def _test_prefix_false_positives(constructor = BloomFilter):
for error_rate in [0.0001, 0.001, 0.01, 0.1, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9]:
a = constructor(error_rate, 10374, prefix="A")
b = constructor(error_rate, 10374, prefix="B")
c = constructor(error_rate, 10374, prefix="C")
d = constructor(error_rate, 10374, prefix="D")
p(a)
print "Estimated errors:", a.error_rate, "->", a.error_rate * b.error_rate, "->", a.error_rate * b.error_rate * c.error_rate, "->", a.error_rate * b.error_rate * c.error_rate * d.error_rate
#we fill each bloomfilter up to its capacity
data = ["%i" % i for i in xrange(a.capacity)]
map(a.add, data)
map(b.add, data)
map(c.add, data)
map(d.add, data)
errors = 0
two_errors = 0
three_errors = 0
four_errors = 0
#we check what happens if we check twice the capacity
for i in xrange(a.capacity * 2):
if "X%i" % i in a:
errors += 1
if "X%i" % i in b:
two_errors += 1
if "X%i" % i in c:
three_errors += 1
if "X%i" % i in d:
four_errors += 1
print "Errors:", errors, "~", errors / (i + 1.0), "Two-Errors:", two_errors, "~", two_errors / (i + 1.0), "Three-Errors:", three_errors, "~", three_errors / (i + 1.0), four_errors, "~", four_errors / (i + 1.0)
print
def test(bits, count, constructor = BloomFilter):
ok = 0
create_begin = time()
bloom = constructor(0.0001, bits)
fill_begin = time()
for i in xrange(count):
if i % 2 == 0:
bloom.add(str(i))
check_begin = time()
for i in xrange(count):
if (str(i) in bloom) == (i % 2 == 0):
ok += 1
write_begin = time()
string = str(bloom)
write_end = time()
print "create: {create:.1f}; fill: {fill:.1f}; check: {check:.1f}; write: {write:.1f}".format(create=fill_begin-create_begin, fill=check_begin-fill_begin, check=write_begin-check_begin, write=write_end-write_begin)
print string.encode("HEX")[:100], "{len} bytes; ({ok}/{total} ~{part:.0%})".format(len=len(string), ok=ok, total=count, part=1.0*ok/count)
def _test_false_positives(constructor = BloomFilter):
for error_rate in [0.001, 0.01, 0.1, 0.5]:
begin = time()
# if constructor == BloomFilter:
# a = constructor(error_rate, 1024*8)
# capacity = a.capacity
# else:
a = constructor(1024*8, error_rate)
capacity = a.get_capacity(error_rate)
print "capacity:", capacity, " error-rate:", error_rate, "bits:", a.size, "bytes:", a.size / 8
data = ["%i" % i for i in xrange(capacity)]
map(a.add, data)
errors = 0
for i in xrange(200000):
if "X%i" % i in a:
errors += 1
end = time()
print "%.3f"%(end-begin), "Errors:", errors, "/", i + 1, " ~ ", errors / (i + 1.0)
print
