VASPKIT

Overview

VASPKIT: Aims at providing a powerful and user-friendly interface to perform high throughput analysis of various material properties from the raw calculated data using the widely-used VASP code. The program can be conveniently run under either the interactive user interface or command line mode.

Using

Use the module name vaspkit to discover versions available and to load the application.

The VASPKIT, comes with a configuration file that needs to be copied into your home folder:

cd ~
cp /apps/u/opt/linux/vaspkit/<version>/how_to_set_environment_variables .vaspkit

You’ll notice within the .vaspkit file that the VASPKIT_UTILITIES_PATH and PYTHON_BIN settings are already appropriately set to the version being used.

The PYTHON_BIN variable points to a python conda environment hat contains the various package dependencies required.

You will need to edit the .vaspkit file according to your needs. For example, which version of VASP are you using, and folder locations for LDA, PBE and GGA paths:

VASP5                         .TRUE.                         # .TRUE. or .FALSE.; Set .FALSE. if you are using vasp.4.x
LDA_PATH                      ~/POTCAR/LDA                   #  Path of LDA potential.
PBE_PATH                      ~/POTCAR/PBE                   #  Path of PBE potential.
GGA_PATH                      ~/POTCAR/GGA                   #  Path of PW91 potential.
...

There is also some details on using PBS to submit jobs. Remember that we use Slurm, and these details will need to be updated accordingly.

Since using VASPKIT is out of our domain knowledge, we are happy to work with a user on how best to achieve this.

Multicore

At this stage, it appears that the vaspkit application can, itself, submit multi node jobs. We are happy to work with a user to better understand how this occurs, and update this section accordingly.