kmeans-clusterer 0.5.0 → 0.5.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 83e84b661282a532269410c4bf91d98f7831cc45
-  data.tar.gz: 2ec3bf6b111d67e4f5beced25b088b59592e3cdf
+  metadata.gz: e1eef92b1d0554e9bd082619863534ce7f74aa64
+  data.tar.gz: 123a573338accb5c7c68e406e4e691bd5235300c
 SHA512:
-  metadata.gz: 7735bf8e71c4fa4300793e40a299b7a4af267821860092b2dd21db16cfc407bc0b2e6a9b99c37846ffbd02b76976242d9719bfa92e244455be14208b1934d21f
-  data.tar.gz: 54f7bf8afff09caf185dbf95c8baeb4ea471e4496b29967ad88b3fc377e0063803c7810295825fe581fd744048964b592197bcfea6d2aa1178133d39e2937b0b
+  metadata.gz: 51dcecd3cf4a8157250d605dcffdc50a0e2d4621436232f438c62aa743d5a01cf06f5c12d01ff214808daf8c62cedb2501b33179c44060cfc0354b2c31d2f592
+  data.tar.gz: 31ad149b1ef15aa0aa64280f4989b7f6476c4529dbf9e15bb3c04497b5756e06438ede5c784a9ef0e179cd96f310414c5f3d80881ec048b046b7e88eef30eb50

data/lib/kmeans-clusterer.rb

@@ -87,19 +87,25 @@ class KMeansClusterer
     opts[:points_matrix] = NMatrix.cast(data, NArray::DFLOAT)
     opts[:row_norms] = opts[:points_matrix].map {|v| v**2}.sum(0)
 
-    runs = opts[:runs].times.map do |i|
+    bestrun = nil
+
+    opts[:runs].times do |i|
       km = new(opts).run
+
       if opts[:log]
         puts "[#{i + 1}] #{km.iterations} iter\t#{km.runtime.round(2)}s\t#{km.error.round(2)} err"
       end
-      km
+      
+      if bestrun.nil? || (km.error < bestrun.error)
+        bestrun = km
+      end
     end
 
-    runs.sort_by {|run| run.error }.first.finish
+    bestrun.finish
   end
 
 
-  attr_reader :k, :points, :clusters, :error, :iterations, :runtime
+  attr_reader :k, :points, :clusters, :error, :mean, :std, :iterations, :runtime
 
 
   def initialize opts = {}
@@ -192,7 +198,7 @@ class KMeansClusterer
     wrap_point Array.new(@points[0].dimension, 0) 
   end
 
-  def silhouette_score
+  def silhouette
     return 1.0 if @k < 2
 
     distances = distance(@centroids, @points_matrix)
@@ -212,6 +218,8 @@ class KMeansClusterer
     scores.reduce(:+) / scores.length # mean score for all points
   end
 
+  alias_method :silhouette_score, :silhouette
+
   private
     def wrap_point point
       return point if point.is_a?(Point)
@@ -323,7 +331,8 @@ class KMeansClusterer
 
     def get_points_for_centroid i
       point_ids = @cluster_point_ids[i]
-      NArray.cast @points_matrix[true, point_ids]
+      points = @points_matrix[true, point_ids]
+      points.empty? ? NArray.dfloat(0) : NArray.cast(points)
     end
 
     def distance x, y, yy = @row_norms

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kmeans-clusterer
 version: !ruby/object:Gem::Version
-  version: 0.5.0
+  version: 0.5.2
 platform: ruby
 authors:
 - Geoff Buesing
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-02-05 00:00:00.000000000 Z
+date: 2015-02-07 00:00:00.000000000 Z
 dependencies: []
 description: k-means clustering. Uses NArray for fast calculations.
 email: gbuesing@gmail.com