Class: DbClustering::Algorithms::Dbscan
- Inherits:
-
Object
- Object
- DbClustering::Algorithms::Dbscan
- Defined in:
- lib/algorithms/density_based/dbscan.rb
Instance Attribute Summary collapse
-
#clusters ⇒ Object
Returns the value of attribute clusters.
-
#datasource ⇒ Object
Returns the value of attribute datasource.
Instance Method Summary collapse
- #cluster(max_distance:, min_neighbors:) ⇒ Object
- #expand_cluster(point, neighbors, max_distance) ⇒ Object
-
#initialize(datasource:, distance_metric:, debug: false) ⇒ Dbscan
constructor
A new instance of Dbscan.
Constructor Details
#initialize(datasource:, distance_metric:, debug: false) ⇒ Dbscan
Returns a new instance of Dbscan.
9 10 11 12 13 14 |
# File 'lib/algorithms/density_based/dbscan.rb', line 9 def initialize(datasource:, distance_metric:, debug: false) @datasource = datasource @distance_metric = distance_metric @clusters = [] @debug = debug end |
Instance Attribute Details
#clusters ⇒ Object
Returns the value of attribute clusters.
7 8 9 |
# File 'lib/algorithms/density_based/dbscan.rb', line 7 def clusters @clusters end |
#datasource ⇒ Object
Returns the value of attribute datasource.
7 8 9 |
# File 'lib/algorithms/density_based/dbscan.rb', line 7 def datasource @datasource end |
Instance Method Details
#cluster(max_distance:, min_neighbors:) ⇒ Object
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 |
# File 'lib/algorithms/density_based/dbscan.rb', line 16 def cluster(max_distance:, min_neighbors:) @clusters = [] cluster = nil if @debug last_printed_progress = 0.0 end @datasource.iterate_all_points do |point, current_index, points_count| neighbors = @datasource.neighbors(point: point, distance_metric: @distance_metric, max_distance: max_distance) if neighbors.count < min_neighbors point.is_noise = true elsif point.cluster.nil? cluster = DbClustering::Models::Cluster.new @clusters << cluster cluster.add(point) (point, neighbors, max_distance) end yield(point, current_index, points_count) if block_given? if @debug point_type_string = point.is_edge_point? ? 'E' : point.is_core_point? ? 'C' : 'N' print point_type_string progress = (current_index + 1) * 100 / points_count.to_f if progress > last_printed_progress + 1 print "[#{progress.to_i}%]" last_printed_progress = progress end end end if @debug print "\n" puts "#{clusters.count} clusters found" end end |
#expand_cluster(point, neighbors, max_distance) ⇒ Object
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 |
# File 'lib/algorithms/density_based/dbscan.rb', line 58 def (point, neighbors, max_distance) neighbors.each do |neighbor| if neighbor.cluster.nil? point.cluster.add(neighbor) if @debug print "+" end neighbors_of_neighbor = @datasource.neighbors(point: neighbor, distance_metric: @distance_metric, max_distance: max_distance) neighbors_of_neighbor.each do |neighbor_of_neighbor| neighbors << neighbor_of_neighbor unless neighbors.include?(neighbor_of_neighbor) end end end end |