data_mining 0.0.5 → 0.0.6

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 8bd3fddec5ee391005c68d0b98688a15ad0049c6
-  data.tar.gz: f0dc879bab93ae64bbf42bb139ca5feb2bab0588
+  metadata.gz: 4803a5f289ea44fdd11459cd36955cbfa0420a16
+  data.tar.gz: 6390a930c494f7eb600db10168fb124f8ce7a339
 SHA512:
-  metadata.gz: d77e59e8ead5dfd7f171ef7a3f47dc9749579f06896200843c7a1d382169910b8fd303f6ae37d07b24fe25515ac2a5fa2b11d734f58bc9216c5f09d7d66983d2
-  data.tar.gz: 27c676e73655384b28d50f4247180dfcfc0c287cf1caf873cfbd014af70e0fcce346e4ea84fc76a48756b7ea96593e14350e8578c388f8e09248b04567e4feee
+  metadata.gz: 8fae4b4508398f56fd95956c3a6bd17bb029c363276bf6873a108ec7d60af62175975f839e5d3038dc32fb31afc04c2e328ceaa24fe6affcb8db044b5f85ee07
+  data.tar.gz: 846a24bc8ab88f0939c12a2545acd0bd4082efc9417b683692e399b7fc535b2ce58728b0a67413d513fa630a24f8f4fc320f53d6e1edc27fbd1edc1824e555bd

data/lib/data_mining/dbscan.rb

@@ -87,17 +87,8 @@ module DataMining
       return false if p1.value.length != p2.value.length
       (p1.value + p2.value).all? { |x| x.is_a? Numeric }
     end
-
-    def euclidean_distance(p1, p2)
-      p1 = p1.value
-      p2 = p2.value
-      Math.sqrt(
-        p1.each_with_index.inject(0) do |sum, (v, i)|
-          sum + ((v - p2[i])**2)
-        end
-      )
-    end
   end
 end
 
 require 'data_mining/point'
+require 'data_mining/euclidean_distance'

data/lib/data_mining/euclidean_distance.rb

@@ -0,0 +1,9 @@
+def euclidean_distance(p1, p2)
+  p1 = p1.value
+  p2 = p2.value
+  Math.sqrt(
+    p1.each_with_index.inject(0) do |sum, (v, i)|
+      sum + ((v - p2[i])**2)
+    end
+  )
+end

data/lib/data_mining/knn.rb

@@ -0,0 +1,35 @@
+module DataMining
+  # k-Nearest Neighbor Classifier
+  class KNearestNeighbor
+    # Find nearest neighbors class
+    #
+    # Arguments:
+    #   training_data: (array of arrays, like [[:c1, value], [:c2, value] .. ])
+    #   k: (integer)
+    def initialize(training_data, k)
+      @data = training_data.map { |i, v| DataMining::Point.new(i, v) }
+      @k    = k
+    end
+
+    def classify(point)
+      count_classes(k_nearest_points(point)).max_by { |_, v| v }.first
+    end
+
+    private
+
+    def count_classes(points)
+      points.each_with_object(Hash.new(0)) do |p, o|
+        o[p.id] += 1
+      end
+    end
+
+    def k_nearest_points(point)
+      @data.sort_by do |item|
+        euclidean_distance(item, DataMining::Point.new(point[0], point[1]))
+      end.take(@k)
+    end
+  end
+end
+
+require 'data_mining/point'
+require 'data_mining/euclidean_distance'

data/lib/data_mining/page_rank.rb

@@ -39,8 +39,9 @@ module DataMining
     end
 
     def next_state
+      current_term = term
       @graph.each_with_object({}) do |(node, _), ranks|
-        ranks[node] = term + @damper * sum_incoming_scores(node)
+        ranks[node] = current_term + @damper * sum_incoming_scores(node)
       end
     end
 
@@ -53,6 +54,7 @@ module DataMining
     end
 
     def pagerank_of_sinknodes
+      return 0 if @sinknodes.empty?
       @ranks.select { |k, _| @sinknodes.include?(k) }.values.inject(:+).to_f
     end
   end

data/lib/data_mining/point.rb

@@ -3,7 +3,7 @@ module DataMining
   class Point
     attr_reader :id, :value
 
-    # Represents a Point for the DBScan Algorithm
+    # Represents a Point
     #
     # Arguments:
     #   id: (symbol)

data/lib/data_mining.rb

@@ -1,6 +1,8 @@
+# A little collection of several Data-Mining-Algorithms
 module DataMining
 end
 
 require 'data_mining/dbscan'
 require 'data_mining/apriori'
 require 'data_mining/page_rank'
+require 'data_mining/knn'

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: data_mining
 version: !ruby/object:Gem::Version
-  version: 0.0.5
+  version: 0.0.6
 platform: ruby
 authors:
 - Manuel Stuefer
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-07-15 00:00:00.000000000 Z
+date: 2015-07-24 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: minitest
@@ -61,6 +61,8 @@ files:
 - lib/data_mining.rb
 - lib/data_mining/apriori.rb
 - lib/data_mining/dbscan.rb
+- lib/data_mining/euclidean_distance.rb
+- lib/data_mining/knn.rb
 - lib/data_mining/page_rank.rb
 - lib/data_mining/point.rb
 homepage: https://github.com/mstuefer/data_mining