Installing Eureqa

From SUMOwiki
Revision as of 17:57, 8 February 2012 by Mnguyen (talk | contribs) (New page: This page details how to get Eureqa ([http://creativemachines.cornell.edu/eureqa]) working in SUMO. ==Installation== * Go to [http://creativemachines.cornell.edu/eureqa the Eureqa websit...)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

This page details how to get Eureqa ([1]) working in SUMO.

Installation

  • Go to the Eureqa website and download the latest Eureqa standalone server for your operating system
  • Place the Eureqa server in the @EureqaFactory folder which is located in the src folder of the SUMO root.
  • Make sure you have a mex compilation of the eureqa_select_SUMO.cpp which is based on code from Johannes Jenkner which you can find here. There are two possibilities either the required mex file is already included with your SUMO installation in the contrib folder (a subfolder of the src folder) or you have to build it your self.

Building the mex files yourself

On linux

  • Install the Boost C++ libraries (Boost.org) following the instructions on the Boost website. Make sure you build the asio, system, thread, date_time and serialization libraries
  • Download the Eureqa API
  • Extract the eureqa_select.zip which you can find in the EureqaModel folder
  • Edit the Makefile to point to your Matlab install, the Boost include files, the Boost library files and the Eureqa API files. Note that you don't need the "<" and ">" brackets
  • Edit Makefile to change FILE = eureqa_select -> FILE = eureqa_select_SUMO
  • Type in make in a linux terminal and the building process should start
  • When the building process is done you will have eureqa_select.mexglx file or a eureqa_select.mexa64 depending on your system architecture to put in the contrib folder.


On Windows using Visual Studio 2010 and Matlab 2009b

Note that these installation instructions require a file (mexversion.rc) which is not available in later versions of Matlab. We have yet to find a solution :).

  • Install the Boost C++ libraries for Windows (Boost.org). The easiest way is by watching this youtube video. Note that if you are working on a 64-bit system you need to add this "address-model=64" option to bjam.