statsample 0.7.0 → 0.8.0

data/test/test_statistics.rb

@@ -1,11 +1,19 @@
-$:.unshift(File.dirname(__FILE__)+'/../lib/')
-require 'statsample'
-require 'test/unit'
-class StatsampleStatisicsTestCase < Test::Unit::TestCase
+require(File.dirname(__FILE__)+'/test_helpers.rb')
+
+class StatsampleStatisicsTestCase < MiniTest::Unit::TestCase
 
   def initialize(*args)
     super
   end
+  def test_p_using_cdf
+    assert_equal(0.25, Statsample::Test.p_using_cdf(0.25, tails=:left))
+    assert_equal(0.75, Statsample::Test.p_using_cdf(0.25, tails=:right))
+    assert_equal(0.50, Statsample::Test.p_using_cdf(0.25, tails=:both))
+    assert_equal(1, Statsample::Test.p_using_cdf(0.50, tails=:both))
+    assert_equal(0.05, Statsample::Test.p_using_cdf(0.025, tails=:both))
+    assert_in_delta(0.05, Statsample::Test.p_using_cdf(0.975, tails=:both),0.0001)
+    
+  end
   def test_recode_repeated
     a=%w{a b c c d d d e}
     exp=["a","b","c_1","c_2","d_1","d_2","d_3","e"]
@@ -23,8 +31,6 @@ class StatsampleStatisicsTestCase < Test::Unit::TestCase
     assert(!"".is_number?)
 
   end
-
-
   def test_estimation_mean
     v=([42]*23+[41]*4+[36]*1+[32]*1+[29]*1+[27]*2+[23]*1+[19]*1+[16]*2+[15]*2+[14,11,10,9,7]+ [6]*3+[5]*2+[4,3]).to_vector(:scale)
     assert_equal(50,v.size)

data/test/test_stest.rb

@@ -1,9 +1,7 @@
-$:.unshift(File.dirname(__FILE__)+'/../lib/')
-require 'statsample'
-require "tempfile"
-require 'test/unit'
+require(File.dirname(__FILE__)+'/test_helpers.rb')
 
-class StatsampleStatisticalTestCase < Test::Unit::TestCase
+
+class StatsampleTestTestCase < MiniTest::Unit::TestCase
   def test_chi_square
     real=Matrix[[95,95],[45,155]]
     expected=Matrix[[68,122],[72,128]]
@@ -22,14 +20,24 @@ class StatsampleStatisticalTestCase < Test::Unit::TestCase
     b=[2,3,4,5,6,12].to_scale
     assert_equal(11, Statsample::Test.u_mannwhitney(a,b).u)
   end
-  
-  
+
+
   def test_levene
-    
     a=[1,2,3,4,5,6,7,8,100,10].to_scale
-    b=[30,40,50,60,70,80,90,100,110,120].to_scale    
+    b=[30,40,50,60,70,80,90,100,110,120].to_scale
     levene=Statsample::Test::Levene.new([a,b])
+    assert_levene(levene)
+  end
+  def test_levene_dataset
+    a=[1,2,3,4,5,6,7,8,100,10].to_scale
+    b=[30,40,50,60,70,80,90,100,110,120].to_scale
+    ds={'a'=>a,'b'=>b}.to_dataset
+    levene=Statsample::Test::Levene.new(ds)
+    assert_levene(levene)
+  end
+  def assert_levene(levene)
     assert_in_delta(0.778, levene.f, 0.001)
     assert_in_delta(0.389, levene.probability, 0.001)
   end
-end
+ 
+end

data/test/test_stratified.rb

@@ -1,19 +1,18 @@
-$:.unshift(File.dirname(__FILE__)+'/../lib/')
-require 'statsample'
-require 'test/unit'
+require(File.dirname(__FILE__)+'/test_helpers.rb')
 
-class StatsampleStratifiedTestCase < Test::Unit::TestCase
 
-	def initialize(*args)
-		super
-	end
-	def test_mean
-		a=[10,20,30,40,50]
-		b=[110,120,130,140]
-		pop=a+b
-		av=a.to_vector(:scale)
-		bv=b.to_vector(:scale)
-		popv=pop.to_vector(:scale)
-		assert_equal(popv.mean,Statsample::StratifiedSample.mean(av,bv))
-	end
+class StatsampleStratifiedTestCase < MiniTest::Unit::TestCase
+
+  def initialize(*args)
+    super
+  end
+  def test_mean
+    a=[10,20,30,40,50]
+    b=[110,120,130,140]
+    pop=a+b
+    av=a.to_vector(:scale)
+    bv=b.to_vector(:scale)
+    popv=pop.to_vector(:scale)
+    assert_equal(popv.mean,Statsample::StratifiedSample.mean(av,bv))
+  end
 end

data/test/test_svg_graph.rb

@@ -1,36 +1,28 @@
-$:.unshift(File.dirname(__FILE__)+'/../lib/')
-require 'statsample'
-require 'tmpdir'
-require 'tempfile'
-require 'tempfile'
-require 'fileutils'
-require 'test/unit'
-begin
-	require 'statsample/graph/svggraph'
-class StatsampleSvgGraphTestCase < Test::Unit::TestCase
+require(File.dirname(__FILE__)+'/test_helpers.rb')
+require('statsample/graph')
+class StatsampleSvgGraphTestCase < MiniTest::Unit::TestCase
 
-  def initialize(*args)
-      @image_path=Dir::tmpdir+"/images"
-      FileUtils.mkdir(@image_path) if !File.exists? @image_path 
-  super
+  def setup
+    @image_path=Dir::tmpdir+"/images"
+    FileUtils.mkdir(@image_path) if !File.exists? @image_path
   end
   def test_histogram
     if Statsample.has_gsl?
       ar=(1..1000).to_a.collect {|a|
-    rand(10)
-    }.to_vector(:scale)
+        rand(10)
+      }.to_vector(:scale)
       h=ar.histogram([0,2,5,11])
       file=Tempfile.new("svg_histogram_only.svg")
       graph = Statsample::Graph::SvgHistogram.new({})
       graph.histogram=h
       file.puts(graph.burn)
     else
-    puts "Statsample::Graph::SvgHistogram.new not tested (no ruby-gsl)"
+      puts "Statsample::Graph::SvgHistogram.new not tested (no ruby-gsl)"
     end
   end
   def assert_svg(msg=nil)
     msg||="%s isn't a svg file"
-    Tempfile.open("svg") do |fp| 
+    Tempfile.open("svg") do |fp|
       yield fp
       fp.close
       fp.open
@@ -49,7 +41,7 @@ class StatsampleSvgGraphTestCase < Test::Unit::TestCase
     assert_svg {|file| vector.svggraph_frequencies(file,800,600, SVG::Graph::PieNoOp, :graph_title=>'Pie') }
     vector.type=:scale
     if Statsample.has_gsl?
-    file=Tempfile.new("svg_histogram.svg").path
+      file=Tempfile.new("svg_histogram.svg").path
       hist=vector.svggraph_histogram(5)
       File.open(file,"wb") {|fp|
         fp.write(hist.burn)
@@ -60,6 +52,3 @@ class StatsampleSvgGraphTestCase < Test::Unit::TestCase
     end
   end
 end
-rescue LoadError
-	puts "You should install SVG::Graph"
-end

data/test/test_test_t.rb

@@ -0,0 +1,40 @@
+require(File.dirname(__FILE__)+'/test_helpers.rb')
+
+class StatsampleTestTTestCase < MiniTest::Unit::TestCase
+  def setup
+    @a=[30.02, 29.99, 30.11, 29.97, 30.01, 29.99].to_scale
+    @b=[29.89, 29.93, 29.72, 29.98, 30.02, 29.98].to_scale
+    @x1=@a.mean
+    @x2=@b.mean
+    @s1=@a.sd
+    @s2=@b.sd
+    @n1=@a.n
+    @n2=@b.n
+  end
+  def test_t_sample_independent_t
+    assert_in_delta(1.959, Statsample::Test::T.two_sample_independent(@x1, @x2, @s1, @s2, @n1, @n2))
+    assert_in_delta(1.959, Statsample::Test::T.two_sample_independent(@x1, @x2, @s1, @s2, @n1, @n2,true))
+  end
+  def test_t_sample_independent_df
+    assert_equal(10,  Statsample::Test::T.df_equal_variance(@n1,@n2))
+    assert_in_delta(7.03,  Statsample::Test::T.df_not_equal_variance(@s1,@s2,@n1,@n2))    
+  end
+  def test_t_sample_independent_object
+    t=Statsample::Test.t_two_samples_independent(@a,@b)
+    assert(t.summary.size>0)
+    assert_in_delta(1.959, t.t_equal_variance)
+    assert_in_delta(1.959, t.t_not_equal_variance)
+    assert_in_delta(10, t.df_equal_variance)
+    assert_in_delta(7.03, t.df_not_equal_variance)
+    assert_in_delta(0.07856, t.probability_equal_variance)
+    assert_in_delta(0.09095, t.probability_not_equal_variance)
+    
+  end
+  def test_t_one_sample
+    u=@a.mean+(1-rand*2)
+    tos=Statsample::Test::T::OneSample.new(@a,{:u=>u})
+    assert_equal((@a.mean-u).quo(@a.sd.quo(Math::sqrt(@a.n))), tos.t)
+    assert_equal(@a.n-1, tos.df)
+    assert(tos.summary.size>0)
+  end
+end

data/test/test_umannwhitney.rb

@@ -1,17 +1,16 @@
-$:.unshift(File.dirname(__FILE__)+'/../lib/')
-require 'statsample'
-require 'test/unit'
+require(File.dirname(__FILE__)+'/test_helpers.rb')
 
-class StatsampleSrsTestCase < Test::Unit::TestCase
-    def test_u_mannwhitney
-      v1=[1,2,3,4,7,8,9,10,14,15].to_scale
-      v2=[5,6,11,12,13,16,17,18,19].to_scale
-      u=Statsample::Test::UMannWhitney.new(v1,v2)
-      assert_equal(73,u.r1)
-      assert_equal(117,u.r2)
-      assert_equal(18,u.u)
-      assert_in_delta(-2.205,u.z,0.001)
-      assert_in_delta(0.027,u.z_probability,0.001)
-      assert_in_delta(0.028,u.exact_probability,0.001)
-    end
+
+class StatsampleUMannWhitneyTestCase < MiniTest::Unit::TestCase
+  def test_u_mannwhitney
+    v1=[1,2,3,4,7,8,9,10,14,15].to_scale
+    v2=[5,6,11,12,13,16,17,18,19].to_scale
+    u=Statsample::Test::UMannWhitney.new(v1,v2)
+    assert_equal(73,u.r1)
+    assert_equal(117,u.r2)
+    assert_equal(18,u.u)
+    assert_in_delta(-2.205,u.z,0.001)
+    assert_in_delta(0.027,u.z_probability,0.001)
+    assert_in_delta(0.028,u.exact_probability,0.001)
+  end
 end

data/test/test_vector.rb

@@ -1,22 +1,18 @@
-$:.unshift(File.dirname(__FILE__)+'/../lib/')
-require 'statsample'
-require 'test/unit'
-require 'tempfile'
-class TestStatsample
-end
-class TestStatsample::TestVector < Test::Unit::TestCase
-  
+require(File.dirname(__FILE__)+'/test_helpers.rb')
+
+class StatsampleTestVector < MiniTest::Unit::TestCase
+
   def setup
-  @c = Statsample::Vector.new([5,5,5,5,5,6,6,7,8,9,10,1,2,3,4,nil,-99,-99], :nominal)
-  @c.missing_values=[-99]
-  
+    @c = Statsample::Vector.new([5,5,5,5,5,6,6,7,8,9,10,1,2,3,4,nil,-99,-99], :nominal)
+    @c.missing_values=[-99]
+
   end
   def test_save_load
     outfile=Tempfile.new("vector.vec")
     @c.save(outfile.path)
     a=Statsample.load(outfile.path)
     assert_equal(@c,a)
-      
+
   end
   def test_lazy_methods
     data=[1,2,3,4,5,nil]
@@ -41,19 +37,19 @@ class TestStatsample::TestVector < Test::Unit::TestCase
     assert_equal(exp2,exp)
   end
   def test_product
-      a=[1,2,3,4,5].to_vector(:scale)
-      assert_equal(120,a.product)
+    a=[1,2,3,4,5].to_vector(:scale)
+    assert_equal(120,a.product)
   end
   def test_matrix
-      a=[1,2,3,4,5].to_vector(:scale)
-      mh=Matrix[[1,2,3,4,5]]
-      mv=Matrix.columns([[1,2,3,4,5]])
-      assert_equal(mh,a.to_matrix)
-      assert_equal(mv,a.to_matrix(:vertical))
-      # 3*4 + 2*5 = 22
-      a=[3,2].to_vector(:scale)
-      b=[4,5].to_vector(:scale)
-      assert_equal(22,(a.to_matrix*b.to_matrix(:vertical))[0,0])
+    a=[1,2,3,4,5].to_vector(:scale)
+    mh=Matrix[[1,2,3,4,5]]
+    mv=Matrix.columns([[1,2,3,4,5]])
+    assert_equal(mh,a.to_matrix)
+    assert_equal(mv,a.to_matrix(:vertical))
+    # 3*4 + 2*5 = 22
+    a=[3,2].to_vector(:scale)
+    b=[4,5].to_vector(:scale)
+    assert_equal(22,(a.to_matrix*b.to_matrix(:vertical))[0,0])
   end
   def test_missing_values
     @c.missing_values=[10]
@@ -149,10 +145,10 @@ class TestStatsample::TestVector < Test::Unit::TestCase
     assert_equal(5,@c.median)
     assert_equal(4,@c.percentil(25))
     assert_equal(7,@c.percentil(75))
-    
+
     v=[200000, 200000, 210000, 220000, 230000, 250000, 250000, 250000, 270000, 300000, 450000, 130000, 140000, 140000, 140000, 145000, 148000, 165000, 170000, 180000, 180000, 180000, 180000, 180000, 180000 ].to_scale
     assert_equal(180000,v.median)
-  
+
   end
   def test_ranked
     v1=[0.8,1.2,1.2,2.3,18].to_vector(:ordinal)
@@ -165,9 +161,9 @@ class TestStatsample::TestVector < Test::Unit::TestCase
     i=0
     factors=a.factors.sort
     [0,1,2,3,4].each{|v|
-        assert(v==factors[i])
-        assert(v.class==factors[i].class,"#{v} - #{v.class} != #{factors[i]} - #{factors[i].class}")
-        i+=1
+      assert(v==factors[i])
+      assert(v.class==factors[i].class,"#{v} - #{v.class} != #{factors[i]} - #{factors[i].class}")
+      i+=1
     }
   end
   def test_vector_standarized
@@ -234,7 +230,7 @@ class TestStatsample::TestVector < Test::Unit::TestCase
     srand(1)
     assert_equal(100, @c.sample_with_replacement(100).size)
     assert_equal(@c.valid_data.to_a.sort, @c.sample_without_replacement(15).sort)
-      
+
   end
   def test_valid_data
     a=Statsample::Vector.new([1,2,3,4,"STRING"])
@@ -273,7 +269,7 @@ class TestStatsample::TestVector < Test::Unit::TestCase
       a.set_valid_data
       assert_equal(3,a.sum)
       b=[1,2,nil,3,4,5,nil,6].to_vector(:scale)
-      assert_equal(21, b.sum)        
+      assert_equal(21, b.sum)
       assert_equal(3.5, b.mean)
       assert_equal(6,b.gsl.size)
       c=[10,20,30,40,50,100,1000,2000,5000].to_scale
@@ -281,7 +277,7 @@ class TestStatsample::TestVector < Test::Unit::TestCase
       assert_in_delta(c.kurtosis, c.kurtosis_slow ,0.0001)
     end
   end
-  def test_vector_matrix        
+  def test_vector_matrix
     v1=%w{a a a b b b c c}.to_vector
     v2=%w{1 3 4 5 6 4 3 2}.to_vector
     v3=%w{1 0 0 0 1 1 1 0}.to_vector
@@ -299,14 +295,14 @@ class TestStatsample::TestVector < Test::Unit::TestCase
     assert_equal(v1.data,v2.data)
     assert_not_same(v1.data,v2.data)
     assert_equal(v1.type,v2.type)
-    
+
     v1.type=:ordinal
     assert_not_equal(v1.type,v2.type)
     assert_equal(v1.missing_values,v2.missing_values)
     assert_not_same(v1.missing_values,v2.missing_values)
     assert_equal(v1.labels,v2.labels)
     assert_not_same(v1.labels,v2.labels)
-    
+
     v3=v1.dup_empty
     assert_equal([],v3.data)
     assert_not_equal(v1.data,v3.data)
@@ -331,7 +327,7 @@ class TestStatsample::TestVector < Test::Unit::TestCase
     assert_equal(exp,a.dichotomize)
     a=  [1,1,1,2,2,2,3].to_vector
     exp=[0,0,0,1,1,1,1].to_scale
-    assert_equal(exp,a.dichotomize)      
+    assert_equal(exp,a.dichotomize)
     a=  [0,0,0,1,2,3,nil].to_vector
     exp=[0,0,0,0,1,1,nil].to_scale
     assert_equal(exp,a.dichotomize(1))
@@ -346,15 +342,15 @@ class TestStatsample::TestVector < Test::Unit::TestCase
     assert(!a.can_be_scale?)
     a.missing_values=["s"]
     assert(a.can_be_scale?)
-    
+
     a=[Date.new(2009,10,10), Date.today(), "2009-10-10", "2009-1-1", nil, "NOW"].to_vector
     assert(a.can_be_date?)
     a=[Date.new(2009,10,10), Date.today(),nil,"sss"].to_vector
-    assert(!a.can_be_date?)      
+    assert(!a.can_be_date?)
   end
   def test_date_vector
     a=[Date.new(2009,10,10), :NOW, "2009-10-10", "2009-1-1", nil, "NOW","MISSING"].to_vector(:date, :missing_values=>["MISSING"])
-    
+
     assert(a.type==:date)
     expected=[Date.new(2009,10,10), Date.today(), Date.new(2009,10,10), Date.new(2009,1,1), nil, Date.today(), nil ]
     assert_equal(expected, a.date_data_with_nils)

data/test/test_xls.rb

@@ -1,42 +1,35 @@
-$:.unshift(File.dirname(__FILE__)+'/../lib/')
-require 'statsample'
-require 'test/unit'
-require 'tempfile'
-begin
-	require 'spreadsheet'
-rescue LoadError
-	puts "You should install spreadsheet (gem install spreadsheet)"
-end
-class StatsampleExcelTestCase < Test::Unit::TestCase
-    def setup
-        @ds=Statsample::Excel.read(File.dirname(__FILE__)+"/test_xls.xls")
-    end    
-    def test_read
-        assert_equal(6,@ds.cases)
-        assert_equal(%w{id name age city a1},@ds.fields)
-        id=[1,2,3,4,5,6].to_vector(:scale)
-        name=["Alex","Claude","Peter","Franz","George","Fernand"].to_vector(:nominal)
-        age=[20,23,25,nil,5.5,nil].to_vector(:scale)
-        city=["New York","London","London","Paris","Tome",nil].to_vector(:nominal)
-        a1=["a,b","b,c","a",nil,"a,b,c",nil].to_vector(:nominal)
-        ds_exp=Statsample::Dataset.new({'id'=>id,'name'=>name,'age'=>age,'city'=>city,'a1'=>a1}, %w{id name age city a1})
-        ds_exp.fields.each{|f|
-            assert_equal(ds_exp[f],@ds[f])
-        }
-        assert_equal(ds_exp,@ds)
-        
-    end
-    def test_nil
-        assert_equal(nil,@ds['age'][5])
+require(File.dirname(__FILE__)+'/test_helpers.rb')
+
+class StatsampleExcelTestCase < MiniTest::Unit::TestCase
+  def setup
+    @ds=Statsample::Excel.read(File.dirname(__FILE__)+"/test_xls.xls")
+  end
+  def test_read
+    assert_equal(6,@ds.cases)
+    assert_equal(%w{id name age city a1},@ds.fields)
+    id=[1,2,3,4,5,6].to_vector(:scale)
+    name=["Alex","Claude","Peter","Franz","George","Fernand"].to_vector(:nominal)
+    age=[20,23,25,nil,5.5,nil].to_vector(:scale)
+    city=["New York","London","London","Paris","Tome",nil].to_vector(:nominal)
+    a1=["a,b","b,c","a",nil,"a,b,c",nil].to_vector(:nominal)
+    ds_exp=Statsample::Dataset.new({'id'=>id,'name'=>name,'age'=>age,'city'=>city,'a1'=>a1}, %w{id name age city a1})
+    ds_exp.fields.each{|f|
+      assert_equal(ds_exp[f],@ds[f])
+    }
+    assert_equal(ds_exp,@ds)
+
+  end
+  def test_nil
+    assert_equal(nil,@ds['age'][5])
+  end
+  def test_write
+    tempfile=Tempfile.new("test_write.xls")
+    Statsample::Excel.write(@ds,tempfile.path)
+    ds2=Statsample::Excel.read(tempfile.path)
+    i=0
+    ds2.each_array do |row|
+      assert_equal(@ds.case_as_array(i),row)
+      i+=1
     end
-    def test_write
-      tempfile=Tempfile.new("test_write.xls")
-        Statsample::Excel.write(@ds,tempfile.path)
-        ds2=Statsample::Excel.read(tempfile.path)
-        i=0
-        ds2.each_array{|row|
-            assert_equal(@ds.case_as_array(i),row)
-               i+=1
-        }
-    end
-end
+  end
+end