Downloads

Home News Downloads Tutorials User Manual Reference Manual Developer Manual Forums Team Members Future Work FAQ Links

Installing into an existing Python environment Topographica itself is platform independent, but it depends on external packages, the installation of which varies by platform. This page assumes you already have Python installed (with Tkinter, if you wish to use the GUI), so is most useful for people who already have a Python environment they want to use, or who are using a platform other than those already described, and who do not need version control. We have tested Topographica with Python 2.5 and 2.6. There are two options for ensuring your Python environment has the appropriate dependencies and then installing Topographica: the first is to use Python's easy_install or pip, which will get the dependencies automatically and then install Topographica; the second is to get the dependencies first yourself, then install Topographica from the Python source (the "python setup.py install" way). (1) easy_install or pip Assuming your system already has easy_install or pip, typing easy_install topographica or pip install topographica (with administrative privileges, e.g. with sudo) should install Topographica into your system Python directory. Alternatively, you can use the appropriate options (e.g. --prefix) for easy_install or pip to choose a different location. If you want to install Topographica's optional dependencies, you can use easy_install or pip to additionally install IPython, MatPlotLib, gmpy, and SciPy (e.g. sudo easy_install ipython gmpy matplotlib scipy). Sometimes, easy_install and pip can encounter problems while installing the packages on which Topographica depends. Fixing these problems is usually straightforward, as described in the FAQ. Once installation has completed, you can proceed to the After Installation instructions. (2) Install from Python source ("python setup.py install") Binaries of Topographica's dependencies are available for many platforms, and many package managers also include them (e.g. apt-get on Ubuntu or Fink/Macports on Mac). First install the required dependencies NumPy and PIL. Optionally, then install the other recommended dependencies MatPlotLib, gmpy, SciPy, and IPython. Once the dependencies are installed, download the source distribution, unpack the archive, and type python setup.py install. Alternatively, choose a different installation location by passing the appropriate option (e.g. --user or --prefix). Once complete, proceed to the After Installation instructions.

Hosted by: James A. Bednar (jbednar@inf.ed.ac.uk) Last update: Thu Aug 5 15:21:43 BST 2010.