Udacity Assessment Answer - Machine Learning Nanodegree

word_frequencies_top.txt

"""Count words."""
count={}

def count_words(s, n):
    """Return the n most frequently occuring words in s."""

    wordlist = s.split()
    wordfreq = [wordlist.count(p) for p in wordlist]
    zip_l=dict(zip(wordlist,wordfreq))
    sort=sorted(zip_l.iteritems(),key=lambda(k,v):(-v,k))
    top_n=sort[0:n]
    return top_n

def test_run():
    """Test count_words() with some inputs."""
    print count_words("cat bat mat cat bat cat", 3)
    print count_words("betty bought a bit of butter but the butter was bitter", 3)

if __name__ == '__main__':
    test_run()