kmeans-clusterer 0.11.2 → 0.11.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 5074292fdb3b22c58c295514f9f80c255bd9c38b
-  data.tar.gz: a2ed2f4c4d288755d551444ac279e57762e14dd3
+  metadata.gz: 07e426d9d9009e4ea9fc432f08b8390f8bb69bc6
+  data.tar.gz: 8cd3c31df1d33de334c63db296fe2d40f4c56eac
 SHA512:
-  metadata.gz: 896ab8eae01c177a4c70550ee73ce22fd57d044c03e099d1148cf017adf1b9be1d9663571a9dcfc23b744776d0d4689e36e5967a3a4fd794a34d08b8026c5e84
-  data.tar.gz: d19418742a21b6d7302d2be7cca6701553c82f3cd2b5cfc6e4289948b2c4f32cc1f05e7861af88f3014538b0faf1f7430e39482732c23b8a0ff9924144a74e65
+  metadata.gz: ffeb5fd9e2167411314674110cb618d5de43460fd32d01572bd400e7905fd7f94052624fa013c3d1309b08c10d5f1afc92b24e87ff9f4eb302c25639f53f3e32
+  data.tar.gz: a4e10c472ec5fc84b4a1444da6fadf5c246310a2aef0f701d58af7cd62231a9aac53fff0c3200ed48badfab5b35b117a9ac8be23aa986cfaa3636a37686fcf12

data/lib/kmeans-clusterer.rb

@@ -260,10 +260,14 @@ class KMeansClusterer
     point_distances = Distance.euclidean @data, @data
 
     scores = @points.map do |point|
-      sort_index = point.centroid_distances.sort_index
-      c1, c2 = sort_index[0], sort_index[1]
-      a = dissimilarity point.id, c1, point_distances
-      b = dissimilarity point.id, c2, point_distances
+      dissimilarities = @clusters.map do |cluster|  
+        dissimilarity(point.id, cluster.id, point_distances)
+      end
+      a = dissimilarities[point.cluster.id]
+      # set to Infinity so we can pick next closest via min()
+      dissimilarities[point.cluster.id] = Float::INFINITY
+      b = dissimilarities.min
+
       (b - a) / [a,b].max
     end
 
@@ -279,7 +283,7 @@ class KMeansClusterer
     def dissimilarity point_id, cluster_id, point_distances
       cluster_point_ids = @cluster_assigns.eq(cluster_id).where
       cluster_point_distances = point_distances[cluster_point_ids, point_id]
-      cluster_point_distances.sum / cluster_point_distances.length
+      cluster_point_distances.mean
     end
 
     def init_centroids
@@ -301,16 +305,11 @@ class KMeansClusterer
 
       while centroid_ids.length < @k
         centroids = @data[true, centroid_ids]
-
         distances = Distance.euclidean(centroids, @data, @row_norms)
+        
+        # squared distances of each point to the nearest centroid
+        d2 = NArray.ref(distances.min(1).flatten)**2
 
-        d2 = []
-        @points_count.times do |i|
-          min_distance = distances[i, true].min
-          d2 << min_distance**2
-        end
-
-        d2 = NArray.cast(d2, @typecode)
         probs = d2 / d2.sum
         cumprobs = probs.cumsum
         r = rand
@@ -334,16 +333,6 @@ class KMeansClusterer
       @points_count.times.to_a.sample @k
     end
 
-    def get_centroid i
-      NArray.ref(@centroids[true, i].flatten)
-    end
-
-    def get_points_for_cluster i
-      point_ids = @cluster_assigns.eq(i).where
-      points = @data[true, point_ids]
-      points.empty? ? NArray.sfloat(0) : NArray.ref(points)
-    end
-
     def origin
       Array.new(@points[0].dimension, 0) 
     end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kmeans-clusterer
 version: !ruby/object:Gem::Version
-  version: 0.11.2
+  version: 0.11.3
 platform: ruby
 authors:
 - Geoff Buesing
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-04-03 00:00:00.000000000 Z
+date: 2015-04-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: narray