Class: Groupie

Inherits:
Object
  • Object
show all
Defined in:
lib/groupie.rb,
lib/groupie/group.rb,
lib/groupie/core_ext/string.rb

Defined Under Namespace

Modules: CoreExt Classes: Group

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initializeGroupie

Returns a new instance of Groupie.



7
8
9
# File 'lib/groupie.rb', line 7

def initialize
  @groups = {}
end

Class Method Details

.versionObject



83
84
85
# File 'lib/groupie.rb', line 83

def self.version
  File.read(File.join(File.dirname(File.expand_path(__FILE__)), "..", "VERSION")).strip
end

Instance Method Details

#[](group) ⇒ Object



11
12
13
# File 'lib/groupie.rb', line 11

def [](group)
  @groups[group] ||= Group.new(group)
end

#classify(entry, strategy = :sum) ⇒ Object



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
57
58
# File 'lib/groupie.rb', line 24

def classify(entry, strategy=:sum)
  results = {}
  total_count = @groups.inject(0) do |sum, name_group|
    group = name_group.last
    count = group.count(entry)
    case strategy
    when :sum
      sum += count
    when :sqrt, :unique
      sum += Math::sqrt(count)
    when :log
      sum += Math::log10(count) if count > 0
    else
      raise "Invalid strategy: #{strategy}"
    end
    next sum
  end
  return results if 0 == total_count

  @groups.each do |name, group|
    count = group.count(entry)
    case strategy
    when :sum
      # keep count
    when :sqrt, :unique
      count = Math::sqrt(count)
    when :log
      count = Math::log10(count) if count > 0
    else
      raise "Invalid strategy: #{strategy}"
    end
    results[name] = count > 0 ? count.to_f / total_count : 0.0
  end
  return results
end

#classify_text(words, strategy = :sum) ⇒ Object

Classify a text by taking the average of all word classifications.



61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
# File 'lib/groupie.rb', line 61

def classify_text(words, strategy=:sum)
  hits = 0
  if strategy==:unique
    words = words & unique_words
  end
  group_score_sums = words.inject({}) do |results, word|
    word_results = classify(word, strategy)
    next results if word_results.empty?
    hits += 1
    results.merge(word_results) do |key, old, new|
      old + new
    end
  end

  averages={}
  group_score_sums.each do |group, sum|
    averages[group] = hits > 0 ? sum / hits : 0
  end

  averages
end

#unique_wordsObject



15
16
17
18
19
20
21
22
# File 'lib/groupie.rb', line 15

def unique_words
  @unique_words ||= (
    total_count = @groups.values.map {|group| group.word_counts}.inject{|total, counts| total.merge(counts){|key,o,n| o+n}}
    median_index = [total_count.values.size * 3 / 4 - 1, 1].max
    median_frequency = total_count.values.sort[median_index]
    total_count.select{|word, count| count <= median_frequency}.map(&:first)
  )
end