apriori-ruby 0.0.4 → 0.0.6

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 34233d76397403835023b9acc4fbc879105ecefd
-  data.tar.gz: 8f7c58b988c48461be7021268bb5957c66dd2488
+  metadata.gz: 0c75fe9bfda0a69c987063e1603416a024b8674e
+  data.tar.gz: 6833a168908a6e2bb901cc745bccc49f51a41c97
 SHA512:
-  metadata.gz: 4713c04c787db63bb82c4a61faca537b67c6f77bc2bcc6f482da3946380c70721e46f2dcbb4ead9d9cbb24e2b33abdd06262d5bd1445b90dce39d3715da2a6ff
-  data.tar.gz: 87d53ba9aef49e50e3761e28108413827442c32280a0fa03b5bf342e52113c15b437542416121488ab5066b1234a68a53f78efb3d7c50a8060bd671e319c6113
+  metadata.gz: eac6f9f9bf0d098ed3cbf26e5248f225feb89eba48f29334d9b901258602ccf4253b0668671980c1e0d35543bcedf2c708a4e7e803f44b897dcc012b97736f5f
+  data.tar.gz: 857a9ee437da98c34ed7aa71eceeba234baa4a0622cdbce4ab3ef96a89136c1ed8b5443551a64d9b4de6d06897c0952cb7b6571efda106d2205533af4a2f716d

data/README.md

@@ -14,3 +14,13 @@ Or add to Gemfile
 ```ruby
 gem 'apriori-ruby'
 ```
+
+Sample Usage
+```ruby
+  test_data = {t1: [1,2,3,4], t2:[1,2,4,5], t3: [2,3,4,5]}
+  item_set = Apriori::ItemSet.new(test_data)
+  support = 50
+  confidence = 60
+  item_set.mine(support, confidence)
+  => {"1=>2"=>100.0, "2=>1"=>66.66666666666666, "1=>4"=>100.0, "4=>1"=>66.66666666666666, "2=>3"=>66.66666666666666, "3=>2"=>100.0, "2=>4"=>100.0, "4=>2"=>100.0, "2=>5"=>66.66666666666666, "5=>2"=>100.0, "3=>4"=>100.0, "4=>3"=>66.66666666666666, "4=>5"=>66.66666666666666, "5=>4"=>100.0, "1=>2,4"=>100.0, "2=>1,4"=>66.66666666666666, "4=>1,2"=>66.66666666666666, "1,2=>4"=>100.0, "1,4=>2"=>100.0, "2,4=>1"=>66.66666666666666}
+```

data/lib/apriori.rb

@@ -1,4 +1,3 @@
-require 'apriori/algorithm.rb'
 require 'apriori/item_set.rb'
 require 'apriori/list.rb'
 require 'set'

data/lib/apriori/item_set.rb

@@ -1,38 +1,33 @@
 module Apriori
   class ItemSet
-    attr_reader :data_set, :item_set, :candidates, :iteration, :min_support, :min_confidence
+    attr_reader :data_set, :min_support, :min_confidence, :candidates
 
     def initialize(data_set)
       @data_set = data_set
     end
 
-    def frequent_item_sets
-      @frequent_item_sets ||= []
+    def mine(min_support=0, min_confidence=0)
+      @min_support, @min_confidence = min_support, min_confidence
+      create_frequent_item_sets(min_support)
+      create_association_rules(min_confidence)
     end
 
     def create_frequent_item_sets min_support
       @min_support = min_support
-      @iteration = 0
-      @candidates = convert_initial_data_set
-      while candidates.any?
-        @iteration += 1
-        @candidates = list.make_candidates
-        frequent_item_sets << list unless iteration == 1
-      end
+      make_item_sets unless frequent_item_sets.any?
       frequent_item_sets
     end
 
     def create_association_rules min_confidence
-      rules ={}
       frequent_item_sets.each do |freq_lists|
         freq_lists.sets.each do |set|
           List.create_subsets(set).each do |combo|
             rule_name = "#{combo.join(',')}=>#{(set.flatten - combo.flatten).join(',')}"
-            rules[rule_name] = confidence(combo.flatten, (set.flatten - combo.flatten))
+            association_rules[rule_name] ||= confidence(combo.flatten, (set.flatten - combo.flatten))
           end
         end
       end
-      rules.select{|k,v| v >= min_confidence}
+      association_rules.select{|name, confidence| confidence >= min_confidence}
     end
 
     def support item
@@ -55,17 +50,32 @@ module Apriori
 
     private
 
-    def list
-      @list ||= {}
-      @list[iteration] ||= List.new(reject_candidates, iteration)
+    def make_item_sets
+      @candidates = initial_data_set
+      iteration = 0
+      while candidates.any?
+        iteration += 1
+        list = List.new(reject_candidates, iteration)
+        frequent_item_sets << list unless iteration == 1
+        @candidates = list.make_candidates
+      end
+    end
+
+    def association_rules
+      @association_rules ||= {}
+    end
+
+    def frequent_item_sets
+      @frequent_item_sets ||= {}
+      @frequent_item_sets[min_support] ||= []
     end
 
     def reject_candidates
       candidates.reject{|item| support(item) < min_support}
     end
 
-    def convert_initial_data_set
-      @data_set.values.flatten.uniq.map{|item| [item]}
+    def initial_data_set
+      @initial_data_set ||= @data_set.values.flatten.uniq.map{|item| [item]}
     end
   end
 end

data/lib/apriori/list.rb

@@ -1,19 +1,19 @@
 module Apriori
   class List
-    attr_reader :sets, :iteration
+    attr_reader :sets, :set_size
 
-    def initialize sets, iteration
+    def initialize sets, set_size
       @sets = sets
-      @iteration = iteration
+      @set_size = set_size
     end
-    
+
     def self.create_subsets set
       (1).upto(set.size - 1).flat_map { |n| set.combination(n).to_a }
     end
 
     def make_candidates
-      if iteration <= 2
-        sets.flatten.combination(iteration).to_a
+      if set_size <= 2
+        sets.flatten.combination(set_size).to_a
       else
         self_join prune
       end

data/lib/apriori/version.rb

@@ -1,3 +1,3 @@
 module Apriori
-  VERSION = "0.0.4"
+  VERSION = "0.0.6"
 end

data/spec/factories/apriori/apriori.rb

@@ -1,8 +1,4 @@
 FactoryGirl.define do
-  factory :algorithm, class: Apriori::Algorithm do
-    initialize_with {new(FactoryGirl.build(:item_set))}
-  end
-
   factory :item_set, class: Apriori::ItemSet do
     initialize_with {new(FactoryGirl.build(:sample_data))}
   end

data/spec/lib/apriori/item_set_spec.rb

@@ -4,6 +4,18 @@ describe Apriori::ItemSet do
     @item_set = Apriori::ItemSet.new(data)
   end
 
+  context '#mine' do
+    it 'returns all association rules meeting the minimum support and confidence' do
+      expect(@item_set.mine(50,90)).to eql({"Mango=>Keychain"=>100.0, "Onion=>Keychain"=>100.0, "Onion=>Eggs"=>100.0, "Eggs=>Keychain"=>100.0, "Yoyo=>Keychain"=>100.0, "Onion=>Keychain,Eggs"=>100.0, "Onion,Keychain=>Eggs"=>100.0, "Onion,Eggs=>Keychain"=>100.0})
+    end
+
+    it 'will allow you to mine the same set multiple times' do
+      @item_set.mine(100,100)
+      expect(@item_set.mine(50,90)).to eql({"Mango=>Keychain"=>100.0, "Onion=>Keychain"=>100.0, "Onion=>Eggs"=>100.0, "Eggs=>Keychain"=>100.0, "Yoyo=>Keychain"=>100.0, "Onion=>Keychain,Eggs"=>100.0, "Onion,Keychain=>Eggs"=>100.0, "Onion,Eggs=>Keychain"=>100.0})
+    end
+
+  end
+
   context '#confidence' do
     it 'will return a rule with support and confidence' do
       set1 = ['Eggs']

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: apriori-ruby
 version: !ruby/object:Gem::Version
-  version: 0.0.4
+  version: 0.0.6
 platform: ruby
 authors:
 - Bryan Mulvihill
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-09-21 00:00:00.000000000 Z
+date: 2014-09-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -69,12 +69,10 @@ files:
 - Rakefile
 - apriori.gemspec
 - lib/apriori.rb
-- lib/apriori/algorithm.rb
 - lib/apriori/item_set.rb
 - lib/apriori/list.rb
 - lib/apriori/version.rb
 - spec/factories/apriori/apriori.rb
-- spec/lib/apriori/algorithm_spec.rb
 - spec/lib/apriori/item_set_spec.rb
 - spec/lib/apriori/list_spec.rb
 - spec/spec_helper.rb
@@ -104,7 +102,6 @@ specification_version: 4
 summary: Ruby implementation of Apriori Algorithm
 test_files:
 - spec/factories/apriori/apriori.rb
-- spec/lib/apriori/algorithm_spec.rb
 - spec/lib/apriori/item_set_spec.rb
 - spec/lib/apriori/list_spec.rb
 - spec/spec_helper.rb

data/lib/apriori/algorithm.rb

@@ -1,16 +0,0 @@
-module Apriori
-  class Algorithm
-    attr_accessor :min_support, :min_confidence, :item_set
-
-    def initialize(item_set)
-      @item_set = item_set
-    end
-
-    def mine(min_support=0, min_confidence=0)
-      @min_support, @min_confidence = min_support, min_confidence
-      item_set.create_frequent_item_sets(min_support)
-      item_set.create_association_rules(min_confidence)
-    end
-
-  end
-end

data/spec/lib/apriori/algorithm_spec.rb

@@ -1,12 +0,0 @@
-describe Apriori::Algorithm do
-  before do
-    @apriori = FactoryGirl.build(:algorithm)
-  end
-
-  context '#mine' do
-    it 'returns all association rules meeting the minimum support and confidence' do
-      @apriori.mine(50)
-    end
-  end
-
-end