def experiment4(n, m, k):
"""
Run experiment 4, where each key is inserted a random number 20 or less times, and all insertions are in a random order
:param n: number of distinct keys
:param m: the size of the bloom filter
:param k: the number of hashes used by the bloom filter
"""
distinct_keys = generate_keys(n)
bf = RevisedExtendedBloomFilter(m, k)
all_keys = []
c_list = {}
for e in distinct_keys:
c = randint(0, 20)
c_list[e] = c
for _ in range(c):
all_keys.append(e)
fisher_yates_shuffle(all_keys)
for e in all_keys:
bf.insert(e)
# Get the false-positive rate
return calc_false_positive_rate_random(bf, distinct_keys, c_list)
def experiment7(n, m, k):
"""
Run experiment 7, where each key is inserted a poisson random variable with lambda=20 times, and shuffled
:param n: number of distinct keys
:param m: the size of the bloom filter
:param k: the number of hashes used by the bloom filter
"""
distinct_keys = generate_keys(n)
bf = RevisedExtendedBloomFilter(m, k)
all_keys = []
c_list = {}
for e in distinct_keys:
c = np.random.poisson(20)
c_list[e] = c
for _ in range(c):
all_keys.append(e)
fisher_yates_shuffle(all_keys)
for e in all_keys:
bf.insert(e)
# Get the false-positive rate
return calc_false_positive_rate_random(bf, distinct_keys, c_list)
def experiment2(n, m, k):
"""
Run experiment 2, where each key is inserted 20 times in a row
:param n: number of distinct keys
:param m: the size of the bloom filter
:param k: the number of hashes used by the bloom filter
"""
distinct_keys = generate_keys(n)
bf = RevisedExtendedBloomFilter(m, k)
for e in distinct_keys:
for _ in range(20):
bf.insert(e)
# Get the false-positive rate
return calc_false_positive_rate(bf, distinct_keys, 20)
def experiment3(n, m, k):
"""
Run experiment 3, where each key is inserted 20 times, but all insertions are in a random order
:param n: number of distinct keys
:param m: the size of the bloom filter
:param k: the number of hashes used by the bloom filter
"""
distinct_keys = generate_keys(n)
bf = RevisedExtendedBloomFilter(m, k)
all_keys = distinct_keys * 20
fisher_yates_shuffle(all_keys)
for e in all_keys:
bf.insert(e)
# Get the false-positive rate
return calc_false_positive_rate(bf, distinct_keys, 20)