bio-statsample-glm 0.1.0 → 0.1.1

data/VERSION

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

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

@@ -22,14 +22,17 @@ module Statsample
     #
     # == Returns
     #    GLM object for given method.
-    def self.glm(x, y, method=:poisson)
+    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
-      #now, #irwls method is available to be called on returned obj
     end
 
 

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

@@ -4,9 +4,6 @@ module Statsample
 
       class Logistic
 
-        # a named vector of coefficients
-        attr_reader :coefficients
-        #
         attr_reader :se
         # The fitted mean values
         attr_reader :fit
@@ -19,11 +16,28 @@ module Statsample
         # Boolean. Tells whether the IRWLS for the given model converged or not
         attr_reader :converged
 
-        def initialize(x, y)
-          @x = x
+        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) }
@@ -91,4 +105,4 @@ module Statsample
 
     end
   end
-end
+end

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

@@ -4,9 +4,6 @@ module Statsample
 
       class Poisson
 
-        # a named vector of coefficients
-        attr_reader :coefficients
-        #
         attr_reader :se
         # The fitted mean values
         attr_reader :fit
@@ -19,11 +16,28 @@ module Statsample
         # Boolean. Tells whether the IRWLS for the given model converged or not
         attr_reader :converged
 
-        def initialize(x, y)
-          @x = x
+        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) }

data/test/helper.rb

@@ -42,6 +42,14 @@ module MiniTest
         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|

data/test/test_glm.rb

@@ -1,37 +1,4 @@
 require(File.expand_path(File.dirname(__FILE__)+'/helper.rb'))
 
 class StatsampleRegressionGlm < 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})
-    end
-
-    context("Logistic") do
-      setup do
-        @glm=Statsample::Regression.glm(@df.to_matrix,@y_log,:binomial)
-        @glm.irwls
-      end
-
-      should "report correct coefficientes" do
-        assert_similar_vector(@glm.coefficients,[0.675603176233325,-0.312493754568903,2.28671333346264])
-      end
-    end
-
-    context("Poisson") do
-      setup do
-        @glm=Statsample::Regression.glm(@df.to_matrix,@y_pois,:poisson)
-        @glm.irwls
-      end
-
-      should "report correct coefficientes" do
-        assert_similar_vector(@glm.coefficients,[0.32993246633711,-0.586359358356708,1.28511323439258])
-      end
-    end
-  end
 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

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: bio-statsample-glm
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
   prerelease: 
 platform: ruby
 authors:
@@ -215,6 +215,8 @@ files:
 - lib/bio-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/bioruby-statsample-glm
 licenses:
 - MIT
@@ -230,7 +232,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: 797002845
+      hash: -988046223
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements: