Built from elements of the [Var]iational [Gl]acier [S]imulator (VarGlaS) codes:

https://github.com/douglas-brinkerhoff/VarGlaS

and the [U] of [M] [F]irn [D]ensification [M]odel (UM-FDM):

https://github.com/pf4d/um-fdm

utilizing the FEniCS project:

http://fenicsproject.org/

and Dolfin-Adjoint:

http://dolfin-adjoint-doc.readthedocs.org/en/latest/

FEniCS 1.6 :

sudo add-apt-repository ppa:fenics-packages/fenics;
sudo apt-get update;
sudo apt-get install fenics;
sudo apt-get dist-upgrade;

After this, install Dolfin-Adjoint as described here, but utilizing my cslvr-modified fork located here.

Latest Python packages and misc. dependencies:

sudo pip install shapely colored termcolor pyproj;

also Gmsh_dynamic and GmshPy from

http://geuz.org/gmsh/

NOTE: GMSH is required to be installed in order to use the mesh generation facilities located in meshing.py. Instructions on installing Gmsh_dynamic can be found on the Qssi wiki .

For a generic install (of gmsh-dynamic 2.10.1) to /usr/local, use

wget http://geuz.org/gmsh/bin/Linux/gmsh-svn-Linux64-dynamic.tgz;
tar -xzvf gmsh-svn-Linux64-dynamic.tgz;
cd gmsh-2.11.0-dynamic-svn-Linux;
sudo cp -r bin/ /usr/local/;
sudo cp -r include/ /usr/local/;
sudo cp -r lib/ /usr/local/;
sudo cp -r share/ /usr/local/;
cd gmshpy;
sudo python setup.py install;
sudo ldconfig;

And basemap:

http://matplotlib.org/basemap/users/installing.html

for a generic install to /usr/local/, use

wget http://sourceforge.net/projects/matplotlib/files/matplotlib-toolkits/basemap-1.0.7/basemap-1.0.7.tar.gz;
tar -xzvf basemap-1.0.7.tar.gz;
cd basemap-1.0.7/geos-3.3.3/;
export GEOS_DIR=/usr/local/;
./configure --prefix=$GEOS_DIR;
make;
sudo make install;
cd ..;
sudo python setup.py install;

Install the program by editing your .bashrc file with

export PYTHONPATH="<PATH TO CSLVR>:$PYTHONPATH"

Test your installation py entering in an ipython terminal

from cslvr import *

You may like to download some data and pre-made meshes. To do this, go into the scripts/ directory and run the download_*.py files there, dependending on your needs.

The data may be accessed via the DataFactory class, and the meshes via the MeshFactory class.