A Bloom filter based on Max Burstein's implementation (http://maxburstein.com/blog/creating-a-simple-bloom-filter/)

bloomfilter.py

from bitarray import bitarray
import mmh3
from math import log, pow, ceil

class BloomFilter:
  def __init__(self, size, hash_count):
		self.size = size
		self.hash_count = hash_count
		self.bit_array = bitarray(size)
		self.bit_array.setall(0)

	def add(self, string):
		for seed in xrange(self.hash_count):
			result = mmh3.hash(string, seed) % self.size
			self.bit_array[result] = 1

	def lookup(self, string):
		for seed in xrange(self.hash_count):
			result = mmh3.hash(string, seed) % self.size
			if self.bit_array[result] == 0:
				return False
		return True

	@staticmethod
	def suggest_sizes(n, p):
		"""Given an expected number of items and probability of false positives, suggests an appropriate size and hash count for a Bloom filter.
	n - expected number of items in filter
	p - probability of false positives (0.0 - 1.0)

	http://en.wikipedia.org/wiki/Bloom_filter#Probability_of_false_positives"""
		if not (0.0 <= p <= 1.0):
			raise ValueError("False probability percentage must be between 0.0 and 1.0")
		if n <= 0:
			raise ValueError("Number of items must be greater than 0")

		l2 = log(2)
		m = -n*log(p)/pow(l2, 2)
		k = (m/n) * l2

		return (int(ceil(m)), int(ceil(k)))

	@staticmethod
	def create_suggested(n, p):
		"""Given an expected number of items and probability of false positives, returns a BloomFilter of appropriate size and hash count.
	n - expected number of items in filter
	p - probability of false positives (0.0 - 1.0)"""
		suggested_sizes = BloomFilter.suggest_sizes(n, p)

		return BloomFilter(suggested_sizes[0], suggested_sizes[1])