Rsquared 0.0.1

data/.travis.yml

@@ -0,0 +1,8 @@
+--- 
+language: ruby
+rvm: 
+  - "1.9.3"
+  - "1.8.7"
+  - jruby-19mode
+
+install: ./install.sh

data/Gemfile

@@ -0,0 +1,4 @@
+source 'https://rubygems.org'
+
+# Specify your gem's dependencies in Rsquared.gemspec
+gemspec

data/LICENSE.txt

@@ -0,0 +1,22 @@
+Copyright (c) 2013 Daniel Cohen
+
+MIT License
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice shall be
+included in all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

data/README.md

@@ -0,0 +1,36 @@
+# Rsquared
+
+A full featured Ruby statistics library with assumption verification to make using statistics easy, 
+even with no background.
+
+[![Build Status](https://travis-ci.org/dacohen/Rsquared.png)](https://travis-ci.org/dacohen/Rsquared)
+
+## Installation
+
+Add this line to your application's Gemfile:
+
+    gem 'Rsquared'
+
+And then execute:
+
+    $ bundle
+
+Or install it yourself as:
+
+    $ gem install Rsquared
+
+## Usage
+
+You can run a statistical test, with assumption checking by supplying an array of numerical data points:
+
+    >> ttest = Rsquared::TTest.new(data)
+    >> ttest.statistic #=> Float
+    >> ttest.significant? #=> Boolean
+
+## Contributing
+
+1. Fork it
+2. Create your feature branch (`git checkout -b my-new-feature`)
+3. Commit your changes (`git commit -am 'Add some feature'`)
+4. Push to the branch (`git push origin my-new-feature`)
+5. Create new Pull Request

data/Rakefile

@@ -0,0 +1,11 @@
+require "bundler/gem_tasks"
+require "rake/testtask"
+
+task :default => :test
+
+Rake::TestTask.new do |t|
+	t.libs << "test"
+	t.test_files = FileList["test/*.rb"]
+	t.verbose = true
+end
+

data/Rsquared.gemspec

@@ -0,0 +1,24 @@
+# coding: utf-8
+lib = File.expand_path('../lib', __FILE__)
+$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
+require 'Rsquared/version'
+
+Gem::Specification.new do |spec|
+  spec.name          = "Rsquared"
+  spec.version       = Rsquared::VERSION
+  spec.authors       = ["Daniel Cohen"]
+  spec.email         = ["dcohen@gatech.edu"]
+  spec.description   = %q{A full-featured Ruby statistics library with assumption verification}
+  spec.summary       = %q{Provides statistical distributions, tests and verifies relevant assumptions}
+  spec.homepage      = "https://github.com/dacohen/Rsquared"
+  spec.license       = "MIT"
+
+  spec.files         = `git ls-files`.split($/)
+  spec.executables   = spec.files.grep(%r{^bin/}) { |f| File.basename(f) }
+  spec.test_files    = spec.files.grep(%r{^(test|spec|features)/})
+  spec.require_paths = ["lib"]
+
+  spec.add_development_dependency "bundler", "~> 1.3"
+  spec.add_development_dependency "rake"
+  spec.add_development_dependency "distribution"
+end

data/install.sh

@@ -0,0 +1,8 @@
+#! /usr/bin/env bash
+
+version=$(ruby -e "require './lib/Rsquared/version.rb'" -e "puts Rsquared::VERSION")
+
+bundle install
+gem build Rsquared.gemspec
+gem install Rsquared-"$version".gem
+

data/lib/Rsquared/version.rb

@@ -0,0 +1,3 @@
+module Rsquared
+  VERSION = "0.0.1"
+end

data/lib/constants.rb

@@ -0,0 +1,24 @@
+KSCV = [0.975, 0.842, 0.708, 0.624, 0.565, 0.521, 0.486, 0.457, 0.432, 0.410, 0.391, 0.375, 0.361, 0.349, 0.338, 0.328, 0.318, 0.309, 0.301, 0.294]
+
+module Rsquared
+       class Upper
+       	     def self.tail
+	     	 return 1
+     	     end
+       end
+
+       class Lower
+       	     def self.tail
+	     	 return -1
+	     end
+       end
+    
+       class Two
+       	     def self.sided
+	     	 return true
+	     end
+       end
+
+       class AssumptionError < StandardError
+       end
+end

data/lib/enumerableext.rb

@@ -0,0 +1,46 @@
+module Enumerable
+       def sum
+       	   return self.inject(0){|acc, i| acc+i}
+       end
+
+       def mean
+       	   return self.sum/self.length.to_f
+       end
+
+       def variance
+       	   varsum = self.inject(0){|acc, i| acc + (i - self.mean)**2}
+	   return(varsum/(self.length.to_f-1.0))
+       end
+
+       def popvariance
+       	   return self.variance*((self.length.to_f-1.0)/self.length.to_f)
+       end
+
+       def popstddev
+       	   return Math.sqrt(self.popvariance)
+       end
+
+       def stddev
+       	   return Math.sqrt(self.variance)
+       end
+
+       def skew
+       	   thirdsum = self.inject(0){|acc, i| acc + (i - self.mean)**3}
+	   thirdmoment = thirdsum/self.length.to_f
+	   return thirdmoment / (self.popvariance)**(3.0/2.0)
+       end
+
+       def kurtosis
+       	   fourthsum = self.inject(0){|acc, i| acc + (i - self.mean)**4}
+	   fourthmoment = fourthsum/self.length.to_f
+	   return (fourthmoment / (self.popvariance)**2)
+       end
+
+       def std
+       	   result = []
+       	   (0..self.length-1).each do |i|
+	   	result[i] = (self[i] - self.mean)/self.stddev
+	   end
+	   return result
+       end
+end

data/lib/rsquared.rb

@@ -0,0 +1,213 @@
+require "Rsquared/version"
+require "rubygems"
+require "distribution"
+require "constants"
+require "complex"
+require "enumerableext.rb"	      
+
+module Rsquared
+
+  ##
+  # KSTest implements the Kolomogorov-Smirnov test for normality
+  # kstest = Rsquared::KSTest.new(data)
+  # kstest.normal? => Boolean, indicates normality of data at 5% confidence
+  # 
+
+  class KSTest
+  	##
+	# Intitializes the test object with an array of numerical data
+	#
+
+  	def initialize(data)
+	    @data = data.std.sort!
+	    fn = 0
+	    d = []
+	    range = @data.max - @data.min
+	    @data.each_with_index do |x, i|
+	    	# Calculate Fn
+		fn = i + 1
+		d[i] = fn/@data.length.to_f - Distribution::Normal::cdf(x)
+		fn = 0.0
+	    end
+	    @ksstat = d.max
+	    return @ksstat
+	 end
+	 
+	 ##
+	 # Returns a boolean indiciating the significance of the test a the 5% level
+	 #
+
+	 def significant?
+	     if @ksstat > Helper::kscv(@data.length) then
+	     	return true
+	     else
+		return false
+	     end
+	 end
+
+	 ##
+	 # Returns logical opposite of significance
+	 #
+
+	 def normal?
+	     !self.significant?
+	 end
+
+	 def inspect
+	     significant?
+	 end
+	 
+	 ##
+	 # Returns the test statistic
+	 #
+
+	 def statistic
+	     @ksstat
+	 end
+  end
+
+  ##
+  # Tests for outliers on either side of the data
+  # grubbs = Rsquared::GrubbsTest.new(data)
+  # grubbs.significant? => Boolean
+  #
+
+  class GrubbsTest
+	  ##
+	  # Initializes the Test object with an array of numerical data
+	  #
+
+	  def initialize(data)
+	     @data = data.sort
+   	     @gstat = [((@data.mean - @data.min)/@data.stddev).abs, ((@data.mean - @data.max)/@data.stddev).abs].max
+	  end
+	  
+	  ##
+	  # Returns a boolean indicating the significance of the test at the 5% level
+	  #	  
+
+	  def significant?(alpha=0.05)
+	     if @gstat > Helper::grubbscv(@data.length, alpha) then
+	     	return true
+	     else
+		return false
+	     end
+	  end
+
+	  def inspect
+	      significant?
+	  end
+
+	  ##
+	  # Returns the test statistic as a float
+	  #
+
+	  def statistic
+	      @gstat
+	  end
+
+	  alias_method :outlier?, :significant?	      
+  end
+  
+  ##
+  # Tests for deviation of sample mean from expected mean
+  # ttest = Rsquared::TTest.new(data, mu0, sided)
+  # mu0 is the expected value of the sample mean
+  # Supply Rsquared::Upper.tail, Rsquared::Lower.tail or Rsquared::Two.sided
+  # Use Upper.tail when you suspect that the sample mean will be greater than the expected mean
+  # Use Lower.tail when you suspect that the sample mean will be smaller than the expected mean
+  # Use Two.sided when you suspect neither
+
+  class TTest
+  	##
+	# Initializes the TTest object with the supplied arguments
+	#
+
+  	def initialize(data, mu0, sided)
+	    @data = data
+  	    @mu0 = mu0
+	    @sided = sided
+	    
+	    if KSTest.new(@data).significant? and @data.length < 40 then
+	       raise AssumptionException, "The data is not close enough to a normal distribution for such a small sample size"
+	    end
+	    if GrubbsTest.new(@data).outlier? then
+	       raise AssumptionException, "Your data has one or more outliers, which the T-Distribution cannot handle"
+	    end
+
+	    @tstat = (@data.mean - @mu0)/(data.stddev/Math.sqrt(@data.length))
+	    @pvalue = Distribution::T::cdf(@tstat, @data.length-1)
+	    if @sided == Upper.tail then
+	       @pvalue = 1.0-@pvalue
+	    elsif @sided == Two.sided then
+	       @pvalue = [(1.0-@pvalue)*2.0, @pvalue*2.0].min
+	    end
+	 end
+
+	 def inspect
+	     @pvalue
+	 end
+
+	 ##
+	 # Returns the t-statistic
+	 #
+
+	 def statistic
+	     @tstat
+	 end
+
+	 ##
+	 # Checks for significance at the supplied alpha level
+	 #
+
+	 def significant?(alpha=0.05)
+	     if @pvalue < alpha then
+	     	return true
+	     else
+		return false
+	     end
+	 end
+  end
+
+  ##
+  # The Helper module implements uncommon statistical functions directly
+  # For use by experts only
+  # = Example
+  #
+  # Rsquared::Helper::kscv(30) => 0.190
+  #
+  module Helper
+	##
+	# kscv(n) => Float
+	# Estimates the 5% critical value of the Kolomogorov-Smirnov distribution given sample size
+	#
+	
+	def kscv(n)
+	    if n < 1 then
+	       return 1.0
+	    elsif n < 21 then
+	       return KSCV[n-1]
+	    elsif n >= 20 and n < 25 then
+	       return 0.270
+	    elsif n >= 25 and n < 30 then
+	       return 0.240
+	    elsif n >= 30 and n < 35 then
+	       return 0.230
+	    elsif n > 35 then
+	       return 1.36/Math.sqrt(n)
+	    end
+	end
+
+	##
+	# grubbscv(n, alpha) => Float
+	# Calculates the Grubbs critical value
+	#
+	
+	def grubbscv(n, alpha)
+	    tcv = Distribution::T::p_value(alpha/(2*n), n-2)
+	    return ((n-1)/Math.sqrt(n))*Math.sqrt(tcv**2/((n-2)+tcv**2))
+	end
+ 
+	 module_function :kscv, :grubbscv
+  end
+end

data/test/enumerable_tests.rb

@@ -0,0 +1,44 @@
+require 'test/unit'
+require 'rsquared'
+
+class EnumerableTest < Test::Unit::TestCase
+
+      def test_sum
+      	  assert_in_delta 15.0, [1,2,3,4,5].sum, 0.001
+      end
+
+      def test_mean
+      	  assert_in_delta 3.0, [1,2,3,4,5].mean, 0.001
+      end
+
+      def test_variance
+      	  assert_in_delta 2.5, [1,2,3,4,5].variance, 0.001
+      end
+      
+      def test_stddev
+      	  assert_in_delta 1.5811, [1,2,3,4,5].stddev, 0.001
+      end
+
+      def test_popvariance
+      	  assert_in_delta 2.0, [1,2,3,4,5].popvariance, 0.001
+      end
+      
+      def test_popstddev
+      	  assert_in_delta 1.4142, [1,2,3,4,5].popstddev, 0.001
+      end
+
+      def test_skew
+      	  assert_in_delta 0.0, [1,2,3,4,5].skew, 0.001
+      end
+
+      def test_kurtosis
+      	  assert_in_delta 1.7, [1,2,3,4,5].kurtosis, 0.1
+      end
+
+      def test_std
+      	  @checkvalues = [-1.2649, -0.63247, 0, 0.63257, 1.2649]
+      	  [1,2,3,4,5].std.each_with_index do |x, i|
+	  	assert_in_delta @checkvalues[i], x, 0.001
+	  end
+      end 
+end

data/test/helper_tests.rb

@@ -0,0 +1,16 @@
+require 'test/unit'
+require 'rsquared'
+
+class HelperTest < Test::Unit::TestCase
+
+      def test_kscv
+      	        assert_equal 0.410, Rsquared::Helper::kscv(10)
+	  	assert_equal 0.240, Rsquared::Helper::kscv(27)
+	  	assert_in_delta 0.20273, Rsquared::Helper::kscv(45), 0.001
+      end
+    
+      def test_grubbscv
+      	  assert_in_delta 2.91, Rsquared::Helper::grubbscv(30, 0.05), 0.01
+      end
+  
+end

data/test/rsquared_tests.rb

@@ -0,0 +1,42 @@
+require 'test/unit'
+require 'rsquared'
+
+module Test::Unit::Assertions
+       def refute(bool, *rest)
+       	   assert(!bool, *rest)
+       end
+end
+
+$data = [-105, 135, 40, 90, -55, -85, 70, 180, 140, -10, -105, 40, 185, -90, -90, 80, 70, -155, 345, 250, 10, -135, 80, 85, -40, 250, -20, 35, 305, -135]
+
+class RsquaredTests < Test::Unit::TestCase
+
+      def test_KS
+	  kstest = Rsquared::KSTest.new($data)	  
+      	  assert_in_delta 0.1046877, kstest.statistic, 0.001
+	  assert kstest.normal?
+      end
+
+      def test_TTest
+      	  ttest = Rsquared::TTest.new($data, 0, Rsquared::Two.sided)
+	  assert_in_delta 0.0763, ttest.inspect, 0.001
+	  refute ttest.significant?
+
+	  ttest = Rsquared::TTest.new($data, 0, Rsquared::Upper.tail)
+	  assert_in_delta 0.03813, ttest.inspect, 0.001
+	  assert ttest.significant?
+      end
+
+      def test_Grubbs
+      	  grubbs = Rsquared::GrubbsTest.new($data)
+	  assert_in_delta 2.21, grubbs.statistic, 0.01
+	  refute grubbs.significant?
+	  refute grubbs.outlier?
+
+	  data = $data + [800]
+	  grubbs = Rsquared::GrubbsTest.new(data)
+	  assert grubbs.significant?
+	  assert grubbs.outlier?
+
+      end
+end

metadata

@@ -0,0 +1,124 @@
+--- !ruby/object:Gem::Specification 
+name: Rsquared
+version: !ruby/object:Gem::Version 
+  hash: 29
+  prerelease: 
+  segments: 
+  - 0
+  - 0
+  - 1
+  version: 0.0.1
+platform: ruby
+authors: 
+- Daniel Cohen
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2013-06-28 00:00:00 Z
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: bundler
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        hash: 9
+        segments: 
+        - 1
+        - 3
+        version: "1.3"
+  type: :development
+  version_requirements: *id001
+- !ruby/object:Gem::Dependency 
+  name: rake
+  prerelease: false
+  requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
+        version: "0"
+  type: :development
+  version_requirements: *id002
+- !ruby/object:Gem::Dependency 
+  name: distribution
+  prerelease: false
+  requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
+        version: "0"
+  type: :development
+  version_requirements: *id003
+description: A full-featured Ruby statistics library with assumption verification
+email: 
+- dcohen@gatech.edu
+executables: []
+
+extensions: []
+
+extra_rdoc_files: []
+
+files: 
+- .travis.yml
+- Gemfile
+- LICENSE.txt
+- README.md
+- Rakefile
+- Rsquared.gemspec
+- install.sh
+- lib/Rsquared/version.rb
+- lib/constants.rb
+- lib/enumerableext.rb
+- lib/rsquared.rb
+- lib/version.rb
+- test/enumerable_tests.rb
+- test/helper_tests.rb
+- test/rsquared_tests.rb
+homepage: https://github.com/dacohen/Rsquared
+licenses: 
+- MIT
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.8.24
+signing_key: 
+specification_version: 3
+summary: Provides statistical distributions, tests and verifies relevant assumptions
+test_files: 
+- test/enumerable_tests.rb
+- test/helper_tests.rb
+- test/rsquared_tests.rb