Booplicate / LFUCache

Python implementation of LFU (Least Frequently Used) cache

LFUCache

Description.

This module implements LFU cache in python using just one class, nevertheless, for the most basic operations (add/get) should be still quite fast. You can also remove cache entries by priority or key, although, this may be slow depending on the size. Uneless you're making something fancy, usually it should be enough to just use the create_lfu_cache decorator on your function.

Example

from lfu_cache import create_lfu_cache

@create_lfu_cache(limit=1024, typed=True)
def expensive_function(arg):
    *heavy processing*
    return arg

That'll cache 1024 precisely stored inputs (since typed is set True) and outputs. Bear in mind, the first execution may take more time, but you save more time with each next call. You can access the original (w/o the decorator, not affected by the cache system) function using the __wrapped__ property, the __cache__ property can be used to access the LFUCache object of the function. The cache object has 2 useful property for debugging: hits and misses, showing how effective the cache was reused.

Limitations.

Arguments to the cached function must be hashable.