courses.csail.mit.edu/6.006/fall11/lectures/lecture2

document_distance.py

import sys
import math

def make_words(document):
    return document.split()

def make_frequency(document):
    frequency = {}
    for w in make_words(document):
        if w in frequency:
            frequency[w] += 1
        else:
            frequency[w] = 1
    return frequency

def dot_product(f1, f2):
    sum = 0
    for k in f1:
        if k in f2:
            sum += f1[k] * f2[k]
    return sum


def d(d1, d2):
    d1f = make_frequency(d1)
    d2f = make_frequency(d2)
    return math.acos(dot_product(d1f, d2f)/math.sqrt(dot_product(d1f, d1f) * dot_product(d2f, d2f)))
    

if __name__ == '__main__':
    file1 = sys.argv[1]
    file2 = sys.argv[2]
    with open(file1) as f1:
        d1 = f1.read()
    with open(file2) as f2:
        d2 = f2.read()
    print(d(d1,d2))