Serialization Benchmark

benchmark.py

"""
Dependencies:

    pip install \
        tabulate ujson msgpack msgpack_numpy numpy pyarrow
"""

import sys
from timeit import timeit
import pickle
from tabulate import tabulate


def get_tests(is_numpy):
    tests = [
        # (title, setup, enc_test, dec_test)
        ('msgpack-python', 'import msgpack; import msgpack_numpy as m; m.patch(); src = msgpack.dumps(d)', 'msgpack.dumps(d)', 'msgpack.loads(src)'),
        ('pyarrow', 'import pyarrow as pa; src = pa.serialize(d).to_buffer()', 'pa.serialize(d).to_buffer()', 'pa.deserialize(src)'),
    ]

    for k in range(2, min(pickle.HIGHEST_PROTOCOL, 3) + 1):
        if sys.version_info.major == 3:
            setup_pickle = 'import pickle ; src = pickle.dumps(d, {})'.format(k)
        else:
            setup_pickle = 'import cPickle as pickle; src = pickle.dumps(d, {})'.format(k)
        tests.append(('pickle-protocol{}'.format(k), setup_pickle, 'pickle.dumps(d, {})'.format(k), 'pickle.loads(src)'))

    if not is_numpy:
        tests.extend([
            ('json', 'import json; src = json.dumps(d)', 'json.dumps(d)', 'json.loads(src)'),
            ('ujson', 'import ujson; src = ujson.dumps(d)', 'ujson.dumps(d)', 'ujson.loads(src)')
            ])
    return tests


def run_tests(tests, data, loops):
    enc_table = []
    dec_table = []

    print("Running tests (%d loops each)" % loops)

    for title, mod, enc, dec in tests:
        mod = data + ' ; ' + mod
        print("Running " + title)

        #print("  [Encode]", enc)
        result = timeit(enc, mod, number=loops)
        enc_table.append([title, result])

        #print("  [Decode]", dec)
        result = timeit(dec, mod, number=loops)
        dec_table.append([title, result])

    enc_table.sort(key=lambda x: x[1])
    enc_table.insert(0, ['Method', 'Seconds'])

    dec_table.sort(key=lambda x: x[1])
    dec_table.insert(0, ['Method', 'Seconds'])

    print("\nEncoding Test (%d loops)" % loops)
    print(tabulate(enc_table, headers="firstrow"))

    print("\nDecoding Test (%d loops)" % loops)
    print(tabulate(dec_table, headers="firstrow"))


print("Benchmarking plain data ...............")
plain_data = '''d = {
    'words': """
        Lorem ipsum dolor sit amet, consectetur adipiscing
        elit. Mauris adipiscing adipiscing placerat.
        Vestibulum augue augue,
        pellentesque quis sollicitudin id, adipiscing.
        """ * 100,
    'list': list(range(100)) * 300,
    'dict': dict((str(i),'a') for i in range(5000)),
    'int': 3000,
    'float': 100.123456
}'''
run_tests(get_tests(False), plain_data, 1000)

print("Benchmarking numpy data ...............")
numpy_data = """
import numpy as np
d = {
  "arrays": [np.random.rand(1000, 353) for k in range(10)]
} """
run_tests(get_tests(True), numpy_data, 300)

Outputs with Python 3.7:

Benchmarking plain data ...............
Running tests (1000 loops each)
Running msgpack-python
Running pyarrow
Running pickle-protocol2
Running pickle-protocol3
Running json
Running ujson

Encoding Test (1000 loops)
Method              Seconds
----------------  ---------
pickle-protocol3   0.608393
pickle-protocol2   0.610825
msgpack-python     1.07399
ujson              1.09344
pyarrow            2.66014
json               2.68677

Decoding Test (1000 loops)
Method              Seconds
----------------  ---------
msgpack-python     0.452591
pickle-protocol2   0.722493
pickle-protocol3   0.723882
pyarrow            0.812569
ujson              0.97694
json               2.04448
Benchmarking numpy data ...............
Running tests (300 loops each)
Running msgpack-python
Running pyarrow
Running pickle-protocol2
Running pickle-protocol3

Encoding Test (300 loops)
Method              Seconds
----------------  ---------
pyarrow             3.3704
pickle-protocol3    6.12827
msgpack-python      6.18487
pickle-protocol2   45.2517

Decoding Test (300 loops)
Method               Seconds
----------------  ----------
pyarrow            0.0100455
pickle-protocol3   1.04249
msgpack-python     1.05066
pickle-protocol2  40.2511