rubystats 0.2.2 → 0.2.3

data/History.txt

@@ -1,3 +1,7 @@
+=== 0.2.3 / 2008-07-06
+* Fixing bug #21100 - problem with Beta distribution calculations when p and q values are small.
+* Minor code cleanup for readability in a few places.
+
 === 0.2.2 / 2008-05-09
 * Added exponential distribution class
 * Replaced constants in NumericalConstants module with uppercase versions to follow Ruby convention  

data/README.txt

@@ -49,11 +49,11 @@ Also includes Fisher's Exact Test
 == SYNOPSIS:
 === Example: normal distribution with mean of 10 and standard deviation of 2
 
-norm = Rubystats::NormalDistribution.new(10, 2)
-cdf = norm.cdf(11)
-pdf = norm.pdf(11)
-puts "CDF(11): #{cdf}"
-puts "PDF(11): #{pdf}"
+ norm = Rubystats::NormalDistribution.new(10, 2)
+ cdf = norm.cdf(11)
+ pdf = norm.pdf(11)
+ puts "CDF(11): #{cdf}"
+ puts "PDF(11): #{pdf}"
 
 Output:
  CDF(11): 0.691462461274013
@@ -61,10 +61,10 @@ Output:
 
 === Example: get some random numbers from a normal distribution
 
-puts "Random numbers from normal distribution:"
-10.times do
-  puts norm.rng
-end
+ puts "Random numbers from normal distribution:"
+ 10.times do
+   puts norm.rng
+ end
 
 (sample) Output:
 

data/Rakefile

@@ -4,6 +4,7 @@ require 'rubygems'
 require 'hoe'
 $:.unshift(File.dirname(__FILE__) + "/lib")
 require 'rubystats'
+#require './lib/rubystats.rb'
 
 Hoe.new('rubystats', Rubystats::VERSION) do |p|
   p.name = "rubystats"
@@ -16,10 +17,15 @@ Hoe.new('rubystats', Rubystats::VERSION) do |p|
   p.remote_rdoc_dir = '' # Release to root
 end
 
-rule '' do |t|
-  system "cd test && ruby ts_stats.rb"
+desc "Run all tests"
+task :test  do 
+  Rake::TestTask.new("test") do |t|
+    t.test_files = FileList['test/tc*.rb']
+    t.verbose = false
+  end
 end
 
+#Use Allison template if available
 ALLISON = "/opt/local/lib/ruby/gems/1.8/gems/allison-2.0.3/lib/allison.rb"
 
 Rake::RDocTask.new do |rd|  

data/lib/rubystats.rb

@@ -1,5 +1,5 @@
 module Rubystats
-  VERSION = '0.2.2'
+  VERSION = '0.2.3'
 end
 
 require 'rubystats/normal_distribution'

data/lib/rubystats/beta_distribution.rb

@@ -10,7 +10,7 @@ module Rubystats
     #dgr_q = degrees of freedom q
     def initialize(dgr_p, dgr_q)
       if dgr_p <= 0 || dgr_q <= 0
-        return nil
+        raise ArgumentError.new("Paramters must be greater than zero.")
       end
       @p = dgr_p.to_f
       @q = dgr_q.to_f
@@ -41,8 +41,9 @@ module Rubystats
         if  (x == 0.0) || (x == 1.0)  
           return 0.0
         else 
-          return Math.exp( - log_beta(@p, @q) + (@p - 1.0) * Math.log(x) + (@q - 1.0) * Math.log(1.0 - x)
-          )
+          return Math.exp( -1.0 * log_beta(@p, @q) + 
+          (@p - 1.0) * Math.log(x) + 
+          (@q - 1.0) * Math.log(1.0 - x))
         end
       end
     end

data/lib/rubystats/modules.rb

@@ -36,26 +36,26 @@ module Rubystats
 
     include Rubystats::NumericalConstants
 
-    @logGAMMACache_res = 0.0
-    @logGAMMACache_x   = 0.0
-    @logBetaCache_res  = 0.0
-    @logBetaCache_p    = 0.0
-    @logBetaCache_q    = 0.0
+    @log_gamma_cache_res = 0.0
+    @log_gamma_cache_x   = 0.0
+    @log_beta_cache_res  = 0.0
+    @log_beta_cache_p    = 0.0
+    @log_beta_cache_q    = 0.0
 
     def log_beta(p,q)
-      if p != @logBetaCache_p || q != @logBetaCache_q 
-        logBetaCache_p = p
-        logBetaCache_q = q
-        if p <= 0.0 || q <= 0.0 || (p + q) > LOG_GAMMA_X_MAX_VALUE
-          logBetaCache_res = 0.0
+      if p != @log_beta_cache_p || q != @log_beta_cache_q 
+        @log_beta_cache_p = p
+        @log_beta_cache_q = q
+        if (p <= 0.0) || (q <= 0.0) || (p + q) > LOG_GAMMA_X_MAX_VALUE
+          @log_beta_cache_res = 0.0
         else
-          logBetaCache_res = log_gamma(p) + log_gamma(q) - log_gamma(p + q)
+          @log_beta_cache_res = log_gamma(p) + log_gamma(q) - log_gamma(p + q)
         end
-        return logBetaCache_res
       end
+      return @log_beta_cache_res
     end
 
-    # GAMMA function.
+    # Gamma function.
     # Based on public domain NETLIB (Fortran) code by W. J. Cody and L. Stoltz<BR>
     # Applied Mathematics Division<BR>
     # Argonne National Laboratory<BR>
@@ -68,9 +68,9 @@ module Rubystats
     # </OL></P><P>
     # From the original documentation:
     # </P><P>
-    # This routine calculates the GAMMA function for a real argument X.
+    # This routine calculates the Gamma function for a real argument X.
     # Computation is based on an algorithm outlined in reference 1.
-    # The program uses rational functions that approximate the GAMMA
+    # The program uses rational functions that approximate the Gamma
     # function to at least 20 significant decimal digits.  Coefficients
     # for the approximation over the interval (1,2) are unpublished.
     # Those for the approximation for X .GE. 12 are from reference 2.
@@ -84,8 +84,8 @@ module Rubystats
     # </P>
     # Author:: Jaco van Kooten
 
-    def GAMMA(x) 
-      # GAMMA related constants
+    def orig_gamma(x) 
+      # Gamma related constants
       g_p = [ -1.71618513886549492533811, 24.7656508055759199108314,
         -379.804256470945635097577, 629.331155312818442661052,
         866.966202790413211295064, -31451.2729688483675254357,
@@ -97,11 +97,11 @@ module Rubystats
       g_c = [-0.001910444077728, 8.4171387781295e-4, -5.952379913043012e-4,
         7.93650793500350248e-4, -0.002777777777777681622553,
         0.08333333333333333331554247, 0.0057083835261 ]
-      fact=1.0
-      i=0
-      n=0
-      y=x
-      parity=false
+      fact = 1.0
+      i = 0
+      n = 0
+      y = x
+      parity = false
       if y <= 0.0 
         # ----------------------------------------------------------------------
         #  Argument is negative
@@ -202,10 +202,31 @@ module Rubystats
       end
     end
 
-    def log_gamma(x)
-      logGAMMACache_res = @logGAMMACache_res
-      logGAMMACache_x = @logGAMMACache_x
+    #TODO test this
+    def gamma(_x)
+      return 0 if _x == 0.0
+
+      p0 = 1.000000000190015
+      p = {1 => 76.18009172947146,
+      2 => -86.50532032941677,
+      3 => 24.01409824083091,
+      4 => -1.231739572450155,
+      5 => 1.208650973866179e-3,
+      6 => -5.395239384953e-6}
+
+      y = x = _x
+      tmp = x + 5.5
+      tmp -= (x + 0.5) * Math.log(tmp)
+
+      summer = p0
+      for j in (1 ... 6)
+        y += 1
+        summer += (p[j] / y)
+      end
+      return Math.exp(0 - tmp + Math.log(2.5066282746310005 * summer / x))
+    end
 
+    def log_gamma(x)
       lg_d1 = -0.5772156649015328605195174
       lg_d2 = 0.4227843350984671393993777
       lg_d4 = 1.791759469228055000094023
@@ -251,12 +272,12 @@ module Rubystats
         -0.002777777777777681622553, 0.08333333333333333331554247,
         0.0057083835261 ]
 
-      # Rough estimate of the fourth root of logGAMMA_xBig
+      # Rough estimate of the fourth root of logGamma_xBig
       lg_frtbig = 2.25e76
       pnt68     = 0.6796875
 
-      if x == logGAMMACache_x
-        return logGAMMACache_res
+      if x == @log_gamma_cache_x
+        return @log_gamma_cache_res
       end
 
       y = x
@@ -280,7 +301,7 @@ module Rubystats
               xnum = xnum * xm1 + lg_p1[i]
               xden = xden * xm1 + lg_q1[i]
             end
-            res = corr * xm1 * (lg_d1 + xm1 * (xnum / xden))
+            res = corr + xm1 * (lg_d1 + xm1 * (xnum / xden))
           else
             xm2 = y - 1.0
             xden = 1.0
@@ -331,13 +352,13 @@ module Rubystats
         res = MAX_VALUE
       end
       # final adjustments and return
-      logGAMMACache_x = x
-      logGAMMACache_res = res
+      @log_gamma_cache_x = x
+      @log_gamma_cache_res = res
       return res
     end
 
 
-    # Incomplete GAMMA function.
+    # Incomplete Gamma function.
     # The computation is based on approximations presented in 
     # Numerical Recipes, Chapter 6.2 (W.H. Press et al, 1992).
     # @param a require a>=0
@@ -345,18 +366,18 @@ module Rubystats
     # @return 0 if x<0, a<=0 or a>2.55E305 to avoid errors and over/underflow
     # @author Jaco van Kooten
 
-    def incomplete_GAMMA(a, x) 
+    def incomplete_gamma(a, x) 
       if x <= 0.0 || a <= 0.0 || a > LOG_GAMMA_X_MAX_VALUE
         return 0.0
       elsif x < (a + 1.0)
-        return GAMMA_series_expansion(a, x)
+        return gamma_series_expansion(a, x)
       else
-        return 1.0-GAMMA_fraction(a, x)
+        return 1.0-gamma_fraction(a, x)
       end
     end
 
     # Author:: Jaco van Kooten
-    def GAMMA_series_expansion(a, x)
+    def gamma_series_expansion(a, x)
       ap  = a
       del = 1.0 / a
       sum = del
@@ -372,7 +393,7 @@ module Rubystats
     end
 
     # Author:: Jaco van Kooten
-    def GAMMA_fraction(a, x) 
+    def gamma_fraction(a, x) 
       b  = x + 1.0 - a
       c  = 1.0 / XMININ
       d  = 1.0 / b
@@ -424,14 +445,13 @@ module Rubystats
         return 0.0
       elsif x >= 1.0
         return 1.0
-      elsif p <= 0.0 || q <= 0.0 || (p + q) > LOG_GAMMA_X_MAX_VALUE
+      elsif (p <= 0.0) || (q <= 0.0) || (p + q) > LOG_GAMMA_X_MAX_VALUE
         return 0.0
       else 
         beta_gam = Math.exp( -log_beta(p, q) + p * Math.log(x) + q * Math.log(1.0 - x) )
         if x < (p + 1.0) / (p + q + 2.0)
           return beta_gam * beta_fraction(x, p, q) / p
         else
-          beta_frac = beta_fraction(1.0 - x, p, q)
           return 1.0 - (beta_gam * beta_fraction(1.0 - x, q, p) / q)
         end
       end
@@ -469,7 +489,7 @@ module Rubystats
         if c.abs < XMININ
           c = XMININ
         end
-        frac *= h * c
+        frac *= (h * c)
         # odd index for d
         d = -(p + m) * (sum_pq + m) * x / ((p + m2) * (p_plus + m2))
         h = 1.0 + d * h

data/lib/rubystats/probability_distribution.rb

@@ -140,7 +140,7 @@ module Rubystats
       x 		= guess
       x_new = guess
       error = 0.0
-      pdf 	= 0.0
+      _pdf 	= 0.0
       dx 		= 1000.0
       i 		= 0
       while ( dx.abs > accuracy && (i += 1) < max_iteration )
@@ -151,15 +151,15 @@ module Rubystats
         else
           x_hi = x
         end
-        pdf = pdf(x)
-        if pdf != 0.0
-          dx = error / pdf
+        _pdf = pdf(x)
+        if _pdf != 0.0
+          dx = error / _pdf
           x_new = x -dx
         end
         # If the NR fails to converge (which for example may be the 
         # case if the initial guess is too rough) we apply a bisection
         # step to determine a more narrow interval around the root.
-        if  x_new < x_lo || x_new > x_hi || pdf == 0.0
+        if  x_new < x_lo || x_new > x_hi || _pdf == 0.0
           x_new = (x_lo + x_hi) / 2.0
           dx = x_new - x
         end

data/test/tc_beta.rb

@@ -40,9 +40,26 @@ class TestBeta < Test::Unit::TestCase
 			inv_cdf = beta.icdf(cdf)
 			assert_in_delta(x, inv_cdf, 0.00000001)
 		end
-
 	end
 
+  def test_low_p_and_q_values
+		p = 1.5 
+		q = 1.0 
+		beta = Rubystats::BetaDistribution.new(p,q)
+
+    #from PHPExcel/PHPMath output
+    expected_icdf_vals = [
+    0.096548938, 0.153261886, 0.200829885, 0.243288080, 0.282310809, 0.318797571, 0.353302082, 0.386195754, 0.417742995, 0.448140475, 0.477539492, 0.506059599, 0.533797339, 0.560832096, 0.587230146, 0.613047546, 0.638332246, 0.663125670, 0.687463910, 0.711378661, 0.734897945, 0.758046692, 0.780847206, 0.803319540, 0.825481812, 0.847350458, 0.868940446, 0.890265460, 0.911338047, 0.932169752, 0.952771225, 0.973152319, 0.993322173]
+    expected_cdf_vals = [
+    0.005196152, 0.014696938, 0.027000000, 0.041569219, 0.058094750, 0.076367532, 0.096234090, 0.117575508, 0.140296115, 0.164316767, 0.189570567, 0.216000000, 0.243554922, 0.272191109, 0.301869177, 0.332553755, 0.364212850, 0.396817338, 0.430340563, 0.464758002, 0.500046998, 0.536186535, 0.573157047, 0.610940259, 0.649519053, 0.688877348, 0.729000000, 0.769872717, 0.811481978, 0.853814968, 0.896859521, 0.940604061, 0.985037563 ]
+    i = 0
+    0.03.step(1.0, 0.03) do |x|
+      assert_in_delta expected_icdf_vals[i], beta.icdf(x), 0.00000001
+      assert_in_delta expected_cdf_vals[i], beta.cdf(x), 0.00000001
+      i += 1
+    end
+  end
+
 	def test_control_limits
 		trials = 50
 		alpha = 0.05

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: rubystats
 version: !ruby/object:Gem::Version 
-  version: 0.2.2
+  version: 0.2.3
 platform: ruby
 authors: 
 - Bryan Donovan - http://www.bryandonovan.com
@@ -9,17 +9,18 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2008-05-09 00:00:00 -07:00
+date: 2008-07-06 00:00:00 -07:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
   name: hoe
+  type: :development
   version_requirement: 
   version_requirements: !ruby/object:Gem::Requirement 
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
-        version: 1.5.1
+        version: 1.7.0
     version: 
 description: Ruby Stats is a port of the statistics libraries from PHPMath. Probability distributions include binomial, beta, and normal distributions with PDF, CDF and inverse CDF as well as Fisher's Exact Test.
 email: b.dondo+rubyforge@gmail.com
@@ -80,7 +81,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
 requirements: []
 
 rubyforge_project: rubystats
-rubygems_version: 1.1.1
+rubygems_version: 1.2.0
 signing_key: 
 specification_version: 2
 summary: Port of PHPMath to Ruby