benchmark_6123.py

from __future__ import print_function

import time
import random

from sklearn.feature_extraction import DictVectorizer


def get_sample():
    return {
        'a': random.random(),
        'b': random.random(),
        'c': random.random(),
        'd': random.random(),
    }


def get_x(n_samples=100000):
    return [get_sample() for i in range(n_samples)]


def run_benchmark():
    results = []
    n_samples = [10, 100, 1000, 10000, 100000, 500000]
    for ns in n_samples:
        X = get_x(ns)
        start = time.time()
        DictVectorizer().fit(X)
        end = time.time()
        results.append((ns, end - start))

    for ns, t in results:
        print('%s: %.3f' % (str(ns).rjust(6, ' '), t))


if __name__ == '__main__':
    run_benchmark()

Results without _validate_dictvectorizer_input:

    10: 0.000
   100: 0.000
  1000: 0.001
 10000: 0.009
100000: 0.096
500000: 0.481

and with:

    10: 0.000
   100: 0.001
  1000: 0.006
 10000: 0.058
100000: 0.603
500000: 2.958