/morris_counter

Memory-efficient probabilistic counter namely Morris Counter

Primary LanguagePythonMIT LicenseMIT

Morris Counter

travis-ci.org pyversion latest version license

Memory-efficient probabilistic counter namely Morris Counter. This module based on the following paper:

  • Robert Morris. Counting large numbers of events in small registers. Communications of the ACM, vol. 21, issue 10, pp. 840-842, 1978.

Currently Morris Counter supports Python 3.5 and higher.

Basic idea of Morris Counter is described as follows:

INSTALLATION

$ pip install morris_counter

While the Morris Counter works builtin modules, using third-party package (numpy and mmh3) leads to improve memory-usage and computation time.

$ pip install numpy mmh3

USAGE

from morris_counter import MorrisCounter

mc = MorrisCounter(size=1000000, dtype='uint8', radix=2, seed=3282)
mc.count('ZOC')
# => 1
mc.increment('ZOC')
mc.count('ZOC')
# => 2
_ = [mc.increment('ZOC') for _ in range(2000)]
mc.count('ZOC')
# => 2048