joblib simple benchmark

joblib_bench.py

import time
import sys

import numpy as np


np.random.seed(0)
N = int(1e7)
big_array = np.random.random(N)

big_dict = {i: np.random.random(10) for i in range(int(1e5))}


def test(joblib_module, to_pickle):
    tic = time.time()
    joblib_module.dump(to_pickle, '/tmp/big_array.gz', compress=1)
    toc = time.time()
    print('Dumping took {:.1f} s'.format(toc - tic))

    tic = time.time()
    joblib_module.load('/tmp/big_array.gz')
    toc = time.time()
    print('Loading took {:.1f} s'.format(toc - tic))


def test_harness(joblib):
    print('=' * 80)
    print('joblib file:', joblib.__file__)
    print('-' * 20)
    print('big array')
    test(joblib, big_array)
    print('-' * 20)
    print('big dict')
    test(joblib, big_dict)


from sklearn.externals import joblib as joblib_stable
import joblib as joblib_dev

print('Python: {0[0]}.{0[1]}'.format(sys.version_info))
test_harness(joblib_stable)
test_harness(joblib_dev)