Latest release

Version 0.4.0 (released 17th October 2009)

Requirements

• One of the following operating systems
  • Linux (tested with Ubuntu 9.10, Jaunty Jackalope)
  • Windows (tested with Windows Vista)
  • Mac OS X (tested with OS X 10.5)
• Python version 2.6
• C/C++ compiler

Please download the 0.4.0 release of Elefant elefant-0.4.0.tar.bz2, and follow the Install guide for installing Elefant on your system. (The Install guide does also provide help if your system does not yet have a Python interpreter or C/C++ compiler installed).

You can browse the documentation of the API from the sidebar, or download the whole html tree containing the API

You can also check out the Elefant source from the SVN repository.

Release information

• The Elefant toolbox is now compatible with Python version 2.6, and prepared for the move to Python3000.
• In the future, we will move more code from Python to C/C++ in order to use the full power of multi-core/multithreaded architectures via OpenMP. This will also enable the use of Elefant libraries without having to install Python (unless the easy to use graphical user interface is wanted).
• This release contains the Stream module as a first step in the direction of providing library support. Stream aims to be a software framework for the implementation of large scale online learning algorithms. Large scale, in this context, should be understood as something that does not fit in the memory of a standard desktop computer. As the objective is to deal with large amounts of data, a lot of effort is being made to optimise the implementation for speed, sometimes at the expense of conciseness - for example, using standard C I/O libraries instead of C++ stream libraries.