Python Distributions

A Python distribution is a software bundle, which contains a Python interpreter and the Python standard library. Installer programs for common operating systems are a frequent mode of distribution. Many Python distributions also have package managers so that you can install or upgrade various Python packages.

Some of the most popular distributions are listed below. Distributions which are marked as “scientific” are ones which come with IPython, numpy, pandas, and matplotlib, at a minimum. All of the distributions provide at least one integrated development environment (IDE) for free. A Python IDE provides a Python-aware code editor integrated with the ability to run code from that editor.

ActiveState ActivePython

Scientific: No, but many scientific packages can be added via the package manager

Platform: AIX, HP-UX, Linux, MacOS X, Solaris, Windows

Overview: http://www.activestate.com/activepython

Downloads: http://www.activestate.com/compare-editions

Package List: too many to list - use search page: http://code.activestate.com/pypm/search:/?tab=name

Package Manager: PyPM

IDE: IDLE, Komodo (must be purchased separately from ActiveState)

Note: ActivePython is one of the oldest Python distributions, but is not particularly geared towards science.

Continuum Analytics Anaconda

Scientific: Yes

Platform: Linux, MacOS X, Windows

Overview: https://store.continuum.io/

Downloads: [Note: Complete distribution is available for free, but requires registration first. Also, academics can get several powerful commercial add-on products for free with proof of affiliation with an educational institution.]

Package List: http://docs.continuum.io/anaconda/pkgs.html

Package Manager: conda

IDE: Spyder

Note: Continuum provides a Python compiler, called Numba, as part of its distribution. This can compile Python code down to machine code and is aware of how to optimize with special consideration for the popular numpy.

Note: Commercial add-on tools are linked against the Intel Math Kernel Library (MKL) for improved numerical performance.

Enthought Canopy

Scientific: Yes

Platform: Linux, MacOS X, Windows

Overview: https://www.enthought.com/products/canopy/

Downloads: [Note: Academics can get the professional version for free by registering for an academic license at https://www.enthought.com/products/canopy/academic/ .]

Package List: https://www.enthought.com/products/canopy/package-index/

Package Manager: Canopy Package Manager

IDE: IDLE, SciTE

Note: Professional version is linked against the Intel Math Kernel Library (MKL) for improved numerical performance.

Python(x,y)

Scientific: Yes

Platform: Windows

Overview: https://code.google.com/p/pythonxy/

Downloads: https://code.google.com/p/pythonxy/wiki/Downloads

Package List: https://code.google.com/p/pythonxy/wiki/StandardPlugins

IDE: SciTE, Spyder

Other Tools: Console (enhanced Windows command line window), WinMerge (differencing and merging of files on Windows)

Note: A variant of this distribution is also available for Linux; please see pythonxy-linux.

WinPython

Scientific: Yes

Platform: Windows

Overview: https://code.google.com/p/winpython/

Downloads: https://code.google.com/p/winpython/downloads/list

Package List: https://code.google.com/p/winpython/wiki/PackageIndex

Package Manager: WinPython Package Manager (WPPM)

IDE: SciTE, Spyder

Other Tools: TortoiseHG (Mercurial version control system integrated into Windows Explorer)

Additional Python Packages

Some Python packages may not be a part of some distributions, but contain files which must be compiled (i.e., they are not “pure Python”). As it can be difficult to compile these files, especially on Windows, there exist third-party repositories of precompiled packages.

Christoph Gohlke’s Windows Binaries

Project Versions

Table Of Contents

Previous topic

Downloads for Presentation

Next topic

Python Packages

This Page