statsample-glm 0.0.1

data/.document

@@ -0,0 +1,5 @@
+lib/**/*.rb
+bin/*
+- 
+features/**/*.feature
+LICENSE.txt

data/.travis.yml

@@ -0,0 +1,13 @@
+language: ruby
+rvm:
+  - 1.9.2
+  - 1.9.3
+  - 2.0.0
+  - jruby-19mode # JRuby in 1.9 mode
+  - rbx-19mode
+#  - 1.8.7
+#  - jruby-18mode # JRuby in 1.8 mode
+#  - rbx-18mode
+
+# uncomment this line if your project needs to run something other than `rake`:
+# script: bundle exec rspec spec

data/Gemfile

@@ -0,0 +1,20 @@
+source "http://rubygems.org"
+
+gem 'statsample', '>=1.2.0'
+# Add dependencies required to use your gem here.
+# Example:
+   gem "activesupport", "= 3.2.10"
+
+# Add dependencies to develop your gem here.
+# Include everything needed to run rake, tests, features, etc.
+group :development do
+  gem "shoulda", ">= 0"
+  gem "rdoc", "~> 3.12"
+  gem "minitest", "~> 4.7.5"
+  gem "cucumber", ">= 0"
+  gem "bundler", "~> 1.3.5"
+  gem "jeweler", "~> 1.8.4"
+  gem "bio", ">= 1.4.2"
+  gem "rdoc", "~> 3.12"
+  gem "mocha", "~> 0.14.0"
+end

data/LICENSE.txt

@@ -0,0 +1,22 @@
+This version of Statsample-GLM is licensed under the BSD 2-clause license.
+
+* http://sciruby.com
+* http://github.com/sciruby/sciruby/wiki/License
+
+You *must* read the Contributor Agreement before contributing code to the SciRuby Project. This is available online:
+
+* http://github.com/sciruby/sciruby/wiki/Contributor-Agreement
+
+-----
+
+Copyright (c) 2010 - 2013, Ruby Science Foundation
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
+
+* Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
+
+* Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+

data/README.rdoc

@@ -0,0 +1,71 @@
+= statsample-glm
+
+{<img
+src="https://secure.travis-ci.org/AnkurGel/statsample-glm.png"
+/>}[http://travis-ci.org/#!/AnkurGel/statsample-glm]
+
+Statsample-GLM is an extension of *Generalized Linear Models* to {Statsample}[https://github.com/SciRuby/statsample], a suite of advance statistics in Ruby.
+
+* {sciruby.com}[http://sciruby.com]
+* {Google+}[https://plus.google.com/109304769076178160953/posts]
+* {Ankur Goel}[http://ankurgoel.com]
+* {Statsample}[https://github.com/SciRuby/statsample]
+
+
+== Description
+Statsample-GLM is an extension of Statsample, and includes many helpful regression techniques for Generalized Linear models such as:
+
+* Poisson Regression
+* Logistic Regression
+* Exponential Regression
+* Iteratively Reweighted Least Squares
+
+Statsample-GLM was created by Ankur Goel as part of Google's Summer of Code 2013. It is the part of {SciRuby}[http://sciruby.com]
+
+Note: This is under active development!
+
+== Installation
+
+        gem install statsample-glm
+
+
+== Usage
+
+To use the library 
+
+        require 'statsample-glm'
+
+You can also go through the blog-posts on {my blog}[http://ankurgoel.com] for descriptive explanation and examples.
+
+
+== Documentation 
+
+The API doc is {online}[http://rubygems.org/gems/statsample-glm]. For more code examples see also the test files in the source tree.
+        
+
+== Contributing
+
+* Fork the project. 
+* Create your feature branch
+* Add/Modify code. 
+* Write equivalent documentation and **tests**.
+* Run `rake test` to verify that all test case passes.
+* Push your branch.
+* Pull request. :)
+
+
+== Project home page
+
+Information on the source tree, documentation, issues and how to contribute, see
+
+  http://github.com/ankurgel/statsample-glm
+
+
+== Biogems.info
+
+This Biogem is published at http://biogems.info/index.html#statsample-glm
+
+== Copyright
+
+Copyright (c) 2013 Ankur Goel. See LICENSE.txt for further details.
+

data/Rakefile

@@ -0,0 +1,48 @@
+# encoding: utf-8
+
+require 'rubygems'
+require 'bundler'
+begin
+  Bundler.setup(:default, :development)
+rescue Bundler::BundlerError => e
+  $stderr.puts e.message
+  $stderr.puts "Run `bundle install` to install missing gems"
+  exit e.status_code
+end
+require 'rake'
+
+require 'jeweler'
+Jeweler::Tasks.new do |gem|
+  # gem is a Gem::Specification... see http://docs.rubygems.org/read/chapter/20 for more options
+  gem.name = "statsample-glm"
+  gem.homepage = "http://github.com/AnkurGel/statsample-glm"
+  gem.license = "MIT"
+  gem.summary = %Q{Generalized Linear Models for Statsample}
+  gem.description = %Q{Statsample-GLM is an extension to Statsample, an advance statistics suite in Ruby. This gem includes modules for Regression techniques such as Poisson Regression, Logistic Regression and Exponential Regression}
+  gem.email = "ankurgel@gmail.com"
+  gem.authors = ["Ankur Goel"]
+  # dependencies defined in Gemfile
+end
+Jeweler::RubygemsDotOrgTasks.new
+
+require 'rake/testtask'
+Rake::TestTask.new(:test) do |test|
+  test.libs << 'lib' << 'test'
+  test.pattern = 'test/**/test_*.rb'
+  test.verbose = true
+end
+
+require 'cucumber/rake/task'
+Cucumber::Rake::Task.new(:features)
+
+task :default => :test
+
+require 'rdoc/task'
+Rake::RDocTask.new do |rdoc|
+  version = File.exist?('VERSION') ? File.read('VERSION') : ""
+
+  rdoc.rdoc_dir = 'rdoc'
+  rdoc.title = "statsample-glm #{version}"
+  rdoc.rdoc_files.include('README*')
+  rdoc.rdoc_files.include('lib/**/*.rb')
+end

data/VERSION

@@ -0,0 +1 @@
+0.0.1

data/features/bio-statsample-glm.feature

@@ -0,0 +1,9 @@
+Feature: something something
+  In order to something something
+  A user something something
+  something something something
+
+  Scenario: something something
+    Given inspiration
+    When I create a sweet new gem
+    Then everyone should see how awesome I am

data/features/support/env.rb

@@ -0,0 +1,15 @@
+require 'bundler'
+begin
+  Bundler.setup(:default, :development)
+rescue Bundler::BundlerError => e
+  $stderr.puts e.message
+  $stderr.puts "Run `bundle install` to install missing gems"
+  exit e.status_code
+end
+
+$LOAD_PATH.unshift(File.dirname(__FILE__) + '/../../lib')
+require 'statsample-glm'
+
+require 'test/unit/assertions'
+
+World(Test::Unit::Assertions)

data/lib/statsample-glm.rb

@@ -0,0 +1,12 @@
+# Please require your code below, respecting the naming conventions in the
+# bioruby directory tree.
+#
+# For example, say you have a plugin named bio-plugin, the only uncommented
+# line in this file would be 
+#
+#   require 'bio/bio-plugin/plugin'
+#
+# In this file only require other files. Avoid other source code.
+
+require 'statsample'
+require 'statsample-glm/regression'

data/lib/statsample-glm/regression.rb

@@ -0,0 +1,69 @@
+require 'statsample-glm/regression/poisson'
+require 'statsample-glm/regression/logistic'
+module Statsample
+  module Regression
+    include Statsample::VectorShorthands
+
+    # = Generalized linear models
+    # == Parameters
+    #
+    # * x = model matrix
+    # * y = response vector
+    # * method = symbol; choice of glm strategy, default = :poisson
+    #
+    # == Usage
+    #    require 'statsample-glm'
+    #    x1=Statsample::Vector.new([0.537322309644812,-0.717124209978434,-0.519166718891331,0.434970973986765,-0.761822002215759,1.51170030921189,0.883854199811195,-0.908689798854196,1.70331977539793,-0.246971150634099,-1.59077593922623,-0.721548040910253,0.467025703920194,-0.510132788447137,0.430106510266798,-0.144353683251536,-1.54943800728303,0.849307651309298,-0.640304240933579,1.31462478279425,-0.399783455165345,0.0453055645017902,-2.58212161987746,-1.16484414309359,-1.08829266466281,-0.243893919684792,-1.96655661929441,0.301335373291024,-0.665832694463588,-0.0120650855753837,1.5116066367604,0.557300353673344,1.12829931872045,0.234443748015922,-2.03486690662651,0.275544751380246,-0.231465849558696,-0.356880153225012,-0.57746647541923,1.35758352580655,1.23971669378224,-0.662466275100489,0.313263561921793,-1.08783223256362,1.41964722846899,1.29325100940785,0.72153880625103,0.440580131022748,0.0351917814720056, -0.142353224879252],:scale)
+    #    x2=Statsample::Vector.new([-0.866655707911859,-0.367820249977585,0.361486610435,0.857332626245179,0.133438466268095,0.716104533073575,1.77206093023382,-0.10136697295802,-0.777086491435508,-0.204573554913706,0.963353531412233,-1.10103024900542,-0.404372761837392,-0.230226345183469,0.0363730246866971,-0.838265540390497,1.12543549657924,-0.57929175648001,-0.747060244805248,0.58946979365152,-0.531952663697324,1.53338594419818,0.521992029051441,1.41631763288724,0.611402316795129,-0.518355638373296,-0.515192557101107,-0.672697937866108,1.84347042325327,-0.21195540664804,-0.269869371631611,0.296155694010096,-2.18097898069634,-1.21314663927206,1.49193669881581,1.38969280369493,-0.400680808117106,-1.87282814976479,1.82394870451051,0.637864732838274,-0.141155946382493,0.0699950644281617,1.32568550595165,-0.412599258349398,0.14436832227506,-1.16507785388489,-2.16782049922428,0.24318371493798,0.258954871320764,-0.151966534521183],:scale)
+    #    y=Statsample::Vector.new([0,0,1,0,1,1,1,1,0,1,1,1,1,0,1,0,1,1,0,1,0,1,1,1,1,0,0,1,1,0,0,1,0,0,1,1,0,0,1,1,0,1,1,1,1,0,0,0,1,1],:scale)
+    #    x=Statsample::Dataset.new({"i"=>intercept,"x1"=>x1,"x2"=>x2})
+    #    obj = Statsample::Regression.glm(x, y, :binomial)
+    #    #=> Logistic Regression object
+    #
+    # == Returns
+    #    GLM object for given method.
+    def self.glm(x, y, method=:gaussian)
+    
+      if method.downcase.to_sym == :poisson
+        obj = Statsample::Regression::GLM::Poisson.new(x,y)
+      elsif method.downcase.to_sym == :binomial
+        obj = Statsample::Regression::GLM::Logistic.new(x,y)
+      else
+        raise("Not implemented yet")
+      end
+      obj.irwls
+      obj
+    end
+
+
+    def self.irwls(x, y, mu, w, j, h, epsilon = 1e-7, max_iter = 100)
+      b = Matrix.column_vector(Array.new(x.column_size,0.0))
+      converged = false
+      1.upto(max_iter) do |i|
+        #conversion from : (solve(j(x,b)) %*% h(x,b,y))
+
+        intermediate = (j.call(x,b).inverse * h.call(x,b,y))
+        b_new = b - intermediate
+
+        if((b_new - b).map(&:abs)).to_a.flatten.inject(:+) < epsilon
+          converged = true
+          b = b_new
+          break
+        end
+        b = b_new
+      end
+      ss = j.call(x,b).inverse.diagonal.map{ |x| -x}.map{ |y| Math.sqrt(y) }
+      values = mu.call(x,b)
+
+      residuals = y - values.column_vectors.map(&:to_a).flatten
+      df_residuals = y.count - x.column_size
+      return [create_vector(b.column_vectors[0]), create_vector(ss), create_vector(values.to_a.flatten),
+              residuals, max_iter, df_residuals, converged]
+    end
+
+    private
+    def self.create_vector(arr)
+      Statsample::Vector.new(arr, :scale)
+    end
+  end
+end

data/lib/statsample-glm/regression/logistic.rb

@@ -0,0 +1,108 @@
+module Statsample
+  module Regression
+    module GLM
+
+      class Logistic
+
+        attr_reader :se
+        # The fitted mean values
+        attr_reader :fit
+        # the _working_ residuals; that is the residuals in the final iteration of the IRWLS fit.
+        attr_reader :residuals
+        # The residuals degree of freedom
+        attr_reader :df
+        # Number of iterations used for convergence
+        attr_reader :iter
+        # Boolean. Tells whether the IRWLS for the given model converged or not
+        attr_reader :converged
+
+        def initialize(ds, y)
+          @ds=ds
+          @fields=@ds.fields
+          @x = ds.to_matrix
+          @y = y
+        end
+
+        # named vector/hash of coefficients
+        # === Parameter
+        # * *type*: symbol; (:array, default). Options = [:array, :hash]
+        def coefficients(type=:array)
+          if type==:array
+            #originally returned as vector; so pass it
+            @coefficients
+          elsif type==:hash
+            h={}
+            @fields.size.times {|i|
+              h[@fields[i]]=@coefficients[i]
+            }
+            h
+          end
+        end
+        def self.mu(x, b)
+          matrix_mul = x * b
+          numerator = matrix_mul.map { |y| Math.exp(y) }
+          denominator = numerator.map { |y| 1 + y }
+
+          numerator.each_with_index { |e, r, c|
+            numerator[r,c] = numerator[r,c].to_f / denominator[r,c].to_f
+          }
+        end
+
+        def self.w(x, b)
+          mus = mu(x,b).column_vectors.map(&:to_a).flatten
+          mus_intermediate = mus.collect { |x| 1 - x }
+          w = mus.zip(mus_intermediate).collect { |x| x.inject(:*) }
+          w_mat = Matrix.I(w.size)
+          w_enum = w.to_enum
+          return w_mat.map do |x|
+            x.eql?(1) ? w_enum.next : x
+          end
+        end
+
+        def self.h(x,b,y)
+          x_t = x.transpose
+          mu_flat = mu(x,b).column_vectors.map(&:to_a).flatten
+          column_data = y.zip(mu_flat).collect { |x| x.inject(:-) }
+          x_t * Matrix.column_vector(column_data)
+        end
+
+        def self.j(x,b)
+          w_matrix = w(x, b)
+          jacobian_matrix = x.transpose * w_matrix * x
+          jacobian_matrix.map { |x| -x }
+        end
+
+        def to_s
+          sprintf("Logistic Regression (Statsample::Regression::GLM;:Logistic)")
+        end
+
+        # = Iteratively reweighted least squares
+        #   Computes irwls for given model and parameters.
+        #
+        # == Usage
+        #    require 'statsample-glm'
+        #    x1=Statsample::Vector.new([0.537322309644812,-0.717124209978434,-0.519166718891331,0.434970973986765,-0.761822002215759,1.51170030921189,0.883854199811195,-0.908689798854196,1.70331977539793,-0.246971150634099,-1.59077593922623,-0.721548040910253,0.467025703920194,-0.510132788447137,0.430106510266798,-0.144353683251536,-1.54943800728303,0.849307651309298,-0.640304240933579,1.31462478279425,-0.399783455165345,0.0453055645017902,-2.58212161987746,-1.16484414309359,-1.08829266466281,-0.243893919684792,-1.96655661929441,0.301335373291024,-0.665832694463588,-0.0120650855753837,1.5116066367604,0.557300353673344,1.12829931872045,0.234443748015922,-2.03486690662651,0.275544751380246,-0.231465849558696,-0.356880153225012,-0.57746647541923,1.35758352580655,1.23971669378224,-0.662466275100489,0.313263561921793,-1.08783223256362,1.41964722846899,1.29325100940785,0.72153880625103,0.440580131022748,0.0351917814720056, -0.142353224879252],:scale)
+        #    x2=Statsample::Vector.new([-0.866655707911859,-0.367820249977585,0.361486610435,0.857332626245179,0.133438466268095,0.716104533073575,1.77206093023382,-0.10136697295802,-0.777086491435508,-0.204573554913706,0.963353531412233,-1.10103024900542,-0.404372761837392,-0.230226345183469,0.0363730246866971,-0.838265540390497,1.12543549657924,-0.57929175648001,-0.747060244805248,0.58946979365152,-0.531952663697324,1.53338594419818,0.521992029051441,1.41631763288724,0.611402316795129,-0.518355638373296,-0.515192557101107,-0.672697937866108,1.84347042325327,-0.21195540664804,-0.269869371631611,0.296155694010096,-2.18097898069634,-1.21314663927206,1.49193669881581,1.38969280369493,-0.400680808117106,-1.87282814976479,1.82394870451051,0.637864732838274,-0.141155946382493,0.0699950644281617,1.32568550595165,-0.412599258349398,0.14436832227506,-1.16507785388489,-2.16782049922428,0.24318371493798,0.258954871320764,-0.151966534521183],:scale)
+        #    y=Statsample::Vector.new([0,0,1,0,1,1,1,1,0,1,1,1,1,0,1,0,1,1,0,1,0,1,1,1,1,0,0,1,1,0,0,1,0,0,1,1,0,0,1,1,0,1,1,1,1,0,0,0,1,1],:scale)
+        #    x=Statsample::Dataset.new({"i"=>intercept,"x1"=>x1,"x2"=>x2})
+        #    obj = Statsample::Regression.glm(x, y, :binomial)
+        #    #=> Logistic Regression object
+        #    obj.irlws
+        #    #=> Array of returned values
+        #    obj.coefficients
+        #    #=> named vector of coefficients
+
+        def irwls
+          x, y = @x, @y
+          #calling irwls on Regression and passing equivalent methods in lambdas.
+          #Ruby_level+=awesome!
+          @coefficients, @se, @fit, @residuals, @df, @iter, @converged = Statsample::Regression.irwls(
+              x,y, ->l,m{self.class.mu(l,m)}, ->l,m{self.class.w(l,m)},
+              ->l,m{self.class.j(l,m)}, ->k,l,m{self.class.h(k,l,m)}
+          )
+        end
+      end
+
+    end
+  end
+end

data/lib/statsample-glm/regression/poisson.rb

@@ -0,0 +1,90 @@
+module Statsample
+  module Regression
+    module GLM
+
+      class Poisson
+
+        attr_reader :se
+        # The fitted mean values
+        attr_reader :fit
+        # the _working_ residuals; that is the residuals in the final iteration of the IRWLS fit.
+        attr_reader :residuals
+        # The residuals degree of freedom
+        attr_reader :df
+        # Number of iterations used for convergence
+        attr_reader :iter
+        # Boolean. Tells whether the IRWLS for the given model converged or not
+        attr_reader :converged
+
+        def initialize(ds, y)
+          @ds=ds
+          @fields=@ds.fields
+          @x = ds.to_matrix
+          @y = y
+        end
+
+        # named vector/hash of coefficients
+        # === Parameter
+        # * *type*: symbol; (:array, default). Options = [:array, :hash]
+        def coefficients(type=:array)
+          if type==:array
+            @coefficients
+          elsif type==:hash
+            h={}
+            @fields.size.times {|i|
+              h[@fields[i]]=@coefficients[i]
+            }
+            h
+          end
+        end
+
+        def self.mu(x, b, link=:log)
+          if link.downcase.to_sym == :log
+            (x * b).map { |y| Math.exp(y) }
+          elsif link.downcase.to_sym == :sqrt
+            (x * b).collect { |y| y**2 }
+          end
+        end
+
+        def self.w(x, b)
+          poisson_mu = mu(x,b)
+          mu_flat = poisson_mu.column_vectors.map(&:to_a).flatten
+
+          w_mat = Matrix.I(mu_flat.size)
+          mu_enum = mu_flat.to_enum
+          return w_mat.map do |x|
+            x.eql?(1) ? mu_enum.next : x
+          end
+        end
+
+        def self.h(x, b, y)
+          x_t = x.transpose
+          mu_flat = mu(x,b).column_vectors.map(&:to_a).flatten
+          column_data = y.zip(mu_flat).collect { |x| x.inject(:-) }
+          x_t * Matrix.columns([column_data])
+        end
+
+        def self.j(x, b)
+          w_matrix = w(x, b)
+          jacobian_matrix = x.transpose * w_matrix * x
+          jacobian_matrix.map { |x| -x }
+        end
+
+        def to_s
+          sprintf("Logistic Regression (Statsample::Regression::GLM;:Logistic)")
+        end
+
+        def irwls
+          x,y = @x,@y
+          #calling irwls on Regression and passing equivalent methods in lambdas.
+          #Ruby_level+=awesome!
+          @coefficients, @se, @fit, @residuals, @df, @iter, @converged = Statsample::Regression.irwls(
+              x,y, ->l,m{self.class.mu(l,m)}, ->l,m{self.class.w(l,m)},
+              ->l,m{self.class.j(l,m)}, ->k,l,m{self.class.h(k,l,m)}
+          )
+        end
+
+      end
+    end
+  end
+end

data/test/helper.rb

@@ -0,0 +1,87 @@
+require 'rubygems'
+require 'bundler'
+begin
+  Bundler.setup(:default, :development)
+rescue Bundler::BundlerError => e
+  $stderr.puts e.message
+  $stderr.puts "Run `bundle install` to install missing gems"
+  exit e.status_code
+end
+require 'minitest/unit'
+require 'shoulda'
+require 'shoulda-context'
+require 'mocha/setup'
+
+
+$LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
+$LOAD_PATH.unshift(File.dirname(__FILE__))
+require 'statsample-glm'
+module MiniTest
+  class Unit
+    class TestCase
+      include Shoulda::Context::Assertions
+      include Shoulda::Context::InstanceMethods
+      extend Shoulda::Context::ClassMethods
+      def self.should_with_gsl(name,&block)
+        should(name) do
+          if Statsample.has_gsl?
+            instance_eval(&block)
+          else
+            skip("Requires GSL")
+          end
+        end
+      end
+    end
+  end
+
+  module Assertions
+    def assert_similar_vector(exp, obs, delta=1e-10,msg=nil)
+      msg||="Different vectors #{exp} - #{obs}"
+      assert_equal(exp.size, obs.size)
+      exp.data_with_nils.each_with_index {|v,i|
+        assert_in_delta(v,obs[i],delta)
+      }
+    end
+    def assert_similar_hash(exp, obs, delta=1e-10,msg=nil)
+      msg||="Different hash #{exp} - #{obs}"
+      assert_equal(exp.size, obs.size)
+      exp.each_key {|k|
+        assert_in_delta(exp[k],obs[k],delta)
+      }
+    end
+
+    def assert_equal_vector(exp,obs,delta=1e-10,msg=nil)
+      assert_equal(exp.size, obs.size, "Different size.#{msg}")
+      exp.size.times {|i|
+        assert_in_delta(exp[i],obs[i],delta, "Different element #{i}. \nExpected:\n#{exp}\nObserved:\n#{obs}.#{msg}")
+      }
+    end
+    def assert_equal_matrix(exp,obs,delta=1e-10,msg=nil)
+      assert_equal(exp.row_size, obs.row_size, "Different row size.#{msg}")
+      assert_equal(exp.column_size, obs.column_size, "Different column size.#{msg}")
+      exp.row_size.times {|i|
+        exp.column_size.times {|j|
+          assert_in_delta(exp[i,j],obs[i,j], delta, "Different element #{i},#{j}\nExpected:\n#{exp}\nObserved:\n#{obs}.#{msg}")
+        }
+      }
+    end
+    alias :assert_raise :assert_raises unless method_defined? :assert_raise
+    alias :assert_not_equal :refute_equal unless method_defined? :assert_not_equal
+    alias :assert_not_same :refute_same unless method_defined? :assert_not_same
+    unless method_defined? :assert_nothing_raised
+      def assert_nothing_raised(msg=nil)
+        msg||="Nothing should be raised, but raised %s"
+        begin
+          yield
+          not_raised=true
+        rescue Exception => e
+          not_raised=false
+          msg=sprintf(msg,e)
+        end
+        assert(not_raised,msg)
+      end
+    end
+  end
+end
+
+MiniTest::Unit.autorun

data/test/test_glm.rb

@@ -0,0 +1,4 @@
+require(File.expand_path(File.dirname(__FILE__)+'/helper.rb'))
+
+class StatsampleRegressionGlm < MiniTest::Unit::TestCase
+end

data/test/test_glm_logistic.rb

@@ -0,0 +1,23 @@
+require(File.expand_path(File.dirname(__FILE__)+'/helper.rb'))
+
+class StatsampleRegressionGlmLogistic < MiniTest::Unit::TestCase
+
+  context("Example") do
+    setup do
+      x1=Statsample::Vector.new([0.537322309644812,-0.717124209978434,-0.519166718891331,0.434970973986765,-0.761822002215759,1.51170030921189,0.883854199811195,-0.908689798854196,1.70331977539793,-0.246971150634099,-1.59077593922623,-0.721548040910253,0.467025703920194,-0.510132788447137,0.430106510266798,-0.144353683251536,-1.54943800728303,0.849307651309298,-0.640304240933579,1.31462478279425,-0.399783455165345,0.0453055645017902,-2.58212161987746,-1.16484414309359,-1.08829266466281,-0.243893919684792,-1.96655661929441,0.301335373291024,-0.665832694463588,-0.0120650855753837,1.5116066367604,0.557300353673344,1.12829931872045,0.234443748015922,-2.03486690662651,0.275544751380246,-0.231465849558696,-0.356880153225012,-0.57746647541923,1.35758352580655,1.23971669378224,-0.662466275100489,0.313263561921793,-1.08783223256362,1.41964722846899,1.29325100940785,0.72153880625103,0.440580131022748,0.0351917814720056, -0.142353224879252],:scale)
+      x2=Statsample::Vector.new([-0.866655707911859,-0.367820249977585,0.361486610435,0.857332626245179,0.133438466268095,0.716104533073575,1.77206093023382,-0.10136697295802,-0.777086491435508,-0.204573554913706,0.963353531412233,-1.10103024900542,-0.404372761837392,-0.230226345183469,0.0363730246866971,-0.838265540390497,1.12543549657924,-0.57929175648001,-0.747060244805248,0.58946979365152,-0.531952663697324,1.53338594419818,0.521992029051441,1.41631763288724,0.611402316795129,-0.518355638373296,-0.515192557101107,-0.672697937866108,1.84347042325327,-0.21195540664804,-0.269869371631611,0.296155694010096,-2.18097898069634,-1.21314663927206,1.49193669881581,1.38969280369493,-0.400680808117106,-1.87282814976479,1.82394870451051,0.637864732838274,-0.141155946382493,0.0699950644281617,1.32568550595165,-0.412599258349398,0.14436832227506,-1.16507785388489,-2.16782049922428,0.24318371493798,0.258954871320764,-0.151966534521183],:scale)
+      @y_log=Statsample::Vector.new([0,0,1,0,1,1,1,1,0,1,1,1,1,0,1,0,1,1,0,1,0,1,1,1,1,0,0,1,1,0,0,1,0,0,1,1,0,0,1,1,0,1,1,1,1,0,0,0,1,1],:scale)
+      @y_pois=Statsample::Vector.new([1,2,1,3,3,1,10,1,1,2,15,0,0,2,1,2,18,2,1,1,1,8,18,13,7,1,1,0,26,0,2,2,0,0,25,7,0,0,21,0,0,1,5,0,3,0,0,1,0,0],:scale)
+      intercept=Statsample::Vector.new([1]*50,:scale)
+      @df=Statsample::Dataset.new({"i"=>intercept,"x1"=>x1,"x2"=>x2})
+      @glm=Statsample::Regression.glm(@df,@y_log,:binomial)
+    end
+    should "report correct coefficientes as array" do
+      assert_similar_vector(@glm.coefficients,[0.675603176233325,-0.312493754568903,2.28671333346264])
+    end
+    should "report correct coefficientes as hash" do
+      assert_similar_hash(@glm.coefficients(:hash), {"i"=>0.675603176233325,"x1"=>-0.312493754568903,"x2"=>2.28671333346264})
+    end
+
+  end
+end

data/test/test_glm_poisson.rb

@@ -0,0 +1,25 @@
+require(File.expand_path(File.dirname(__FILE__)+'/helper.rb'))
+
+class StatsampleRegressionGlmPoisson < MiniTest::Unit::TestCase
+
+  context("Example") do
+    setup do
+      x1=Statsample::Vector.new([0.537322309644812,-0.717124209978434,-0.519166718891331,0.434970973986765,-0.761822002215759,1.51170030921189,0.883854199811195,-0.908689798854196,1.70331977539793,-0.246971150634099,-1.59077593922623,-0.721548040910253,0.467025703920194,-0.510132788447137,0.430106510266798,-0.144353683251536,-1.54943800728303,0.849307651309298,-0.640304240933579,1.31462478279425,-0.399783455165345,0.0453055645017902,-2.58212161987746,-1.16484414309359,-1.08829266466281,-0.243893919684792,-1.96655661929441,0.301335373291024,-0.665832694463588,-0.0120650855753837,1.5116066367604,0.557300353673344,1.12829931872045,0.234443748015922,-2.03486690662651,0.275544751380246,-0.231465849558696,-0.356880153225012,-0.57746647541923,1.35758352580655,1.23971669378224,-0.662466275100489,0.313263561921793,-1.08783223256362,1.41964722846899,1.29325100940785,0.72153880625103,0.440580131022748,0.0351917814720056, -0.142353224879252],:scale)
+      x2=Statsample::Vector.new([-0.866655707911859,-0.367820249977585,0.361486610435,0.857332626245179,0.133438466268095,0.716104533073575,1.77206093023382,-0.10136697295802,-0.777086491435508,-0.204573554913706,0.963353531412233,-1.10103024900542,-0.404372761837392,-0.230226345183469,0.0363730246866971,-0.838265540390497,1.12543549657924,-0.57929175648001,-0.747060244805248,0.58946979365152,-0.531952663697324,1.53338594419818,0.521992029051441,1.41631763288724,0.611402316795129,-0.518355638373296,-0.515192557101107,-0.672697937866108,1.84347042325327,-0.21195540664804,-0.269869371631611,0.296155694010096,-2.18097898069634,-1.21314663927206,1.49193669881581,1.38969280369493,-0.400680808117106,-1.87282814976479,1.82394870451051,0.637864732838274,-0.141155946382493,0.0699950644281617,1.32568550595165,-0.412599258349398,0.14436832227506,-1.16507785388489,-2.16782049922428,0.24318371493798,0.258954871320764,-0.151966534521183],:scale)
+      @y_log=Statsample::Vector.new([0,0,1,0,1,1,1,1,0,1,1,1,1,0,1,0,1,1,0,1,0,1,1,1,1,0,0,1,1,0,0,1,0,0,1,1,0,0,1,1,0,1,1,1,1,0,0,0,1,1],:scale)
+      @y_pois=Statsample::Vector.new([1,2,1,3,3,1,10,1,1,2,15,0,0,2,1,2,18,2,1,1,1,8,18,13,7,1,1,0,26,0,2,2,0,0,25,7,0,0,21,0,0,1,5,0,3,0,0,1,0,0],:scale)
+      intercept=Statsample::Vector.new([1]*50,:scale)
+      @df=Statsample::Dataset.new({"i"=>intercept,"x1"=>x1,"x2"=>x2})
+      @glm=Statsample::Regression.glm(@df,@y_pois,:poisson)
+
+    end
+    should "report correct coefficientes as array" do
+      assert_similar_vector(@glm.coefficients,[0.32993246633711,-0.586359358356708,1.28511323439258])
+    end # should
+    should "report correct coefficientes as hash" do
+      assert_similar_hash(@glm.coefficients(:hash), {"i"=>0.32993246633711,"x1"=>-0.586359358356708, "x2"=>1.28511323439258})
+    end # should
+
+  end # context
+end # class
+

metadata

@@ -0,0 +1,246 @@
+--- !ruby/object:Gem::Specification
+name: statsample-glm
+version: !ruby/object:Gem::Version
+  version: 0.0.1
+  prerelease: 
+platform: ruby
+authors:
+- Ankur Goel
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2013-09-25 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: statsample
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 1.2.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 1.2.0
+- !ruby/object:Gem::Dependency
+  name: activesupport
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.2.10
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.2.10
+- !ruby/object:Gem::Dependency
+  name: shoulda
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: rdoc
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '3.12'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '3.12'
+- !ruby/object:Gem::Dependency
+  name: minitest
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 4.7.5
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 4.7.5
+- !ruby/object:Gem::Dependency
+  name: cucumber
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: bundler
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.3.5
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.3.5
+- !ruby/object:Gem::Dependency
+  name: jeweler
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.8.4
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.8.4
+- !ruby/object:Gem::Dependency
+  name: bio
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 1.4.2
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 1.4.2
+- !ruby/object:Gem::Dependency
+  name: rdoc
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '3.12'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '3.12'
+- !ruby/object:Gem::Dependency
+  name: mocha
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.14.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.14.0
+description: Statsample-GLM is an extension to Statsample, an advance statistics suite
+  in Ruby. This gem includes modules for Regression techniques such as Poisson Regression,
+  Logistic Regression and Exponential Regression
+email: ankurgel@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files:
+- LICENSE.txt
+- README.rdoc
+files:
+- .document
+- .travis.yml
+- Gemfile
+- LICENSE.txt
+- README.rdoc
+- Rakefile
+- VERSION
+- features/bio-statsample-glm.feature
+- features/step_definitions/bio-statsample-glm_steps.rb
+- features/support/env.rb
+- lib/statsample-glm.rb
+- lib/statsample-glm/regression.rb
+- lib/statsample-glm/regression/logistic.rb
+- lib/statsample-glm/regression/poisson.rb
+- test/helper.rb
+- test/test_glm.rb
+- test/test_glm_logistic.rb
+- test/test_glm_poisson.rb
+homepage: http://github.com/AnkurGel/statsample-glm
+licenses:
+- MIT
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+      segments:
+      - 0
+      hash: 891920669
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 1.8.25
+signing_key: 
+specification_version: 3
+summary: Generalized Linear Models for Statsample
+test_files: []