mathdatech/Python
1
0
Fork 0
mirror of https://github.com/TheAlgorithms/Python.git synced 2025-03-03 13:28:40 +00:00
Code Issues Packages Projects Releases Wiki Activity

from fail to erro

Browse Source
This commit is contained in:
Isidro Arias 2023-04-06 14:07:01 +02:00
parent e4d39db77e
commit 7629686cc7
1 changed files with 11 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
data_structures/hashing/bloom_filter.py
View File

@ -73,30 +73,29 @@ def random_string(size):
def test_probability(m=64, n=20): def test_probability(m=64, n=20):
b = Bloom(size=m) b = Bloom(size=m)
k = len(b.HASH_FUNCTIONS)
estimated_error_rate_beforehand = (1 - (1 - 1 / m) ** (k * n)) ** k
added = {random_string(10) for i in range(n)} added = {random_string(10) for i in range(n)}
for a in added: for a in added:
b.add(a) b.add(a)
k = len(b.HASH_FUNCTIONS)
n_ones = bin(b.bitstring).count("1") n_ones = bin(b.bitstring).count("1")
expected_probability = (n_ones / m) ** k estimated_error_rate = (n_ones / m) ** k
expected_probability_wikipedia = (1 - (1 - 1 / m) ** (k * n)) ** k
not_added = {random_string(10) for i in range(1000)} not_added = {random_string(10) for i in range(1000)}
fails = 0 errors = 0
for string in not_added: for string in not_added:
if b.exists(string): if b.exists(string):
fails += 1 errors += 1
fail_rate = fails / len(not_added) error_rate = errors / len(not_added)
print(f"total = {len(not_added)}, fails = {fails}, fail_rate = {fail_rate}") print(f"total = {len(not_added)}, errors = {errors}, error_rate = {error_rate}")
print(f"{expected_probability=}") print(f"{estimated_error_rate=}")
print(f"{expected_probability_wikipedia=}") print(f"{estimated_error_rate_beforehand=}")
assert ( assert (
abs(expected_probability - fail_rate) <= 0.05 abs(estimated_error_rate - error_rate) <= 0.05
) # 5% absolute margin calculated experiementally ) # 5% absolute margin calculated experiementally