dict_UserDict_RichDict_benchmarks.py

import timeit
import pprint
import rounder # requires installation from PyPi


def compare(
    __snippet1,
    __snippet2,
    *args,
    number=10_000_000,
    repeat=7,
    setup="from collections import UserDict"):
    snippets = [__snippet1, __snippet2, *args]
    results = {}
    for i, snippet in enumerate(snippets):
        name = snippet if len(snippet) < 30 else f"snippet {i + 1}"
        results[name] = min(timeit.repeat(
            snippet, number=number, repeat=repeat, setup=setup
            )) / number,
    results = rounder.signif_object(results, digits=4)
    pprint.pprint(results)


setup = """from collections import UserDict
from collections.abc import Sequence
from typing import Callable

def try_calculate(func: Callable, *args, **kwargs):
    try:
        return func(*args, **kwargs)
    except TypeError:
        return float("nan")

class RichDict(dict):
    measures = {
        "sum": sum,
        "n": len,
        "mean": lambda x: sum(x) / len(x),
    }
    
    def summarize(self):
        statistics = {}
        for k, v in self.items():
            if isinstance(v, str):
                statistics[k] = {"n": len(v)}
            elif isinstance(v, Sequence):
                statistics[k] = {
                    name: try_calculate(func, v)
                    for name, func
                    in self.measures.items()
                }
        return statistics

"""

compare("RichDict()", "dict()", "UserDict()", setup=setup)

###
compare(
    "UserDict({i: i**2 for i in range(1000)})",
    "RichDict({i: i**2 for i in range(1000)})",
    "{i: i**2 for i in range(1000)}",
    number=100_000,
    setup=setup)

###
setup += """d = {'x': 1, 'y': 2, 'z': 3}
ud = UserDict(d)
rd = RichDict(d)
"""
compare("ud['x']", "rd['x']", "d['x']", setup=setup)

compare(
    "ud.get('a', None)",
    "rd.get('a', None)",
    "d.get('a', None)",
    setup=setup,
    number=1_000_000)

###
compare("'a' in ud", "'a' in rd", "'a' in d", setup=setup, number=1_000_000)

###
setup += """d_loop = {str(i): i for i in range(100_000)}
ud_loop = UserDict(d)
rd_loop = RichDict(d)
"""
compare(
    "for i, v in ud_loop.items(): pass",
    "for i, v in rd_loop.items(): pass",
    "for i, v in d_loop.items(): pass",
    setup=setup,
    number=10_000
)

###
setup += """
d = {str(i): i for i in range(100_000)}
rd = RichDict(d)
ud = UserDict(d)
"""
compare(
    "for i, v in ud.items(): pass",
    "for i, v in rd.items(): pass",
    "for i, v in d.items(): pass",
    setup=setup,
    number=10
)