weighted_random 0.0.3 → 0.0.4

data/.gitignore

@@ -2,3 +2,4 @@
 .bundle
 Gemfile.lock
 pkg/*
+*.swp

data/README.rdoc

@@ -1,7 +1,8 @@
 = Weighted Random
+
 Weighted randomization extension for ActiveRecord.
 
-Loads records with weight value for randomization and gives ability of weighted randomization of them.
+Gives ability to get weighted random record from particular database relation.
 
 == Installation
 
@@ -30,29 +31,23 @@ Load weighted randomization stuff into desired model and set accessibillity of p
 
 == Usage
 
-=== Load data
+=== Importing data
 
-Load records with <code>weight</code> attribute for randomization, here is an example of loading last names and their frequencies from CSV file:
+To import data for randomization you have to provide an array of hashes with <code>weight</code> attributes to
+<code>create_with_cumulative_weight</code> method, for example:
 
-last_names.csv:
-  Smith,10
-  Johnson,8
-  Williams,7
-  Jones,6
-  Brown,6
-  Davis,5
-  Miller,4
-  Wilson,3
+  FirstName.create_with_cumulative_weight([
+    {:name => "Smith",    :weight => 10},
+    {:name => "Johnson",  :weight => 8},
+    {:name => "Williams", :weight => 7},
+    {:name => "Jones",    :weight => 6},
+    {:name => "Brown",    :weight => 6},
+    {:name => "Davis",    :weight => 5},
+    {:name => "Miller",   :weight => 4},
+    {:name => "Wilson",   :weight => 3}
+  ])
 
-seeds.rb:
-  csv = CSV.open(File.expand_path("seeds/first_names.csv", File.dirname(__FILE__)))
-  names = csv.collect { |name, weight| {:name => name, :weight => to_i} }
-  csv.close
-  LastName.create_with_cumulative_weight(names)
-
-Simply you have to supply collection of hashes, each with <code>weight</code> integer property, to
-<code>create_with_cumulative_weight</code> method of your desired model class, so it will load them all
-into database with additional <code>cumulative_weight</code> attribute, which is reqired for faster weighted randomization.
+This method will save all this hashes in database with additional <code>cumulative_weight</code> attribute.
 
 === Weighted randomization
 
@@ -60,7 +55,7 @@ To get weighed random record simply run:
 
   LastName.weighted_rand
 
-=== Demonstration
+Demonstration:
 
   10.times { puts LastName.weighted_rand.name }
   Johnson
@@ -74,6 +69,26 @@ To get weighed random record simply run:
   Miller
   Jones
 
+== Import data from CSV
+
+Here is an example of importing last names and their frequencies from CSV file:
+
+db/seeds/last_names.csv:
+  name,weight
+  Smith,10
+  Johnson,8
+  Williams,7
+  Jones,6
+  Brown,6
+  Davis,5
+  Miller,4
+  Wilson,3
+
+db/seeds.rb:
+  LastName.create_with_cumulative_weight(
+    CSV.table(File.expand_path("seeds/last_names.csv", File.dirname(__FILE__))).collect(&:to_hash)
+  )
+
 == Author
 
 Szymon Przybył (http://github.com/apocalyptiq)

data/lib/weighted_random/railtie.rb

@@ -4,9 +4,15 @@ module WeightedRandom
     class Railtie < Rails::Railtie
       initializer 'weighted_random.insert_into_active_record' do
         ActiveSupport.on_load :active_record do
-          ActiveRecord::Base.send(:extend, WeightedRandom::Loader)
+            WeightedRandom::Railtie.insert
         end
       end
     end
   end
+
+  class Railtie
+    def self.insert
+      ActiveRecord::Base.send(:extend, WeightedRandom::Loader)
+    end
+  end
 end

data/lib/weighted_random/version.rb

@@ -1,3 +1,3 @@
 module WeightedRandom
-  VERSION = "0.0.3"
+  VERSION = "0.0.4"
 end

data/lib/weighted_random.rb

@@ -10,14 +10,16 @@ module WeightedRandom
 
   module ClassMethods
     def weighted_rand
-      self.where("cumulative_weight >= #{rand(self.maximum('cumulative_weight'))}").order('cumulative_weight').first
+      self.where("cumulative_weight > #{Kernel.rand(self.maximum('cumulative_weight'))}").order('cumulative_weight').limit(1).first
     end
 
     def create_with_cumulative_weight(collection)
-      self.create! self.compute_and_insert_cumulative_weight(collection)
+      self.create WeightedRandom.set_cumulative_weight(collection)
     end
+  end
 
-    def compute_and_insert_cumulative_weight(collection)
+  class << self
+    def set_cumulative_weight(collection)
       weight_sum = 0
       collection.collect do |item|
         weight_sum += item[:weight]

data/spec/auto_management_of_cumulative_weight_spec.rb

@@ -0,0 +1,30 @@
+require 'spec_helper'
+
+describe WeightedRandom, "auto management of cumulative weight" do
+
+  def cumulative_weight_of(name)
+    TestModel.find_by_name(name).cumulative_weight
+  end
+
+  describe "#create_with_cumulative_weight sets cumulative_weight to" do
+    before(:all) do
+      TestModel.create_with_cumulative_weight [
+        {:name => 'first-50',  :weight => 50},
+        {:name => 'second-1', :weight => 1},
+        {:name => 'last-10',  :weight => 10}
+      ]
+    end
+
+    specify "50 for first record with weight 50" do
+      cumulative_weight_of('first-50').should be(50)
+    end
+
+    specify "51 for second record with weight 1" do
+      cumulative_weight_of('second-1').should be(51)
+    end
+
+    specify "61 for last record with weight 10" do
+      cumulative_weight_of('last-10').should be(61)
+    end
+  end
+end

data/spec/config/database.yml

@@ -0,0 +1,3 @@
+test:
+  adapter: sqlite3
+  database: ":memory:"

data/spec/spec_helper.rb

@@ -0,0 +1,29 @@
+require 'rspec'
+
+require 'active_record'
+require 'active_support'
+
+DIR = File.dirname(__FILE__)
+ROOT = File.join(DIR, '..')
+
+# Load WeightedRandom module!
+require File.join(ROOT, 'lib/weighted_random')
+
+# Load extension inserter
+WeightedRandom::Railtie.insert
+
+# Database configuration and connection
+config = YAML::load(IO.read(DIR + '/config/database.yml'))
+ActiveRecord::Base.establish_connection(config['test'])
+
+# Test model
+class TestModel < ActiveRecord::Base
+  weighted_randomizable
+  attr_accessible :name, :weight, :cumulative_weight
+
+  connection.create_table self.table_name, :force => true do |t|
+    t.string :name
+    t.integer :weight
+    t.integer :cumulative_weight
+  end
+end

data/spec/weighted_randomization_spec.rb

@@ -0,0 +1,88 @@
+require 'spec_helper'
+
+describe WeightedRandom, "weighted randomization" do
+
+  def times_drawed(name)
+    counter = 0
+    rand_times.times { counter += 1 if TestModel.weighted_rand.name == name }
+    counter
+  end
+
+  def reload_with_data(data)
+    TestModel.destroy_all
+    TestModel.create_with_cumulative_weight(data)
+  end
+
+  context "in 1000 rands" do
+    let(:rand_times) { 1000 }
+
+    before(:each) do
+      # Set random number generator seed, so every test will receive the same sequence of numbers
+      Kernel.srand(77)
+    end
+
+    describe "deviation tolerance" do
+      before(:all) do
+        reload_with_data [
+          {:name => 'first-50',  :weight => 50},
+          {:name => 'second-25', :weight => 25},
+          {:name => 'third-10',  :weight => 10},
+          {:name => 'fourth-5',  :weight => 5},
+          {:name => 'fifth-3',   :weight => 3},
+          {:name => 'sixth-2',   :weight => 2},
+          {:name => 'seventh-1', :weight => 1},
+          {:name => 'last-4',    :weight => 4}
+        ]
+      end
+
+      it "draws record with weight 50% of overall within 50 of 500 times (5%   of 50%)" do
+        times_drawed('first-50').should be_within(50).of(500)
+      end
+
+      it "draws record with weight 25% of overall within 40 of 250 times (4%   of 25%)" do
+        times_drawed('second-25').should be_within(40).of(250)
+      end
+
+      it "draws record with weight 10% of overall within 25 of 100 times (2.5% of 10%)" do
+        times_drawed('third-10').should be_within(25).of(100)
+      end
+
+      it "draws record with weight 5%  of overall within 20 of 50  times (2%   of 5%)" do
+        times_drawed('fourth-5').should be_within(20).of(50)
+      end
+
+      it "draws record with weight 3%  of overall within 15 of 30  times (1.5% of 3%)" do
+        times_drawed('fifth-3').should be_within(15).of(30)
+      end
+
+      it "draws record with weight 2%  of overall within 10 of 20  times (1%   of 2%)" do
+        times_drawed('sixth-2').should be_within(10).of(20)
+      end
+
+      it "draws record with weight 1%  of overall within 7  of 10  times (0.7% of 1%)" do
+        times_drawed('seventh-1').should be_within(7).of(10)
+      end
+    end
+
+    describe "boundary records with weight 1 (10% of overall)" do
+      before(:all) do
+        reload_with_data [
+          {:name => 'first-1',  :weight => 1},
+          {:name => 'second-8', :weight => 8},
+          {:name => 'last-1',   :weight => 1}
+        ]
+      end
+
+      context "draws within 25 of 100 times (2.5% of 10%)" do
+        it "first" do
+          times_drawed('first-1').should be_within(25).of(100)
+        end
+
+        it "last" do
+          times_drawed('last-1').should be_within(25).of(100)
+        end
+      end
+    end
+  end
+
+end

data/weighted_random.gemspec

@@ -14,6 +14,8 @@ Gem::Specification.new do |s|
 
   s.rubyforge_project = "weighted_random"
 
+  s.add_development_dependency('rspec')
+
   s.files         = `git ls-files`.split("\n")
   s.test_files    = `git ls-files -- {test,spec,features}/*`.split("\n")
   s.executables   = `git ls-files -- bin/*`.split("\n").map{ |f| File.basename(f) }

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: weighted_random
 version: !ruby/object:Gem::Version
-  version: 0.0.3
+  version: 0.0.4
   prerelease: 
 platform: ruby
 authors:
@@ -9,9 +9,20 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2011-04-12 00:00:00.000000000 +02:00
+date: 2011-04-13 00:00:00.000000000 +02:00
 default_executable: 
-dependencies: []
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: &72406200 !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: *72406200
 description: ActiveRecord extension for weighted randomization which supplies loading
   records with weight for randomize into database and weighted randomization of them
 email:
@@ -27,6 +38,10 @@ files:
 - lib/weighted_random.rb
 - lib/weighted_random/railtie.rb
 - lib/weighted_random/version.rb
+- spec/auto_management_of_cumulative_weight_spec.rb
+- spec/config/database.yml
+- spec/spec_helper.rb
+- spec/weighted_randomization_spec.rb
 - weighted_random.gemspec
 has_rdoc: true
 homepage: https://github.com/apocalyptiq/weighted_random