lib/epitome/corpus.rb in epitome-0.2.0 vs lib/epitome/corpus.rb in epitome-0.2.1

@@ -1,7 +1,8 @@
 require 'matrix'
 require 'stopwords'
+require 'pry'
 
 module Epitome
   class Corpus
     attr_reader :original_corpus
     def initialize(document_collection, lang="en")
@@ -93,10 +94,13 @@
     end
 
     def idf(word)
       # Number of documents in which word appears
       # Inverse Frequency Smooth (as per wikipedia article)
-      Math.log( @n_docs / n_docs_including_w(word) )
+      result = Math.log( @n_docs / n_docs_including_w(word) )
+
+      # Return 1 to avoid words having all the same td_idf by multiplying by 0
+      return result == 0 ? 1.0 : result 
     end
 
     def tf(sentence, word)
       # Number of occurences of word in sentence
       sentence.scan(word).count