aliastable 3.1.0 → 4.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: daff6ffd485a6eabe1a2341625be2ffd7b43e821b66d6b0a50363f5822f22cdd
-  data.tar.gz: cffa205ec028890811239ab49dfaf36111d53c749a8b90971fd658da5ad0fe5c
+  metadata.gz: 88cae2e262098e88436d41789056946a36ec55270d4e52c33242f5018b687595
+  data.tar.gz: e6b9c4befb4950137807729b171178b0210c794693a1045dbd94ea10606aa593
 SHA512:
-  metadata.gz: f227b5d8d874414e8ff7ab1dea7295cf1cd212f9a7cdebd8df90cb2e74ff95ff07bda966df03ccc205abdf13397990de16988ccc011d9a025ab2964ac3e08a5e
-  data.tar.gz: 4b2742b75bc5e1d8c089bcf653bb7b37c2c0a6757a754d97cace15b1b13a1dc2744dbc294641bf4790d238a7e9832d3f01c031d0f74fad043086fef251cb1148
+  metadata.gz: 294e9171f62c9602f1eb3fe9886daeb1fb899431d906e26a33d95c0af4398bad147a62b7e2eee8f7d52a736dd4aa57a67bbf2e5b767ed847817eae59409b7bcf
+  data.tar.gz: 545e97af416b1d7ad20f81444811343eafb0d0ec74a30ce2b5b754cfb4336d8aefb337b907fa4b1b7453da796b5d4b48da941b9840bfeb878c3bff4bc7f614e8

data/aliastable.gemspec

@@ -1,27 +1,20 @@
 # -*- ruby -*-
-_VERSION = "3.1.0"
+_VERSION = "4.0.0"
 
 Gem::Specification.new do |s|
   s.name = "aliastable"
   s.version = _VERSION
-  s.date = "2018-01-16"
+  s.date = "2023-01-21"
   s.summary = "Efficiently generate random outcomes from an arbitrary categorical distribution."
   s.email = "pjs@alum.mit.edu"
   s.description = "If a categorical distribution has k distinct values, traditional approaches will require O(k) work to pick an outcome with the correct probabilities.  This algorithm uses conditional probability to construct a table which will yield outcomes with the correct probabilities. Table generation requires O(k) time, but subsequent generation is done in O(1) time."
   s.author = "Paul J Sanchez"
+  s.homepage = "https://bitbucket.org/paul_j_sanchez/aliastable"
   s.files = %w[
     aliastable.gemspec
     lib/aliastable.rb
-    Rakefile
-    test/infile.bad.1
-    test/infile.bad.2
-    test/infile.bad.3
-    test/infile.good.1
-    test/infile.good.2
-    test/infile.good.3
-    test/infile.good.4
-    test/test_alias.rb
   ]
-  s.required_ruby_version = '>= 2.0.0'
-  s.license = 'LGPL'
+  s.required_ruby_version = '>= 2.6.0'
+  s.license = 'MIT'
+  s.metadata["homepage_uri"] = s.homepage
 end

data/lib/aliastable.rb

@@ -9,6 +9,7 @@
 # probabilities.
 #
 class AliasTable
+  include Enumerable
   # Construct an alias table from a set of values and their associated
   # probabilities.  Values and their probabilities must be synchronized,
   # i.e., they must be arrays of the same length.  Values can be
@@ -28,34 +29,52 @@ class AliasTable
   def initialize(x_set, p_value)
     fail 'x_set & p_value must have same length.' if x_set.size != p_value.size
     fail 'p_values must be positive' unless p_value.all? { |value| value > 0 }
-    @p_primary = p_value.map(&:rationalize)
-    fail 'p_values must sum to 1' unless @p_primary.reduce(:+) == Rational(1)
-    @x = x_set.clone.freeze
-    @alias = Array.new(@x.length)
-    parity = Rational(1, @x.length)
-    group = @p_primary.each_index.group_by { |i| @p_primary[i] <=> parity }
+    p_primary = p_value.map(&:rationalize)
+    fail 'p_values must sum to 1' unless p_primary.reduce(:+) == Rational(1)
+    x = x_set.clone.freeze
+    len = x.length
+    col_alias = Array.new(len)
+    parity = Rational(1, len)
+    group = p_primary.each_index.group_by { |i| p_primary[i] <=> parity }
     deficit_set = group[-1]
     surplus_set = group[1]
-    until deficit_set.empty?
-      deficit = deficit_set.pop
-      surplus = surplus_set.pop
-      @p_primary[surplus] -= parity - @p_primary[deficit]
-      @p_primary[deficit] /= parity
-      @alias[deficit] = @x[surplus]
-      if @p_primary[surplus] == parity
-        @p_primary[surplus] = Rational(1)
-      else
-        (@p_primary[surplus] < parity ? deficit_set : surplus_set) << surplus
+    if deficit_set.nil?
+      @enum = Enumerator.new { |y| loop { y << x[rand(len)] } }.lazy
+    else
+      until deficit_set.empty?
+        deficit = deficit_set.pop
+        surplus = surplus_set.pop
+        p_primary[surplus] -= parity - p_primary[deficit]
+        p_primary[deficit] /= parity
+        col_alias[deficit] = x[surplus]
+        if p_primary[surplus] == parity
+          p_primary[surplus] = Rational(1)
+        else
+          (p_primary[surplus] < parity ? deficit_set : surplus_set) << surplus
+        end
       end
+      @enum = Enumerator.new do |y|
+        loop do
+          column = rand(len)
+          y << ((rand <= p_primary[column]) ? x[column] : col_alias[column])
+        end
+      end.lazy
     end
   end
 
-  # Return a random outcome from this object's distribution.
-  # The generate method is O(1) time, but is not an inversion
+  def each(&block)
+    @enum.each(&block)
+  end
+
+  # Return a random outcome from this object's distribution. The
+  # next (aka generate) method is O(1) time, but is not an inversion
   # since two uniforms are used for each value that gets generated.
+  # The exception is that when all probabilities are equal, it is
+  # a true inversion.
   #
-  def generate
-    column = rand(@x.length)
-    rand <= @p_primary[column] ? @x[column] : @alias[column]
+  def next
+    @enum.next
   end
+
+  alias generate next   # for backwards compatibility
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: aliastable
 version: !ruby/object:Gem::Version
-  version: 3.1.0
+  version: 4.0.0
 platform: ruby
 authors:
 - Paul J Sanchez
-autorequire: 
+autorequire:
 bindir: bin
 cert_chain: []
-date: 2018-01-16 00:00:00.000000000 Z
+date: 2023-01-21 00:00:00.000000000 Z
 dependencies: []
 description: If a categorical distribution has k distinct values, traditional approaches
   will require O(k) work to pick an outcome with the correct probabilities.  This
@@ -20,22 +20,14 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- Rakefile
 - aliastable.gemspec
 - lib/aliastable.rb
-- test/infile.bad.1
-- test/infile.bad.2
-- test/infile.bad.3
-- test/infile.good.1
-- test/infile.good.2
-- test/infile.good.3
-- test/infile.good.4
-- test/test_alias.rb
-homepage: 
+homepage: https://bitbucket.org/paul_j_sanchez/aliastable
 licenses:
-- LGPL
-metadata: {}
-post_install_message: 
+- MIT
+metadata:
+  homepage_uri: https://bitbucket.org/paul_j_sanchez/aliastable
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -43,16 +35,15 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: 2.0.0
+      version: 2.6.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: 
-rubygems_version: 2.7.4
-signing_key: 
+rubygems_version: 3.4.3
+signing_key:
 specification_version: 4
 summary: Efficiently generate random outcomes from an arbitrary categorical distribution.
 test_files: []

data/Rakefile

@@ -1,8 +0,0 @@
-require 'rake/testtask'
-
-Rake::TestTask.new do |t|
-  t.libs << 'test'
-end
-
-desc "Run tests"
-task :default => :test

data/test/infile.bad.1

@@ -1,5 +0,0 @@
-1,0.1
-3,0.2
-7,-0.3
-42,0.4
-8096,0.6

data/test/infile.bad.2

@@ -1,4 +0,0 @@
-1,0.1
-3,0.2
-7,0.3
-42,0.3

data/test/infile.bad.3

@@ -1,3 +0,0 @@
-1,0.333333
-3,0.333333
-7,0.333333

data/test/infile.good.1

@@ -1,4 +0,0 @@
-1,0.1
-3,0.2
-7,0.3
-42,0.4

data/test/infile.good.2

@@ -1,4 +0,0 @@
-aardvark,0.01
-baboon,0.02
-coati,0.07
-doggie,0.9

data/test/infile.good.3

@@ -1,22 +0,0 @@
-1,1/253
-2,2/253
-3,3/253
-4,4/253
-5,5/253
-6,6/253
-7,7/253
-8,8/253
-9,9/253
-10,10/253
-11,11/253
-12,12/253
-13,13/253
-14,14/253
-15,15/253
-16,16/253
-17,17/253
-18,18/253
-19,19/253
-20,20/253
-21,21/253
-22,22/253

data/test/infile.good.4

@@ -1,3 +0,0 @@
-1,0.333333
-3,0.333333
-7,0.333334

data/test/test_alias.rb

@@ -1,44 +0,0 @@
-#!/usr/bin/env ruby -w
-
-require_relative '../lib/aliastable.rb'
-
-nvars = 1_000_000
-begin
-  at = AliasTable.new(%w(yes no), [0.3, 0.3, 0.4])
-  nvars.times { print at.generate, "\n" }
-rescue RuntimeError => e
-  p e
-  puts
-end
-Dir['test/infile.*'].each do |f_name|
-  x = []
-  probs = []
-  f = File.open(f_name, 'r')
-  counts = {}
-  expected_counts = {}
-  while line = f.gets
-    inputs = line.strip.split(/[\s,;:]+/)
-    x << inputs[0]
-    counts[inputs[0]] = 0
-    probs << Rational(inputs[1])
-    n_hat = probs[-1] * nvars
-    half_width = 2.5 * Math.sqrt(n_hat * (1.0 - probs[-1])) if n_hat > 0
-    expected_counts[inputs[0]] = [n_hat, half_width]
-  end
-  f.close
-  p x
-  p probs
-  begin
-    at = AliasTable.new(x, probs)
-    nvars.times { counts[at.generate] += 1 }
-    puts 'All values should be in range almost always:'
-    counts.each_key do |k|
-      printf "%s: Allowable Range = %d, Expected - Observed = %d\n",
-             k, expected_counts[k][1], expected_counts[k][0] - counts[k]
-    end
-    puts
-  rescue RuntimeError => e
-    p e
-    puts
-  end
-end