apriori-ruby 0.0.6 → 0.0.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 0c75fe9bfda0a69c987063e1603416a024b8674e
-  data.tar.gz: 6833a168908a6e2bb901cc745bccc49f51a41c97
+  metadata.gz: d476a1f153c942bce7b6ab5bcb6e3482547fa01b
+  data.tar.gz: 0cd9fe4e61335cb42e00ee7934ede7d8cd896658
 SHA512:
-  metadata.gz: eac6f9f9bf0d098ed3cbf26e5248f225feb89eba48f29334d9b901258602ccf4253b0668671980c1e0d35543bcedf2c708a4e7e803f44b897dcc012b97736f5f
-  data.tar.gz: 857a9ee437da98c34ed7aa71eceeba234baa4a0622cdbce4ab3ef96a89136c1ed8b5443551a64d9b4de6d06897c0952cb7b6571efda106d2205533af4a2f716d
+  metadata.gz: 945a4ef7009955375496c5666133a7e56929b33171a7336e0614160442793db8874367c55c5d1a58d7890e150f0fc99adbf680a162886d6c536d72f078b4296f
+  data.tar.gz: 141573132f47069e3504a86be9b68d0cd3765204894b37d069afe1768a5d0de9a41967e32bea4b785790b2c11f7e2458d15bc3519dc09cdb927aa04befd6cddc

data/README.md

@@ -4,7 +4,6 @@ http://en.wikipedia.org/wiki/Apriori_algorithm
 
 Implementation Project for CS 634 - Data Mining
 
-**Project is still in progress**  
 Requirements: Ruby 2.1.X  
 Installation
 ```bash

data/lib/apriori/item_set.rb

@@ -1,6 +1,6 @@
 module Apriori
   class ItemSet
-    attr_reader :data_set, :min_support, :min_confidence, :candidates
+    attr_reader :data_set, :min_support, :min_confidence
 
     def initialize(data_set)
       @data_set = data_set
@@ -19,15 +19,8 @@ module Apriori
     end
 
     def create_association_rules min_confidence
-      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(',')}"
-            association_rules[rule_name] ||= confidence(combo.flatten, (set.flatten - combo.flatten))
-          end
-        end
-      end
-      association_rules.select{|name, confidence| confidence >= min_confidence}
+      make_association_rules unless frequent_item_sets.empty?
+      association_rules.select{ |name, confidence| confidence >= min_confidence }
     end
 
     def support item
@@ -44,25 +37,32 @@ module Apriori
       end.reject {|item| item == false }.size
     end
 
-    def contains_all? set, subset
-      set.to_set.superset? subset.to_set
-    end
-
     private
+    attr_reader :candidates
 
     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
+        list = List.new(reject_candidates)
+        frequent_item_sets << list
         @candidates = list.make_candidates
       end
     end
 
+    def make_association_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(',')}"
+            association_rules[rule_name] ||= confidence(combo.flatten, (set.flatten - combo.flatten))
+          end
+        end
+      end
+    end
+
     def association_rules
       @association_rules ||= {}
+      @association_rules[min_support] ||= {}
     end
 
     def frequent_item_sets
@@ -70,6 +70,10 @@ module Apriori
       @frequent_item_sets[min_support] ||= []
     end
 
+    def contains_all? set, subset
+      set.to_set.superset? subset.to_set
+    end
+
     def reject_candidates
       candidates.reject{|item| support(item) < min_support}
     end

data/lib/apriori/list.rb

@@ -1,21 +1,25 @@
 module Apriori
   class List
-    attr_reader :sets, :set_size
+    attr_reader :sets
 
-    def initialize sets, set_size
+    def self.create_subsets set
+      (1).upto(set.size - 1).flat_map { |n| set.combination(n).to_a }
+    end
+    
+    def initialize sets
       @sets = sets
-      @set_size = set_size
     end
 
-    def self.create_subsets set
-      (1).upto(set.size - 1).flat_map { |n| set.combination(n).to_a }
+    def size
+      return 0 if sets.empty?
+      sets.first.size
     end
 
     def make_candidates
-      if set_size <= 2
-        sets.flatten.combination(set_size).to_a
+      if (size + 1) <= 2
+        sets.flatten.combination(size + 1).to_a
       else
-        self_join prune
+        self_join(prune)
       end
     end
 

data/lib/apriori/version.rb

@@ -1,3 +1,3 @@
 module Apriori
-  VERSION = "0.0.6"
+  VERSION = "0.0.8"
 end

data/spec/lib/apriori/item_set_spec.rb

@@ -31,6 +31,13 @@ describe Apriori::ItemSet do
     end
   end
 
+  context '#create_frequent_item_sets' do
+    it 'creates frequent item sets for a given support' do
+      @set = Apriori::ItemSet.new({:t1 => ['1','2','3'], :t2 => ['1','2','4'], :t3 => ['1','4','5']})
+      expect(@set.create_frequent_item_sets(60).first.sets).to eql([['1'],['2'],['4']])
+    end
+  end
+
   context '#create_association_rules' do
     it 'creates association rules for all combinations' do
       @set = Apriori::ItemSet.new({:t1 => ['1','2','3'], :t2 => ['1','2','4'], :t3 => ['1','4','5']})
@@ -46,17 +53,4 @@ describe Apriori::ItemSet do
     end
   end
 
-  context '#contains_all?' do
-    it 'will return true if one array contains all elements of sub set' do
-      set = [1,2,3]
-      subset = [2,1]
-      expect(@item_set.contains_all?(set, subset)).to be true
-    end
-
-    it 'will return false if one array does not contain all the elements of a sub set' do
-      set = [1,2,3,4]
-      subset = [2,1,5]
-      expect(@item_set.contains_all?(set, subset)).to be false
-    end
-  end
 end

data/spec/lib/apriori/list_spec.rb

@@ -1,20 +1,20 @@
 describe Apriori::List do
   context '#make_candidates' do
     it 'will return all combinations for first two iterations' do
-      array =['Hi1','Hi2','Hi3']
-      @list = Apriori::List.new(array,2)
+      array =[['Hi1'],['Hi2'],['Hi3']]
+      @list = Apriori::List.new(array)
       expect(@list.make_candidates).to eql([['Hi1','Hi2'],['Hi1','Hi3'],['Hi2','Hi3']])
     end
 
     it 'will return the self join after the first two iterations' do
       array = [['Hi1','Hi2'], ['Hi1','Hi3'], ['Blah1', 'Blah2'], ['Blah1', 'Blah3']]
-      @list = Apriori::List.new(array,3)
+      @list = Apriori::List.new(array)
       expect(@list.make_candidates).to eql([['Hi1','Hi2','Hi3'],['Blah1','Blah2','Blah3']])
     end
 
     it 'will prune elements' do
       array = [['Hi1','Hi2'], ['Hi1','Hi3'], ['Blah1', 'Blah2'], ['Blah1', 'Blah3'], ['Blarg1', 'Blarg2']]
-      @list = Apriori::List.new(array,3)
+      @list = Apriori::List.new(array)
       expect(@list.make_candidates).to eql([['Hi1','Hi2','Hi3'],['Blah1','Blah2','Blah3']])
     end
   end
@@ -30,4 +30,18 @@ describe Apriori::List do
       expect(Apriori::List.create_subsets([1,2,3])).to eql([[1],[2],[3],[1,2],[1,3],[2,3]])
     end
   end
+
+  context '#set_size' do
+    it 'will return the size of the sets' do
+      array = [['Hi1','Hi2'], ['Hi1','Hi3'], ['Blah1', 'Blah2'], ['Blah1', 'Blah3'], ['Blarg1', 'Blarg2']]
+      @list = Apriori::List.new(array)
+      expect(@list.size).to eql(2)
+    end
+
+    it 'will return zero if there is no sets' do
+      array = []
+      @list = Apriori::List.new(array)
+      expect(@list.size).to eql(0)
+    end
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: apriori-ruby
 version: !ruby/object:Gem::Version
-  version: 0.0.6
+  version: 0.0.8
 platform: ruby
 authors:
 - Bryan Mulvihill
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-09-28 00:00:00.000000000 Z
+date: 2014-10-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler