Modular Remapping Model Installation

Software Requirements

The model and analysis code consist of the Python package grid_remap and its subpackages (core, analysis, and tools). You need to have a basic Python environment set up for scientific computing, including numpy, scipy, traits, matplotlib, and PIL for image output. Ideally, you should run the code interactively at an IPython command prompt.

The easiest way to set up a complete Python environment is to install the Enthought Python Distribution (EPD). The model code was developed and tested on EPD v6.2. The latest version of the EPD can be found here:

It is available for all major operating systems. The grid_remap software should also run on all major platforms, but any platform-specific bugs or errors can be emailed to me at

If you don't want to install the entire EPD, the most critical libraries needed can be downloaded and installed from their respective websites:

This can be done, but it really is a pain. The EPD is a big download, but it Just Works(TM).

Software Installation

Download the zip-file archive from ModelDB. Then you need to extract the archive and run the file from a terminal:

$ tar xzf grid_remap.tar.gz
$ cd grid_remap/
$ sudo python install

This will ask for your administrator password and install the grid_remap package at the standard location on your computer for third-party python libraries.

Now, you can test to make sure everything is working properly by importing the package in IPython:

$ ipython -pylab
In [0]: from grid_remap import *

Starting IPython in the -pylab mode will enable a Matlab(R)-like environment with figure, plotting, and basic analysis commands available. The first time you do this import, the software will ask you to set up a directory where analysis output will be saved. It can be any directory that you will have full access to. When the import is finished, you can run the whos command in IPython to see the simulation and analysis classes that have been loaded (most of the classes needed are loaded by default).