Class: Categorize::Models::Cluster

Inherits:
AbstractModel show all
Defined in:
lib/categorize/models/cluster.rb

Direct Known Subclasses

HierarchicalCluster

Instance Attribute Summary collapse

Instance Method Summary collapse

Methods inherited from AbstractModel

#build_dataset, #build_vars, #vectorize

Constructor Details

#initialize ⇒ Cluster

Returns a new instance of Cluster.



9
10
11
12
13
 
# File 'lib/categorize/models/cluster.rb', line 9

def initialize
  @num_clusters = 10
  @clusterer = Ai4r::Clusterers::WardLinkage.new
  super
end

Instance Attribute Details

#num_clusters ⇒ Object

Returns the value of attribute num_clusters.



7
8
9
 
# File 'lib/categorize/models/cluster.rb', line 7

def num_clusters
  @num_clusters
end

Instance Method Details

#build_categories(clusters) ⇒ Object 



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
 
# File 'lib/categorize/models/cluster.rb', line 22

def build_categories(clusters)
  clusters_to_records = Hash[clusters.each_with_index.map do |cluster, i|
    [i, cluster.data_items.map { |v| @vectors.index(v) }]
  end]

  @query_terms ||= @query.split.map(&:downcase)

  categories = clusters_to_records.map do |cluster, records|
    term_vectors = records.map { |r| @vectors[r] }.transpose
    tf = term_vectors.map { |f| f.reduce(&:+) }
    get_bigram_max(records, tf)
  end

  records = clusters_to_records.values

  # merge categories with the same label
  categories_records = []
  categories.each_with_index do |category, i|
    j = categories[0...i].index(category)

    if j && categories_records[j]
      categories_records[j].last + records.shift
    else
      categories_records << [category, records.shift]
    end
  end

  categories_records
end

#model(query, records_to_tokens) ⇒ Object 



15
16
17
18
19
20
 
# File 'lib/categorize/models/cluster.rb', line 15

def model(query, records_to_tokens)
  @query = query
  dataset = build_vars(records_to_tokens)
  @clusterer.build(dataset, @num_clusters)
  build_categories(@clusterer.clusters)
end