Merge pull request #196 from aparibocci/max_k_most_frequent_words

Solving the `Top k most frequent words` problem using a max-heap

Author: StepfenShawn

strings/max_k_most_frequent_words.rb

@@ -0,0 +1,36 @@
+require_relative '../data_structures/heaps/max_heap'
+
+##
+# This class represents a word count information
+# (i.e. how many occurrences for a word).
+
+class WordCount
+  include Comparable
+
+  attr_reader :word
+  attr_reader :occurrences
+
+  def <=>(other)
+    occurrences <=> other.occurrences
+  end
+
+  def initialize(word, occurrences)
+    @word = word
+    @occurrences = occurrences
+  end
+end
+
+##
+# Returns the `k` most frequently occurring words, in non-increasing order of occurrence.
+# In this context, a word is defined as an element in the provided list.
+#
+# In case `k` is greater than the number of distinct words, a value of `k` equal
+# to the number of distinct words will be considered, instead.
+
+def max_k_most_frequent_words(words, k)
+  count_by_word = words.tally
+  heap = MaxHeap.new(count_by_word.map { |w, c| WordCount.new(w, c) })
+  most_frequent_words = []
+  [k, count_by_word.size].min.times { most_frequent_words.append(heap.extract_max.word) }
+  most_frequent_words
+end

strings/max_k_most_frequent_words_test.rb

@@ -0,0 +1,28 @@
+require 'minitest/autorun'
+require_relative 'max_k_most_frequent_words'
+
+class TestMaxKMostFrequentWords < Minitest::Test
+  def test_top_3_frequent_words
+    assert max_k_most_frequent_words(['a', 'b', 'c', 'a', 'c', 'c'], 3) == ['c', 'a', 'b']
+  end
+
+  def test_top_2_frequent_words
+    assert max_k_most_frequent_words(['a', 'b', 'c', 'a', 'c', 'c'], 2) == ['c', 'a']
+  end
+
+  def test_top_frequent_word
+    assert max_k_most_frequent_words(['a', 'b', 'c', 'a', 'c', 'c'], 1) == ['c']
+  end
+
+  def test_no_frequent_word_given_zero_k
+    assert max_k_most_frequent_words(['a', 'b', 'c', 'a', 'c', 'c'], 0) == []
+  end
+
+  def test_no_frequent_word_given_empty_word_list
+    assert max_k_most_frequent_words([], 1) == []
+  end
+
+  def test_all_frequent_words_given_k_too_large
+    assert max_k_most_frequent_words(['a', 'a'], 2) == ['a']
+  end
+end