Compute the inverse document frequency

idf.py

from __future__ import division

from collections import Counter
import math

def idf(documents):
    freq = Counter()
    for doc in documents:
        words = set(doc.split())
        freq.update(words)
    res = dict()
    for w, f in freq.items():
        res[w] = math.log(len(documents) / f)
    return res

inverse document frequency