kmeans-clusterer 0.7.0 → 0.8.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 280cf2639965cde59fcff41909375d284637825d
-  data.tar.gz: e9d9fc0db3828b9bfc37def487200cad77a4d5a7
+  metadata.gz: 38086107da4ad5ec21f6daa33166b4d8542f3a59
+  data.tar.gz: f7fed19b6df6b6c9dbfb21548bc07bd1d997b56e
 SHA512:
-  metadata.gz: 7cd63b7d8f844ea17972ce535d4f0b1ea29d61524351431ada328fdbf30736006739ced8f9c1b344f834c0c993f6491ef203f22761f30ef224795a6d38ee4996
-  data.tar.gz: 63ed5fa36fe27785c124877ac9a56d30190f55aec1a08c6ab161f79a24980f91f553c552f7956fed7d3f9bc25b4d1cbcb98df97d06e51a56c17eb5d290381019
+  metadata.gz: 2e4603c448b30875ba49eb694974ca5e43c0f4270d9d7005608f9bad086eff1bab769f23bb92a8756a024f037a1cc03dd568cb02122d7841c3a28b2556571893
+  data.tar.gz: 40ed7fcf8c7db56d7abf50c072cb7b928c501951e3dbf8a3d2276508c4a5ce2bae2b2bb12159c54ce999cd322b721f12f3b1cf3b46f7903aabc79dbf03f8ada4

data/lib/kmeans-clusterer.rb

@@ -23,6 +23,22 @@ class KMeansClusterer
     end
   end
 
+  module Distance
+    def self.euclidean x, y, yy = nil
+      if x.is_a?(NMatrix) && y.is_a?(NMatrix)
+        xx = x.map {|v| v**2}.sum(0)
+        yy ||= y.map {|v| v**2}.sum(0)
+        xy = x * y.transpose
+        distance = xy * -2
+        distance += xx
+        distance += yy.transpose
+        NMath.sqrt distance
+      else
+        NMath.sqrt ((x - y)**2).sum(0)
+      end
+    end
+  end
+
 
   class Point
     attr_reader :id, :data
@@ -66,6 +82,14 @@ class KMeansClusterer
       point.cluster = self
       @points << point
     end
+
+    def sorted_points point = @centroid
+      point = point.data if point.is_a?(Point)
+      point = NArray.cast(point, @centroid.typecode) unless point.is_a?(NArray)
+      points_data = NArray.cast(@points.map(&:data))
+      distances = Distance.euclidean(points_data, point)
+      @points.sort_by.with_index {|p, i| distances[i] }
+    end
   end
 
 
@@ -108,7 +132,7 @@ class KMeansClusterer
   end
 
 
-  attr_reader :k, :points, :clusters, :error, :mean, :std, :iterations, :runtime
+  attr_reader :k, :points, :clusters, :centroids, :error, :mean, :std, :iterations, :runtime
 
 
   def initialize opts = {}
@@ -136,7 +160,7 @@ class KMeansClusterer
     loop do
       @iterations +=1
 
-      distances = distance(@centroids, @points_matrix)
+      distances = Distance.euclidean(@centroids, @points_matrix, @row_norms)
 
       # assign point ids to @cluster_point_ids
       @points_count.times do |i|
@@ -157,7 +181,7 @@ class KMeansClusterer
         else
           points = @points_matrix[true, point_ids]
           newcenter = points.mean(1)
-          moves << distance(centroid, newcenter)
+          moves << Distance.euclidean(centroid, newcenter)
         end
 
         updated_centroids << newcenter
@@ -185,27 +209,23 @@ class KMeansClusterer
   def predict data
     data = NMatrix.cast(data, @typecode)
     data, _m, _s = Scaler.scale(data, @mean, @std, @typecode) if @scale_data
-    distances = distance(@centroids, data, nil)
+    distances = Distance.euclidean(@centroids, data)
     data.shape[1].times.map do |i|
       distances[i, true].sort_index[0] # index of closest cluster
     end
   end
 
   def sorted_clusters point = origin
-    point = wrap_point point
-    centroids = get_cluster_centroids
-    distances = distance(centroids, point.data)
+    point = point.data if point.is_a?(Point)
+    point = NArray.cast(point, @typecode) unless point.is_a?(NArray)
+    distances = Distance.euclidean(NArray.ref(@centroids), point)
     @clusters.sort_by.with_index {|c, i| distances[i] }
   end
 
-  def origin
-    wrap_point Array.new(@points[0].dimension, 0) 
-  end
-
   def silhouette
     return 1.0 if @k < 2
 
-    distances = distance(@centroids, @points_matrix)
+    distances = Distance.euclidean(@centroids, @points_matrix, @row_norms)
 
     scores = @points_count.times.map do |i|
       point = get_point i
@@ -222,20 +242,14 @@ class KMeansClusterer
     scores.reduce(:+) / scores.length # mean score for all points
   end
 
-  alias_method :silhouette_score, :silhouette
-
   def inspect
     %{#<#{self.class.name} k:#{@k} iterations:#{@iterations} error:#{@error} runtime:#{@runtime}>}
   end
 
   private
-    def wrap_point point
-      return point if point.is_a?(Point)
-      Point.new(0, NArray.cast(point, @typecode))
-    end
 
     def dissimilarity points, point
-      distances = distance points, point
+      distances = Distance.euclidean points, point
       distances.sum / distances.length.to_f
     end
 
@@ -259,7 +273,7 @@ class KMeansClusterer
       while centroid_ids.length < @k
         centroids = @points_matrix[true, centroid_ids]
 
-        distances = distance(centroids, @points_matrix)
+        distances = Distance.euclidean(centroids, @points_matrix, @row_norms)
 
         d2 = []
         @points_count.times do |i|
@@ -288,11 +302,7 @@ class KMeansClusterer
     end
 
     def pick_k_random_indexes
-      @points_count.times.to_a.shuffle.slice(0, @k)
-    end
-
-    def get_cluster_centroids
-      NArray.to_na @clusters.map {|c| c.centroid.data }
+      @points_count.times.to_a.sample @k
     end
 
     def set_points
@@ -321,7 +331,7 @@ class KMeansClusterer
         if points.empty?
           0
         else
-          distances = distance points, centroid
+          distances = Distance.euclidean points, centroid
           (distances**2).sum
         end
       end
@@ -343,17 +353,7 @@ class KMeansClusterer
       points.empty? ? NArray.sfloat(0) : NArray.ref(points)
     end
 
-    def distance x, y, yy = @row_norms
-      if x.is_a?(NMatrix) && y.is_a?(NMatrix)
-        xx = x.map {|v| v**2}.sum(0)
-        yy ||= y.map {|v| v**2}.sum(0)
-        xy = x * y.transpose
-        distance = xy * -2
-        distance += xx
-        distance += yy.transpose
-        NMath.sqrt distance
-      else
-        NMath.sqrt ((x - y)**2).sum(0)
-      end
+    def origin
+      Array.new(@points[0].dimension, 0) 
     end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kmeans-clusterer
 version: !ruby/object:Gem::Version
-  version: 0.7.0
+  version: 0.8.0
 platform: ruby
 authors:
 - Geoff Buesing
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-02-13 00:00:00.000000000 Z
+date: 2015-02-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: narray