Python Sorted Container Types: SortedList, SortedDict, and SortedSet
Python’s standard library is great until you need a sorted collections type. Many will attest that you can get really far without one, but the moment you really need a sorted list, dict, or set, you’re faced with a dozen different implementations, most using C-extensions without great documentation and benchmarking.
In Python, we can do better. And we can do it in pure-Python!
>>> sl = sortedcontainers.SortedList(xrange(10000000)) >>> 1234567 in sl True >>> sl 7654321 >>> sl.add(1234567) >>> sl.count(1234567) 2 >>> sl *= 3 >>> len(sl) 30000003
Note: don’t try this without at least a half gigabyte of memory. In Python an integer requires about 24 bytes. SortedList will add about 8 bytes per object stored in the container. That’s pretty hard to beat as it’s the cost of a pointer to each object. It’s also 66% less overhead than a typical binary tree implementation (e.g. red-black tree, avl tree, aa tree, splay tree, treap, etc.) for which every node must also store two pointers to children nodes.
SortedContainers takes all of the work out of Python sorted collections - making your deployment and use of Python easy. There’s no need to install a C compiler or pre-build and distribute custom extensions. Performance is a feature and testing has 100% coverage with unit tests and hours of stress.
Alex Martelli, Wikipedia
Good stuff! … I like the simple, effective implementation idea of splitting the sorted containers into smaller “fragments” to avoid the O(N) insertion costs.
Jeff Knupp, Review of SortedContainers
That last part, “fast as C-extensions,” was difficult to believe. I would need some sort of Performance Comparison to be convinced this is true. The author includes this in the docs. It is.
Kevin Samuel, Formations Python
I’m quite amazed, not just by the code quality (it’s incredibly readable and has more comment than code, wow), but the actual amount of work you put at stuff that is not code: documentation, benchmarking, implementation explanations. Even the git log is clean and the unit tests run out of the box on Python 2 and 3.
Mark Summerfield, a short plea for Python Sorted Collections
Python’s “batteries included” standard library seems to have a battery missing. And the argument that “we never had it before” has worn thin. It is time that Python offered a full range of collection classes out of the box, including sorted ones.
- Fully documented
- Benchmark comparison (alternatives, runtimes, load-factors)
- 100% test coverage
- Hours of stress testing
- Performance matters (often faster than C implementations)
- Compatible API (nearly identical to popular blist and rbtree modules)
- Feature-rich (e.g. get the five largest keys in a sorted dict: d.iloc[-5:])
- Pragmatic design (e.g. SortedSet is a Python set with a SortedList index)
- Developed on Python 2.7
- Tested on CPython 2.7, 3.2, 3.3, 3.4, 3.5, 3.6 and PyPy, PyPy3
$ pip install sortedcontainers
You can access documentation in the interpreter with Python’s built-in help function:
>>> from sortedcontainers import SortedList, SortedSet, SortedDict >>> help(SortedList)
Complete documentation including performance comparisons is available at http://www.grantjenks.com/docs/sortedcontainers/ .
For those wanting more details, this part of the documentation describes introduction, implementation, performance, and development.
Collaborators are welcome!
- Check for open issues or open a fresh issue to start a discussion around a bug. There is a Contributor Friendly tag for issues that should be used by people who are not very familiar with the codebase yet.
- Fork the SortedContainers repository on GitHub and start making your changes to a new branch.
- Write a test which shows that the bug was fixed.
- Send a pull request and bug the maintainer until it gets merged and published.
Copyright 2014-2016 Grant Jenks
Licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License. You may obtain a copy of the License at
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.
Release history Release notifications
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size & hash SHA256 hash help||File type||Python version||Upload date|
|sortedcontainers-1.5.10-py2.py3-none-any.whl (32.0 kB) Copy SHA256 hash SHA256||Wheel||py2.py3||Apr 22, 2018|
|sortedcontainers-1.5.10.tar.gz (12.0 MB) Copy SHA256 hash SHA256||Source||None||Apr 22, 2018|