Word error rate script

wer.py

from __future__ import division
import os
import numpy
import argparse


def wer(r, h):
    d = numpy.zeros((len(r)+1)*(len(h)+1), dtype=numpy.uint8)
    d = d.reshape((len(r)+1, len(h)+1))
    for i in xrange(len(r)+1):
        for j in xrange(len(h)+1):
            if i == 0:
                d[0][j] = j
            elif j == 0:
                d[i][0] = i

    for i in xrange(1, len(r)+1):
        for j in xrange(1, len(h)+1):
            if r[i-1] == h[j-1]:
                d[i][j] = d[i-1][j-1]
            else:
                d[i][j] = min(d[i-1][j-1] + 1, d[i][j-1] + 1, d[i-1][j] + 1)

    return d[len(r)][len(h)]


def get_contents(filename):
    with open(filename) as f:
        return f.read().lower()

def read_wer(rfile, hfile):
    r = get_contents(rfile)
    h = get_contents(hfile)
    return wer(r, h)


if __name__ == '__main__':
    parser = argparse.ArgumentParser(description="Word Error Rate")
    parser.add_argument('truth', type=str, nargs=1, help='Path to file containing ground truth')
    parser.add_argument('pred', type=str, nargs=1, help='Path to file containing word prediction')
    args = parser.parse_args()
    truth_file = os.path.join(os.getcwd(), args.truth[0])
    pred_file = os.path.join(os.getcwd(), args.pred[0])
    print 1 - (read_wer(truth_file, pred_file) / len(get_contents(truth_file)))