work with vectors of different length" do
+    expect { Measurable.chebyshev(@u, [1, 3, 5, 7]) }.to raise_error(ArgumentError)
+  end
+
+  it "can be extended separately" do
+    klass = Class.new do
+      extend Measurable::Chebyshev
+    end
+
+    x = klass.chebyshev(@u, @v)
+    expect(x).to be_within(TOLERANCE).of(3.1)
+  end
+
+  it "can be included separately" do
+    klass = Class.new do
+      include Measurable::Chebyshev
+    end
+
+    x = klass.new.chebyshev(@u, @v)
+    expect(x).to be_within(TOLERANCE).of(3.1)
+  end
+
+end

data/spec/cosine_spec.rb

@@ -1,29 +1,77 @@
-describe "Cosine distance" do
-  
-  before :all do
-    @u = [1, 2]
-    @v = [2, 3]
-    @w = [4, 5]
-  end
-  
-  it "accepts two arguments" do
-    expect { Measurable.cosine(@u, @v) }.to_not raise_error
-    expect { Measurable.cosine(@u, @v, @w) }.to raise_error(ArgumentError)
-  end
-  
-  it "should be symmetric" do
-    x = Measurable.cosine(@u, @v)
-    y = Measurable.cosine(@v, @u)
+describe "Cosine" do
 
-    x.should be_within(TOLERANCE).of(y)
-  end
+  context "Similarity" do
+    before :all do
+      @u = [1, 2]
+      @v = [2, 3]
+      @w = [4, 5]
+    end
+
+    it "accepts two arguments" do
+      expect { Measurable.cosine_similarity(@u, @v) }.to_not raise_error
+      expect { Measurable.cosine_similarity(@u, @v, @w) }.to raise_error(ArgumentError)
+    end
+
+    it "should be symmetric" do
+      x = Measurable.cosine_similarity(@u, @v)
+      y = Measurable.cosine_similarity(@v, @u)
+
+      x.should be_within(TOLERANCE).of(y)
+    end
+
+    it "should return the correct value" do
+      x = Measurable.cosine_similarity(@u, @v)
+      x.should be_within(TOLERANCE).of(0.992277877)
+    end
 
-  it "should return the correct value" do
-    x = Measurable.cosine(@u, @v)
-    x.should be_within(TOLERANCE).of(0.992277877)
+    it "shouldn't work with vectors of different length" do
+      expect { Measurable.cosine_similarity(@u, [1, 3, 5, 7]) }.to raise_error(ArgumentError)
+    end
+
+    it "can be extended separately" do
+      klass = Class.new do
+        extend Measurable::Cosine
+      end
+      x = klass.cosine_similarity(@u, @v)
+      x.should be_within(TOLERANCE).of(0.992277877)
+    end
+
+    it "can be extended separately" do
+      klass = Class.new do
+        include Measurable::Cosine
+      end
+      x = klass.new.cosine_similarity(@u, @v)
+      x.should be_within(TOLERANCE).of(0.992277877)
+    end
   end
 
-  it "shouldn't work with vectors of different length" do
-    expect { Measurable.cosine(@u, [1, 3, 5, 7]) }.to raise_error
+  context "Distance" do
+    before :all do
+      @u = [1, 2]
+      @v = [2, 3]
+      @w = [4, 5]
+    end
+
+    it "accepts two arguments" do
+      expect { Measurable.cosine_distance(@u, @v) }.to_not raise_error
+      expect { Measurable.cosine_distance(@u, @v, @w) }.to raise_error(ArgumentError)
+    end
+
+    it "should be symmetric" do
+      x = Measurable.cosine_distance(@u, @v)
+      y = Measurable.cosine_distance(@v, @u)
+
+      x.should be_within(TOLERANCE).of(y)
+    end
+
+    it "should return the correct value" do
+      x = Measurable.cosine_distance(@u, @v)
+      # TODO: Use a real example.
+      x.should be_within(TOLERANCE).of(1.0 - 0.992277877)
+    end
+
+    it "shouldn't work with vectors of different length" do
+      expect { Measurable.cosine_distance(@u, [1, 3, 5, 7]) }.to raise_error(ArgumentError)
+    end
   end
-end
+end

data/spec/euclidean_spec.rb

@@ -1,22 +1,22 @@
 describe "Euclidean" do
-  
+
   before :all do
     @u = [1, 3, 16]
     @v = [1, 4, 16]
     @w = [4, 5, 6]
   end
-  
+
   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" do
       Measurable.euclidean(@u, @v).should == Measurable.euclidean(@v, @u)
     end
@@ -25,27 +25,43 @@ describe "Euclidean" 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
+      expect { Measurable.euclidean(@u, [2, 2, 2, 2]) }.to raise_error(ArgumentError)
+    end
+
+    it "can be extended separately" do
+      klass = Class.new do
+        extend Measurable::Euclidean
+      end
+
+      klass.euclidean([3, 4]).should == 5
+    end
+
+    it "can be included separately" do
+      klass = Class.new do
+        include Measurable::Euclidean
+      end
+
+      klass.new.euclidean([3, 4]).should == 5
     end
   end
-  
+
   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
 
@@ -53,9 +69,9 @@ describe "Euclidean" 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    
+      expect { Measurable.euclidean_squared(@u, [2, 2, 2, 2]) }.to raise_error(ArgumentError)
+    end
   end
-end
+end

data/spec/hamming_spec.rb

@@ -0,0 +1,46 @@
+describe "Hamming distance" do
+
+  before :all do
+    @u = "Hi, I'm a test string!"
+    @v = "Hello, not a test omg."
+    @w = "Hey there, a test wtf!"
+  end
+
+  it "accepts two arguments" do
+    expect { Measurable.hamming(@u, @v) }.to_not raise_error
+    expect { Measurable.hamming(@u, @v, @w) }.to raise_error(ArgumentError)
+  end
+
+  it "should be symmetric" do
+    x = Measurable.hamming(@u, @v)
+    y = Measurable.hamming(@v, @u)
+
+    x.should be(y)
+  end
+
+  it "should return the correct value" do
+    x = Measurable.hamming(@u, @v)
+    x.should be(17)
+  end
+
+  it "shouldn't work with strings of different length" do
+    expect { Measurable.hamming(@u, "smallstring") }.to raise_error(ArgumentError)
+    expect { Measurable.hamming(@u, "largestring" * 20) }.to raise_error(ArgumentError)
+  end
+
+  it "can be extended separately" do
+    klass = Class.new do
+      extend Measurable::Hamming
+    end
+
+    klass.hamming(@u, @v).should == 17
+  end
+
+  it "can be included separately" do
+    klass = Class.new do
+      include Measurable::Hamming
+    end
+
+    klass.new.hamming(@u, @v).should == 17
+  end
+end

data/spec/haversine_spec.rb

@@ -32,6 +32,26 @@ describe "Haversine distance" do
   end
 
   it "should only work with [lat, long] vectors" do
-    expect { Measurable.haversine([2, 4], [1, 3, 5, 7]) }.to raise_error
+    expect { Measurable.haversine([2, 4], [1, 3, 5, 7]) }.to raise_error(ArgumentError)
   end
-end
+
+  it "can be extended seperately" do
+    klass = Class.new do
+      extend Measurable::Haversine
+    end
+
+    x = klass.haversine(@u, @v, :km)
+
+    x.should be_within(@haversine_tolerance).of(18533)
+  end
+
+  it "can be included seperately" do
+    klass = Class.new do
+      include Measurable::Haversine
+    end
+
+    x = klass.new.haversine(@u, @v, :km)
+
+    x.should be_within(@haversine_tolerance).of(18533)
+  end
+end

data/spec/jaccard_spec.rb

@@ -16,18 +16,39 @@ describe "Jaccard" do
       x = Measurable.jaccard_index(@u, @v)
       y = Measurable.jaccard_index(@v, @u)
 
-      x.should be_within(TOLERANCE).of(y)
+      expect(x).to be_within(TOLERANCE).of(y)
     end
 
     it "should return the correct value" do
       x = Measurable.jaccard_index(@u, @v)
 
-      x.should be_within(TOLERANCE).of(2.0 / 3.0)
+      expect(x).to be_within(TOLERANCE).of(1.0 / 2.0)
     end
 
-    it "shouldn't work with vectors of different length" do
-      expect { Measurable.jaccard_index(@u, [1, 2, 3, 4]) }.to raise_error
+    it "should work with vectors of different length" do
+      expect { Measurable.jaccard_index(@u, [1, 2, 3, 4]) }.to_not raise_error
     end
+
+    it "can be extended separately" do
+      klass = Class.new do
+        extend Measurable::Jaccard
+      end
+
+      x = klass.jaccard_index(@u, @v)
+
+      expect(x).to be_within(TOLERANCE).of(1.0 / 2.0)
+    end
+
+    it "can be included separately" do
+      klass = Class.new do
+        include Measurable::Jaccard
+      end
+
+      x = klass.new.jaccard_index(@u, @v)
+
+      expect(x).to be_within(TOLERANCE).of(1.0 / 2.0)
+    end
+
   end
 
   context "Distance" do
@@ -38,25 +59,25 @@ describe "Jaccard" do
     end
 
     it "accepts two arguments" do
-      expect { Measurable.jaccard(@u, @v) }.to_not raise_error
-      expect { Measurable.jaccard(@u, @v, @w) }.to raise_error(ArgumentError)
+      expect { Measurable.jaccard_dissimilarity(@u, @v) }.to_not raise_error
+      expect { Measurable.jaccard_dissimilarity(@u, @v, @w) }.to raise_error(ArgumentError)
     end
 
     it "should be symmetric" do
-      x = Measurable.jaccard(@u, @v)
-      y = Measurable.jaccard(@v, @u)
+      x = Measurable.jaccard_dissimilarity(@u, @v)
+      y = Measurable.jaccard_dissimilarity(@v, @u)
 
-      x.should be_within(TOLERANCE).of(y)
+      expect(x).to be_within(TOLERANCE).of(y)
     end
 
     it "should return the correct value" do
-      x = Measurable.jaccard(@u, @v)
+      x = Measurable.jaccard_dissimilarity(@u, @v)
 
-      x.should be_within(TOLERANCE).of(1.0 / 3.0)
+      expect(x).to be_within(TOLERANCE).of(1.0 / 2.0)
     end
 
-    it "shouldn't work with vectors of different length" do
-      expect { Measurable.jaccard(@u, [1, 2, 3, 4]) }.to raise_error
+    it "should work with vectors of different length" do
+      expect { Measurable.jaccard_dissimilarity(@u, [1, 2, 3, 4]) }.to_not raise_error
     end
   end
-end
+end