pombredanne / intdict-1

High performance integer->object mapping in python

intdict

intdict is a high performance integer -> PyObject data structure intended for mapping 64 bit integers such as what might be returned by hash or id to a Python object.

Installation

pip install intdict

Usage

>>> from intdict import intdict
>>> d = intdict()
>>> d[1] = 'one'
>>> d[2] = 'two'
>>> d[1]
'one'
>>> d[3]
KeyError

You may also preallocate intdict objects to a specific size for even better performance.

>>> d = intdict(10000)

Cavets

intdict stores all signed 64 bit integers except for the value of -1. This happens to also be exactly the range of the hash function on a 64 bit machine and a superset of the range of values id can return.

Benchmarks

If you can live within intdict's mild restrictions, here are the performance numbers you can enjoy. These timings were taken from an Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz running Ubuntu 15.04.

Storing 1,000,000 sequential numbers

Dict type	Seconds	Memory
dict	15.3	1,153Mb
int dict	1.3	773Mb
prealloced	0.3	773Mb

Storing 1,000,000 random numbers

Dict type	Seconds	Memory
dict	3	1,234Mb
intdict	1.5	854Mb
prealloced	1.3	854Mb