AFLOW
AFLOW is a package written by collaborators at Duke University. It
will help automate the crucial quantum-mechanical calculations that
you will perform. The instructions for installing aflow can be found
here. Follow the instructions on slides 13 and 14 if
installing on MaryLou. Once you have it installed, you’ll need to
modify a file called aflow.rc
located in your root:
emacs ~/.aflow.rc
- Find the tag labeled
DEFAULT_VASP_BIN
and set it equal tovasp6_serial
- Find the tag labeled
DEFAULT_VASP_MPI_BIN
and set it equal tovasp6
- Find the tag labeled
DEFAULT_VASP_POTCAR_DIRECTORIES
and replace/somewhere/
with/fslhome/glh43/src/vasp
. - Save the file and exit.
aBuild
aBuild is a python package that I built to help automate file construction and data analysis that will be needed. When installing python packages, it is best to do it inside something called a virtual environment. You can install python packages inside of your virtual environment without disturbing the packages installed outside of the virtual environment.
Setting up your virtual environment
Make a directory called environments and go into that directory:
mkdir ~/environments
cd ~/environments
Create the virtual environment
virtualenv alloys
Enter the virtual environment:
workon alloys
You can exit the environment like this:
deactivate alloys
Installing aBuild
1.Register for an account on github.
2.Navigate to B. Nelson’s aBuild repository (found here)
3.Fork B. Nelson’s reposity. This will create a copy of this code
that is yours to edit.
4.[back on ML] Make a directory called codes
and go into it:
mkdir ~/codes
cd ~/codes
5.[Back on ML] Type the following to clone your aBuild repository to ML:
git clone https://github.com/<username>/aBuild.git .
(make sure you replace my username with yours.)
6.Install aBuild (while in your virtual environment)
pip install -e ~/codes/aBuild