Set up your environment for diffusion MRI processing

Create a user directory

 mkdir -p /research/graphics/users/$USER/sandbox

Set up your environment

Edit ~/.environment and add the following lines:

# in the umask section
umask 002

# anywhere; this loads the licenses for the NAG numerics library
# change to profile.sh for bash
source /com/nag/profile.csh

# in the environment variable section
setenvvar       CVSROOT            /research/graphics/cvsroot/
setenvvar       G                  /research/graphics/users/$USER/sandbox
setenvvar       GTO                /research/graphics/users/$USER/sandbox
setenvvar       GARCH              linux
setenvvar       G_COMPILER         gcc4
setenvvar       G_IMPORT           /research/graphics/tools/linux/import

# in the path section
pathappend       PATH       "$G/bin"
pathappend       PATH       "/research/graphics/tools/linux/bin"
setenvifnot      PYTHONPATH "$G/bin"
pathappend       PYTHONPATH "$G/bin"

Then execute: source ~/.environment

If you're working on your own machine outside the CS department, you may need to set $GARCH to another value, OSX or WIN32. Since you won't have direct access to the CS filesystem, you should remove the $G_IMPORT line. You will also have to manually administrate all your own system's tools (for example, Diffusion Toolkit and FSL) and make sure that they're on your $PATH; within the department, $G_IMPORT takes care of this for you.

Check out, build, and install common code

mkdir $G/install_linux
cd $G && cvs co common/build common/utility common/gg common/mri common/libcurvecollection
cd $G/common/utility/nag && make install
cd $G/common/utility/port && make all && make install
cd $G/common/gg && make all && make install
cd $G/common/mri && make all && make install

Check out, build, and install pipeline code

cd $G && cvs co project/brain/pipeline
cd $G/project/brain/pipeline && make all && make install

Put everything on your $PATH

cd $G
mkdir -p bin
common/build/kit/linkbins.py $G/install_linux/bin bin