Frequent Distribution sorted by frequency 

 1 import nltk
 2 
 3 def freq_sorted(text,ranklimit):
 4     fd=nltk.FreqDist(text)
 5     cumulative = 0.0
 6     for rank, (word,freq) in enumerate(sorted(fd.items(), key=lambda x: (-1*x[1], x[0]))[:ranklimit]):
 7         cumulative += fd[word] * 100 / fd.N()
 8         print "%3d %6.2f%% %s" % (rank+1, cumulative, word)
 9 
10 def test_freq_sorted():
11     freq_sorted(nltk.corpus.brown.words(),15)

结果为：

  1   5.00% the
  2  10.00% ,
  3  14.00% .
  4  17.00% of
  5  19.00% and
  6  21.00% to
  7  22.00% a
  8  23.00% in
  9  23.00% that
 10  23.00% is
 11  23.00% was
 12  23.00% for
 13  23.00% ``
 14  23.00% ''
 15  23.00% The

 

posted on 2015-04-23 13:42  karis  阅读(155)  评论(0)  编辑  收藏  举报

导航