ml 0.1.0 → 0.1.1

data/VERSION

@@ -1 +1 @@
-0.1.0
+0.1.1

data/lib/data/generator.rb

@@ -1,98 +1,97 @@
 require 'matrix'
 
 module ML
-  # Generating sample points on 2D plane
-  class Generator2D
-    # Generate point from line
-    #
-    # @param [Array] [a,b,c] for ax+by+c=0
-    # @param [Number] x value
-    # @return [Array] 
-    def self.point_from_line coef, x
-      [x, (-coef[2]-(coef[0] * x))/coef[1]]
-    end
-
-    # Initialize a generator
-    #
-    # @param [Integer] x range
-    # @param [Integer] y range
-    def initialize x_range = 100, y_range = 100
-      @x_range = x_range
-      @y_range = y_range
-    end
+  module Data
+    # Generating sample points on 2D plane
+    class Generator2D
+      # Generate point from line
+      #
+      # @param [Array] coef [a,b,c] for ax+by+c=0
+      # @param [Number] x x value
+      # @return [Array] point
+      def self.point_from_line coef, x
+        [x, (-coef[2]-(coef[0] * x))/coef[1]]
+      end
 
-    # Generate two groups of points on 2d plain
-    #
-    # @param [Integer] the number of points of each set
-    # @param [Array] [a,b,c] for ax+by+c=0
-    # @return [Array] two sets of points
-    def points_2d points, coef = [-1.0, 1.0, 0.0]
-      result = []
-      # for each group
-      [1, -1].each do |grp|
-        result << []
+      # Initialize a generator
+      #
+      # @param [Integer] x_range x range
+      # @param [Integer] y_range y range
+      def initialize x_range = 100, y_range = 100
+        @x_range = x_range
+        @y_range = y_range
+      end
 
-        points.times do
-          while true
-            point = generate_point
-            prod = Matrix.column_vector(point).transpose * Matrix.column_vector(coef)
-            if (prod[0,0] <=> 0) == grp
-              result[-1] << point
-              break
+      # Generate two groups of points on 2d plain
+      #
+      # @param [Integer] points the number of points of each set
+      # @param [Array] coef [a,b,c] for ax+by+c=0
+      # @return [Hash] key: points, value: supervised value
+      def points_2d points, coef = [-1.0, 1.0, 0.0]
+        result = {}
+        # for each group
+        [1, -1].each do |grp|
+          points.times do
+            while true
+              point = generate_point
+              prod = Matrix.column_vector(point).transpose * Matrix.column_vector(coef)
+              if (prod[0,0] <=> 0) == grp
+                result[point] = grp
+                break
+              end
             end
           end
         end
+        result
       end
-      result
-    end
 
-  private
-    def generate_point
-      [@x_range * rand, @y_range * rand, 1.0]
-    end
-  end
-
-  # General generator for n-dimentional space
-  class Generator
-    # Initial generator
-    #
-    # @param [Integer] dimension
-    def initialize dim
-      @dim = dim
+    private
+      def generate_point
+        [@x_range * rand, @y_range * rand, 1.0]
+      end
     end
 
-    # Generate two groups of points
-    #
-    # @param [Integer] the number of points of each set
-    # @param [Array] array of the size of dimension to specify the hyper plane
-    # @return [Array] two sets of points
-    def points points, coef
-      result = []
-      # for each group
-      [1, -1].each do |grp|
-        result << []
+    # General generator for n-dimentional space
+    class Generator
+      # Initial generator
+      #
+      # @param [Integer] dim dimension
+      def initialize dim
+        @dim = dim
+      end
 
-        points.times do
-          while true
-            point = Generator.generate_vector(@dim)
-            prod = Matrix.column_vector(point).transpose * Matrix.column_vector(coef)
-            if (prod[0,0] <=> 0) == grp
-              result[-1] << point
-              break
+      # Generate two groups of points
+      #
+      # @param [Integer] points the number of points of each set
+      # @param [Array] coef array of the size of dimension to specify the hyper plane
+      # @return [Hash] key: points, value: supervised value
+      def points points, coef
+        result = {}
+        # for each group
+        [1, -1].each do |grp|
+          points.times do
+            while true
+              point = Generator.generate_vector(@dim, 100)
+              prod = Matrix.column_vector(point).transpose * Matrix.column_vector(coef)
+              if (prod[0,0] <=> 0) == grp
+                result[point] = grp
+                break
+              end
             end
           end
         end
+        result
       end
-      result
-    end
 
-    # Generating a random vector
-    #
-    # @param [Integer] the dimension of the vector
-    # @return [Array] random vector
-    def self.generate_vector dim
-      result = Array.new(dim) { rand - 0.5 } 
-      result << 1.0
+      # Generating a random vector
+      #
+      # @param [Integer] dim the dimension of the vector
+      # @param [Integer] scale the scale of each component
+      # @return [Array] random vector
+      def self.generate_vector dim, scale = 1
+        result = Array.new(dim) { (rand - 0.5) * scale } 
+        result << 1.0
+      end
     end
   end
 end

data/lib/data/parser.rb

@@ -1,27 +1,31 @@
 module ML
-  # Parser for traing/testing data
-  class Parser
-    # Parse the vector file with supervised result
-    #
-    # @return [Hash] map from data to supervised result
-    def parse_supervised filename
-      result = {}
-      lines = IO.readlines(filename)
+  module Data
+    # Parser for traing/testing data
+    class Parser
+      # Parse the vector file with supervised result
+      #
+      # @param [String] filename filename of the input data
+      # @return [Hash] map from data to supervised result
+      def parse_supervised filename
+        result = {}
+        lines = IO.readlines(filename)
 
-      lines.each do |line|
-        splitted = line.split.map(&:to_f)
-        result[splitted[1..-1] + [1.0]] = splitted[0]
-      end
+        lines.each do |line|
+          splitted = line.split.map(&:to_f)
+          result[splitted[1..-1] + [1.0]] = splitted[0]
+        end
 
-      result
-    end
+        result
+      end
 
-    # Parse the vector file
-    #
-    # @return [Array] array of vectors
-    def parse_unsupervised filename
-      lines = IO.readlines(filename)
-      lines.map {|line| line.split.map(&:to_f) }
+      # Parse the vector file
+      #
+      # @param [String] filename filename of the input data
+      # @return [Array] array of vectors
+      def parse_unsupervised filename
+        lines = IO.readlines(filename)
+        lines.map {|line| line.split.map(&:to_f) }
+      end
     end
   end
 end

data/lib/data/plotter.rb

@@ -1,77 +1,80 @@
 require 'rubyvis'
 
 module ML
-  # Plotting the data to svg
-  class Plotter
-    # Initializer of plotter
-    #
-    # @param [Integer] x value range
-    # @param [Integer] y value range
-    # @param [Integer] x plot size
-    # @param [Integer] y plot size
-    def initialize x_range = 100, y_range = 100, x_size = 100, y_size = 100
-      @x_range = x_range
-      @y_range = y_range
-      @x_size = x_size
-      @y_size = y_size
+  module Data
+    # Plotting the data to svg
+    class Plotter
+      # Initializer of plotter
+      #
+      # @param [Integer] x_range x value range
+      # @param [Integer] y_range y value range
+      # @param [Integer] x_size x plot size
+      # @param [Integer] y_size y plot size
+      def initialize x_range = 100, y_range = 100, x_size = 100, y_size = 100
+        @x_range = x_range
+        @y_range = y_range
+        @x_size = x_size
+        @y_size = y_size
 
-      @x = pv.Scale.linear(0, @x_range).range(0, @x_size)
-      @y = pv.Scale.linear(0, @y_range).range(0, @y_size)
+        @x = pv.Scale.linear(0, @x_range).range(0, @x_size)
+        @y = pv.Scale.linear(0, @y_range).range(0, @y_size)
 
-      @vis = pv.Panel.new.width(@x_size).height(@y_size)
-               .bottom(20).left(20).right(10).top(5)
-      
-      @vis.add(pv.Rule).data(@y.ticks()).bottom(@y)
-          .stroke_style(lambda {|d| d!=0 ? "#eee" : "#000"})
-          .anchor("left").add(pv.Label)
-          .visible(lambda {|d| d > 0 and d < x_range})
-          .text(@y.tick_format)
- 
-      @vis.add(pv.Rule).data(@x.ticks()).left(@x)
-          .stroke_style(lambda {|d| d!=0 ? "#eee" : "#000"})
-          .anchor("bottom").add(pv.Label)
-          .visible(lambda {|d| d > 0 and d < y_range})
-          .text(@x.tick_format)
+        @vis = pv.Panel.new.width(@x_size).height(@y_size)
+                 .bottom(20).left(20).right(10).top(5)
+        
+        @vis.add(pv.Rule).data(@y.ticks()).bottom(@y)
+            .stroke_style(lambda {|d| d!=0 ? "#eee" : "#000"})
+            .anchor("left").add(pv.Label)
+            .visible(lambda {|d| d > 0 and d < x_range})
+            .text(@y.tick_format)
+   
+        @vis.add(pv.Rule).data(@x.ticks()).left(@x)
+            .stroke_style(lambda {|d| d!=0 ? "#eee" : "#000"})
+            .anchor("bottom").add(pv.Label)
+            .visible(lambda {|d| d > 0 and d < y_range})
+            .text(@x.tick_format)
 
-      yield(self) if block_given?
-    end
+        yield(self) if block_given?
+      end
 
-    # Plotting points with shape and color
-    #
-    # @param [Array] points to plot
-    # @param [String] shape of the points
-    # @param [String] color of the points
-    def dot points, shape = "circle", color = "#000"
-      # FIXME: dirty hack for instance_exec
-      x = @x
-      y = @y
+      # Plotting points with shape and color
+      #
+      # @param [Array] points points to plot
+      # @param [String] shape shape of the points
+      # @param [String] color color of the points
+      def dot points, shape = "circle", color = "#000"
+        # FIXME: dirty hack for instance_exec
+        x = @x
+        y = @y
 
-      @vis.add(pv.Dot).data(points)         
-          .left(lambda {|d| x.scale(d[0])})
-          .bottom(lambda {|d| y.scale(d[1])})
-          .shape(shape)
-          .stroke_style(color)
-    end
+        @vis.add(pv.Dot).data(points)         
+            .left(lambda {|d| x.scale(d[0])})
+            .bottom(lambda {|d| y.scale(d[1])})
+            .shape(shape)
+            .stroke_style(color)
+      end
 
-    # Plotting line with color
-    #
-    # @param [Array] 2 points which represents line
-    def line points, color = "#000"
-      x = @x
-      y = @y
+      # Plotting line with color
+      #
+      # @param [Array] points 2 points which represents line
+      # @param [String] color color of the line
+      def line points, color = "#000"
+        x = @x
+        y = @y
 
-      @vis.add(pv.Line).data(points)
-          .left(lambda {|d| x.scale(d[0])})
-          .bottom(lambda {|d| y.scale(d[1])})
-          .stroke_style(color)
-    end
+        @vis.add(pv.Line).data(points)
+            .left(lambda {|d| x.scale(d[0])})
+            .bottom(lambda {|d| y.scale(d[1])})
+            .stroke_style(color)
+      end
 
-    # Convert to svg
-    #
-    # @return [String] svg plot
-    def to_svg
-      @vis.render
-      @vis.to_svg
+      # Convert to svg
+      #
+      # @return [String] svg plot
+      def to_svg
+        @vis.render
+        @vis.to_svg
+      end
     end
   end
 end

data/lib/method/perceptron.rb

@@ -1,66 +1,67 @@
 require 'matrix'
 
 module ML
-  # Implementation of Perceptron Learning Algorithm
-  class PerceptronLearner
-    # Initialize a perceptron learner
-    #
-    # @param [Integer] the number of dimension
-    def initialize dim
-      @dim = dim
-      @w = Matrix.column_vector(Array.new(dim + 1, 0))
-    end
+  module Learner
+    # Implementation of Perceptron Learning Algorithm
+    class PerceptronLearner
+      # Initialize a perceptron learner
+      #
+      # @param [Integer] dim the number of dimension
+      def initialize dim
+        @dim = dim
+        @w = Matrix.column_vector(Array.new(dim + 1, 0))
+      end
 
-    # Train with supervised data
-    #
-    # @param [Hash] supervised input data (mapping from array to integer)
-    # @return [PerceptronLearner] self object 
-    def train! data
-      @update = 0
-      while true
-        misclassified = false
+      # Train with supervised data
+      #
+      # @param [Hash] data supervised input data (mapping from array to integer)
+      def train! data
+        @update = 0
+        while true
+          misclassified = false
 
-        for dat, result in data
-          aug_data = Matrix.column_vector(dat)
+          for dat, result in data
+            aug_data = Matrix.column_vector(dat)
 
-          if classify(aug_data) != result
-            misclassified = true
+            if classify(aug_data) != result
+              misclassified = true
 
-            @w = @w + result * aug_data
-            @update += 1
-            break
+              @w = @w + result * aug_data
+              @update += 1
+              break
+            end
           end
-        end
 
-        break unless misclassified
+          break unless misclassified
+        end
       end
-    end
 
-    # The final coefficient of the line
-    #
-    # @return [line] [a,b,c] for ax+by+c=0
-    def line
-      @w.column(0).to_a
-    end
+      # The final coefficient of the line
+      #
+      # @return [Array] [a,b,c] for ax+by+c=0
+      def line
+        @w.column(0).to_a
+      end
 
-    # The number for updates
-    #
-    # @return [Integer] update count
-    def update_count
-      @update
-    end
+      # The number for updates
+      #
+      # @return [Integer] update count
+      def update_count
+        @update
+      end
 
-    # Predict certain data
-    #
-    # @param [Array] data in question
-    # @param [Integer] prediction
-    def predict data
-      classify Matrix.column_vector(data + [1.0])  
-    end
+      # Predict certain data
+      #
+      # @param [Array] data data in question
+      # @return [Integer] prediction
+      def predict data
+        classify Matrix.column_vector(data + [1.0])  
+      end
 
-  private
-    def classify data
-      (@w.transpose * data)[0,0] <=> 0
+    private
+      def classify data
+        (@w.transpose * data)[0,0] <=> 0
+      end
     end
   end
 end

data/lib/ml.rb

@@ -7,7 +7,15 @@ require 'data/parser'
 
 require 'method/perceptron'
 
+# Top namespace for machine learning algorithms
 module ML
+  # Data processing module
+  module Data
+  end
+
+  # Learning algorithms
+  module Learner
+  end
 end
 
 MachingLearning = ML

data/ml.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = "ml"
-  s.version = "0.1.0"
+  s.version = "0.1.1"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Andrew Liu"]
-  s.date = "2011-10-03"
+  s.date = "2011-10-04"
   s.description = "Machine learning library in Ruby"
   s.email = "andrewliu33@gmail.com"
   s.extra_rdoc_files = [
@@ -29,7 +29,8 @@ Gem::Specification.new do |s|
     "lib/method/perceptron.rb",
     "lib/ml.rb",
     "ml.gemspec",
-    "spec/learning_spec.rb",
+    "spec/data_spec.rb",
+    "spec/learner_spec.rb",
     "spec/spec_helper.rb"
   ]
   s.homepage = "http://github.com/eggegg/ml"

data/spec/data_spec.rb

@@ -0,0 +1,24 @@
+require 'spec_helper'
+
+describe "Data" do
+  it "should plot a single plot" do
+    graph = ML::Data::Plotter.new(100, 100, 100, 100) do |p|
+      p.dot [[10,10], [20,20]]
+      p.line [[0,50], [100,50]]
+    end
+
+    graph.to_svg.kind_of?(String).should == true
+  end
+
+  it "should generate points" do
+    hyperplane = ML::Data::Generator.generate_vector(4)
+    hyperplane.size.should == 5
+
+    generator = ML::Data::Generator.new(4)
+    points = generator.points(10, hyperplane)
+
+    points.kind_of?(Hash).should == true
+    points.first.kind_of?(Array).should == true
+    points.first.first.size.should == 5
+  end
+end

data/spec/learner_spec.rb

@@ -0,0 +1,35 @@
+require 'spec_helper'
+
+describe "Learner" do
+  describe "Perceptron Learner" do
+    it "should run perceptron learning in 2d" do
+      learner = ML::Learner::PerceptronLearner.new(2)
+
+      generator = ML::Data::Generator2D.new
+      data = generator.points_2d(10)
+
+      learner.train! data
+
+      line = learner.line
+      line.should.kind_of?(Array).should == true
+      line.size.should == 3
+
+      learner.update_count.should > 0
+    end
+
+    it "should run perceptron learning in hyperspace" do
+      learner = ML::Learner::PerceptronLearner.new(4)
+
+      generator = ML::Data::Generator.new(4)
+      data = generator.points(10, ML::Data::Generator.generate_vector(4))
+
+      learner.train! data
+
+      line = learner.line
+      line.should.kind_of?(Array).should == true
+      line.size.should == 5
+
+      learner.update_count.should > 0
+    end
+  end
+end

data/spec/spec_helper.rb

@@ -11,6 +11,6 @@ require 'bacon'
 
 $LOAD_PATH.unshift(File.dirname(__FILE__))
 $LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
-require 'learning'
+require 'ml'
 
 Bacon.summary_on_exit

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ml
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
   prerelease: 
 platform: ruby
 authors:
@@ -9,11 +9,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2011-10-03 00:00:00.000000000Z
+date: 2011-10-04 00:00:00.000000000Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rubyvis
-  requirement: &2154897000 !ruby/object:Gem::Requirement
+  requirement: &2158153020 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -21,10 +21,10 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *2154897000
+  version_requirements: *2158153020
 - !ruby/object:Gem::Dependency
   name: nokogiri
-  requirement: &2154896440 !ruby/object:Gem::Requirement
+  requirement: &2158152400 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -32,10 +32,10 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *2154896440
+  version_requirements: *2158152400
 - !ruby/object:Gem::Dependency
   name: bacon
-  requirement: &2154895820 !ruby/object:Gem::Requirement
+  requirement: &2158151760 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -43,10 +43,10 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *2154895820
+  version_requirements: *2158151760
 - !ruby/object:Gem::Dependency
   name: yard
-  requirement: &2154895300 !ruby/object:Gem::Requirement
+  requirement: &2158151260 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -54,10 +54,10 @@ dependencies:
         version: 0.6.0
   type: :development
   prerelease: false
-  version_requirements: *2154895300
+  version_requirements: *2158151260
 - !ruby/object:Gem::Dependency
   name: bundler
-  requirement: &2154894780 !ruby/object:Gem::Requirement
+  requirement: &2158150660 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -65,10 +65,10 @@ dependencies:
         version: 1.0.0
   type: :development
   prerelease: false
-  version_requirements: *2154894780
+  version_requirements: *2158150660
 - !ruby/object:Gem::Dependency
   name: jeweler
-  requirement: &2154894220 !ruby/object:Gem::Requirement
+  requirement: &2158150040 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -76,10 +76,10 @@ dependencies:
         version: 1.6.4
   type: :development
   prerelease: false
-  version_requirements: *2154894220
+  version_requirements: *2158150040
 - !ruby/object:Gem::Dependency
   name: rcov
-  requirement: &2154893720 !ruby/object:Gem::Requirement
+  requirement: &2158149480 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -87,7 +87,7 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *2154893720
+  version_requirements: *2158149480
 description: Machine learning library in Ruby
 email: andrewliu33@gmail.com
 executables: []
@@ -108,7 +108,8 @@ files:
 - lib/method/perceptron.rb
 - lib/ml.rb
 - ml.gemspec
-- spec/learning_spec.rb
+- spec/data_spec.rb
+- spec/learner_spec.rb
 - spec/spec_helper.rb
 homepage: http://github.com/eggegg/ml
 licenses:
@@ -125,7 +126,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: 3039846434950023552
+      hash: -2452539594856608419
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements:

data/spec/learning_spec.rb

@@ -1,7 +0,0 @@
-require 'spec_helper'
-
-describe "Learning" do
-  it "fails" do
-    should.flunk "hey buddy, you should probably rename this file and start specing for real"
-  end
-end