RapidFuzz/bench/benchmark_fuzz.py

# todo combine benchmarks of scorers into common code base
import timeit

import numpy as np
import pandas


def benchmark(name, func, setup, lengths, count):
    print(f"starting {name}")
    start = timeit.default_timer()
    results = []
    for length in lengths:
        test = timeit.Timer(func, setup=setup.format(length, count))
        results.append(min(test.timeit(number=1) for _ in range(7)) / count)
    stop = timeit.default_timer()
    print(f"finished {name}, Runtime: ", stop - start)
    return results

setup ="""
from rapidfuzz import fuzz as rfuzz
from fuzzywuzzy import fuzz
import string
import random
random.seed(18)
characters = string.ascii_letters + string.digits + string.whitespace + string.punctuation
a      = ''.join(random.choice(characters) for _ in range({0}))
b_list = [''.join(random.choice(characters) for _ in range({0})) for _ in range({1})]
"""

lengths = list(range(1,512,2))
count = 1000

def scorer_benchmark(funcname):
    time_rapidfuzz = benchmark("rapidfuzz",
        f'[rfuzz.{funcname}(a, b) for b in b_list]',
        setup, lengths, count)

    time_fuzzywuzzy = benchmark("fuzzywuzzy",
        f'[fuzz.{funcname}(a, b) for b in b_list]',
        setup, lengths, count)

    df = pandas.DataFrame(data={
        "length": lengths,
        "rapidfuzz": time_rapidfuzz,
        "fuzzywuzzy": time_fuzzywuzzy,
    })

    df.to_csv(f"results/{funcname}.csv", sep=',',index=False)

scorer_benchmark("ratio")
scorer_benchmark("partial_ratio")
scorer_benchmark("token_sort_ratio")
scorer_benchmark("token_set_ratio")
scorer_benchmark("partial_token_sort_ratio")
scorer_benchmark("partial_token_set_ratio")
scorer_benchmark("WRatio")

# token_ratio is unique to RapidFuzz
time_token_ratio = benchmark("token_ratio",
    f'[rfuzz.token_ratio(a, b, processor=None) for b in b_list]',
    setup, lengths, count)

# this gets very slow, so only benchmark it for smaller values
time_token_ratio_simple = benchmark("fuzzywuzzy",
    f'[max(rfuzz.token_sort_ratio(a, b, processor=None), rfuzz.token_set_ratio(a, b, processor=None)) for b in b_list]',
    setup, lengths, count)

df = pandas.DataFrame(data={
    "length": lengths,
    "token_ratio": time_token_ratio,
    "max(token_sort_ratio, token_set_ratio)": time_token_ratio_simple,
})

df.to_csv(f"results/token_ratio.csv", sep=',',index=False)

# partial_token_ratio is unique to RapidFuzz
time_partial_token_ratio = benchmark("token_ratio",
    f'[rfuzz.partial_token_ratio(a, b, processor=None) for b in b_list]',
    setup, lengths, count)

# this gets very slow, so only benchmark it for smaller values
time_partial_token_ratio_simple = benchmark("fuzzywuzzy",
    f'[max(rfuzz.partial_token_sort_ratio(a, b, processor=None), rfuzz.partial_token_set_ratio(a, b, processor=None)) for b in b_list]',
    setup, lengths, count)

df = pandas.DataFrame(data={
    "length": lengths,
    "partial_token_ratio": time_partial_token_ratio,
    "max(partial_token_sort_ratio, partial_token_set_ratio)": time_partial_token_ratio_simple,
})

df.to_csv(f"results/partial_token_ratio.csv", sep=',',index=False)