Download

This page contains nightly builds from the code repository. These are typically stable and we recommend using them.

Windows

Full package: Snapshot of Orange with Python 2.7 and required libraries
This package is recommended to those installing Orange for the first time. It includes all required libraries (Python, PythonWin, NumPy, PyQt, PyQwt ...), though it will not change any libraries you might already have.
(Also available: Orange for Python 2.6, Orange for Python 2.5)
Pure Orange: Snapshot of Orange for Python 2.7
Use this version if you are updating from an earlier snapshot. You can install it over your existing installation.
(Also available: Orange for Python 2.6, Orange for Python 2.5)

Mac OS X

Bundle: Orange Snapshot
This is an universal bundle with everything packed in and ready for an unadvanced user.
easy_install/pip:
Orange is available as a PyPi package. To install run
```
easy_install numpy && easy_install orange
```
Fink:
If you would like to feel the real power of Orange you should try its scripting part. Although this is possible through Python packed in a bundle (see Use Orange From DMG) we recommend using a Fink package management system to deploy Orange system-wide. You should have Fink, X11 version 2.3.0 or later and Xcode installed on a system and run this install script to add information about our packages and our binary packages repository (so you will not have to compile everything on your machine) to your Fink installation. You should run this script whenever you would like to update information about our packages (and thus be able to update them). Orange is split into two packages: orange-gui-svn (graphical interface) and orange-svn, with suffix for Python version. Many example scripts also require the matplotlib Fink package.

Linux

Debian

We have a repository with Orange packages for i386 and amd64 platforms. You should add the repository to your list of repositories by adding those two lines to /etc/apt/sources.list file:

deb http://orange.biolab.si/debian squeeze main
deb-src http://orange.biolab.si/debian squeeze main

After updating the list of known packages (aptitude update) you can install, for example, orange-canvas package (aptitude install orange-canvas). Packages are not signed as they are made automatically so you will probably be warned about this.

Sources

setup.py

To build and install Orange you can use the setup.py in the root orange directory (requires GCC, Python and numpy development headers). To use it unpack the nightly sources and run:

python setup.py build
sudo python setup.py install

This will also install orange-canvas script so you can start Orange Canvas from the command line.

To install orange locally run:

python setup.py install --user

This will install orange in /home/<username>/.local/lib/pythonX.Y/site-packages/orange/.

make (for developers)

Source distribution also includes makefiles for Linux and Mac OS X and project files for MS Visual Studio 2008. To compile Orange, you will need an appropriate compiler, Python and NumPy.

See building from source wiki page for more information.

C4.5 files

We have no rights to distribute the original C4.5 in any form, but we can provide assistance for building an Orange plug-in from the sources you download yourself from the Ross Quinlan's site. The files required for compiling C4.5 are here and instructions here.