Vasputil

Vasputil is a set of command-line utilities and Python libraries designed to make life with VASP easier. The command-line utilities (Python scripts), can be used directly or as examples of what can be done with the provided python modules. These are

vasputil_atomsdistance

Measure distance between two atoms, can also measure projected distances.

vasputil_atomsmoved

See which atoms have moved between two POSCAR files. If lattices are compatible, takes into account periodic boundary conditions. Optionally print out moved distances and the distance vector.

vasputil_chgarith

Simple arithmetic on charge density in CHG/CHGCAR format files. Supports +,-,*,/ for the elemental arithmetic operations and 'avg' for calculating the average.

vasputil_chgplaneplot

Plot charge density in a plane (generated e.g. by lev00) as a pseudocolor, contour (colour or BW), or filled contour plot. Output is either to a window on the screen or to a file.

vasputil_direct2cartesian

Convert a POSCAR file from direct to cartesian coordinates.

vasputil_dosplot

Plot density-of-states. Can plot total, total integrated, and site-projected DOS. For site-projected DOS, different orbitals can be plotted.

vasputil_interpolate

Interpolate coordinates between two POSCAR files, either one intermediate image where one can specify where in the interval the new coordinates will be, or create many evenly spaced images.

vasputil_nearestneighbors

Print out nearest neighbor table, based on minimum distance or number of nearest neighbors.

vasputil_plane2atom

Calculate the distance between an atom and a plane defined by three atoms.

vasputil_poscar2xyz

Convert a POSCAR format file to XYZ format.

vasputil_zlayers

Find layers in the z direction and print interlayer distances.

Documentation for these utilities is provided via the -h command line option.

Web Page

https://github.com/jabl/vasputil

Main home page; download, source tree, issue tracker, etc.

https://users.aalto.fi/~jblomqvi/vasputil

Older releases, and a copy of the README page.

Requirements

For full functionality, vasputil requires the following python libraries:

1. ASE (Atomic Simulation Environment) version 3.4 or newer. For older ASE releases, vasputil 5.3 is compatible with ASE 3.3.1, vasputil 5.2 is compatible with ASE 3.2, vasputil 5.1 with ASE 3.1, and vasputil 5.0 with ASE 3.0.
2. NumPy
3. matplotlib
4. SciPy

Finally, to regenerate the README.html file, the rst2html utility is needed.

Without all the libraries, a subset of the functionality is available. The matplotlib library is needed only by the vasputil_chgplaneplot, vasputil_dosplot and vasputil_zlayers utilities. The SciPy library is needed only by vasputil_chgplaneplot, and only in case the figure shape is not square. The remaining utilities require only ASE and NumPy.

Except for ASE, the other three libraries (and their dependencies) should already be installed on your system. If not, at least on Linux they should be available from the default package repositories, and installation should be a breeze. If you don't have access to a package management system (e.g. on Linux) which includes matplotlib and its dependencies, you can get all at once by installing one of the python distributions targeted at science, such as SAGE, Enthought Python Distribution, or Python(x,y). Another solution if you already have python installed, is to install matplotlib and other libraries via easy_install from setuptools (a sort of package management system for python packages).

Installation Instructions

1. Unpack the program distribution and go to the directory.

   tar xvzf vasputil-$(VERSION).tar.gz

   or unzip vasputil-$(VERSION).zip

   cd vasputil-$(VERSION)

2. Install the program into the standard python user site directory

   python setup.py install --user

   Make sure that the user site binary directory is in your PATH. On Linux, the default directory is ~/.local/bin, for more information see PEP 370 .

   In case your python version is older than 2.6, you can instead install with the --home or --prefix options, e.g.

   python setup.py install --home=~/.local

   This will install the scripts into ~/.local/bin and the python modules into ~/.local/lib/python/vasputil. You should make sure that the directory of the python modules is found on the python search path. E.g. if you installed it like the above, you should have

   export PYTHONPATH=$HOME/.local/lib/python

   in your shell initialization script (~/.bash_profile or equivalent). For installation into some other directory, see python setup.py --help and the documentation for distutils .

   For SUSE Linux users: There is apparently something weird with the python installation in SUSE, and the --home option does not work properly. Instead install vasputil with

   python setup.py install --prefix=~/.local

   In this case the library modules will instead be installed in ~/.local/lib{64}/python{X.Y}/site-packages/vasputil.

3. Optionally, run the testsuite to make sure it works.

   python runtests.py

Interactive Usage

The python modules are designed to be used from the interactive python prompt as well, to provide a sort of MATLAB-like interactive environment where you can manipulate coordinates stored in arrays, or plot data etc. For this, the IPython enhanced interactive environment is recommended. Start with the matplotlib stuff preloaded with ipython -pylab. For those familiar with MATLAB, see NumPy for MATLAB Users . After that, just import what you need from the vasputil module and get going! See the python scripts for reference, or pydoc vasputil, pydoc vasputil.supercell etc. to read documentation, or the ? operator in IPython.

With IPython you can also run the scripts in the interactive environment with the %run command, and this will import the variables in the script into the ipython environment. This is useful if you want to do something similar, but not quite what the command-line interface allows.

Manual

This section describes how to accomplish specific tasks. It does not document every feature of vasputil, as most of the tasks are hopefully simple enough that the usage should be self-evident from the help instruction given by the -h option to the command-line utilities.

Plotting charge density in a plane

Use the vasputil_chgarith tool to create a suitable input charge density file (e.g. the charge density difference when some species adsorbs on a surface), and then use the lev00 utility to create the charge density data in a plane. The data is written to a file (default name out.dat_1) that can be read with the vasputil_chgplaneplot utility which can then plot this data.

Generating supercells

vasputil 5.x does not contain the supercell generator previously found in vasputil 4.x. To create supercells, the ASE supercell generator is recommended. This contains functionality to create surface supercells as well. In order to generate a supercell for VASP, first create the supercell following the ASE lattice instructions. This is probably easiest done interactively using IPython. Assuming you have imported ase as from ase import * and your supercell is referenced via the variable atoms, you can visualize your results with view(atoms). Finally, write out the supercell to the file POSCAR.out with write('POSCAR.out', atoms, format='vasp').

Another option is the tetr utility, which also contains a supercell generator.

Plotting projected Density-of-States

The vasputil_dosplot contains a simple command-line utility for plotting a single orbital. While this is nice for quickly looking at the DOS, for publication plots you probably want to look at that utility and create custom versions of it in order to create the specific plots you want, with multiple subplots and multiple graphs per plot.