mathdatech/Python
1
0
Fork 0
mirror of https://github.com/TheAlgorithms/Python.git synced 2025-04-28 08:33:36 +00:00
Code Issues Packages Projects Releases Wiki Activity

has functions constant

Browse Source
This commit is contained in:
Isidro Arias 2023-04-06 11:20:41 +02:00
parent 173ab0ea96
commit 08bc970b7a
1 changed files with 6 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
data_structures/hashing/bloom_filter.py
View File

@ -7,6 +7,9 @@ import string
class Bloom: class Bloom:
# number of hash functions is fixed
HASH_FUNCTIONS = (sha256, md5)
def __init__(self, size=8): def __init__(self, size=8):
self.bitstring = 0b0 self.bitstring = 0b0
self.size = size self.size = size
@ -42,7 +45,7 @@ class Bloom:
def hash(self, value): def hash(self, value):
res = 0b0 res = 0b0
for func in (sha256, md5): for func in HASH_FUNCTIONS:
b = func(value.encode()).digest() b = func(value.encode()).digest()
position = int.from_bytes(b, "little") % self.size position = int.from_bytes(b, "little") % self.size
res |= 2**position res |= 2**position
@ -74,8 +77,7 @@ def test_probability(m=64, n=20):
for a in added: for a in added:
b.add(a) b.add(a)
# number of hash functions is fixed k = len(b.HASH_FUNCIONS)
k = 2
n_ones = bin(b.bitstring).count("1") n_ones = bin(b.bitstring).count("1")
expected_probability = (n_ones / m) ** k expected_probability = (n_ones / m) ** k
@ -95,7 +97,7 @@ def test_probability(m=64, n=20):
assert ( assert (
abs(expected_probability - fail_rate) <= 0.05 abs(expected_probability - fail_rate) <= 0.05
) # 5% margin calculated experiementally ) # 5% absolute margin calculated experiementally
if __name__ == "__main__": if __name__ == "__main__":