qe-tools

qe-tools is a collection of tools for interacting with Quantum Espresso. There are two types of tools so far:

• Tools which will run QE for you
• Tools to process QE outputs

List of tools

qe-test.py

This script runs validation tests. If the test is named TEST, it expects runs TEST/RUN with a qe-run input file TEST.RUN.in. A special run is called 'ref' and should also contain the outputs of qe-run from a trusted build. The runs are executed and outputs compared to the 'ref' run. The 'ref' run is also re-executed to validate the build. Here is an example.

This script looks for a json config file in the test directory, TEST/config, that stores the number of bands, the number of atoms, and error tolerances. The default config is:

{
  "nb":     opts.nb,
  "etot":   0.005,
  "efermi": 0.1,
  "force":  0.01,
  "stress": 1.,
  "bands":  0.01,
  "atoms": 1.
}

This script can also handle VASP. The structure is the same as for QE, but with the contents of TEST/RUN being the VASP input files, usually just INCAR, POSCAR, POTCAR, KPOINTS.

qe-run.py

This script parses and runs a special input file. It reads the file sequentially, adding or overwriting namelist options and keeping only the most recently read card of each type. When it hits a "@run" directive, it dumps state as a valid pwscf input file and runs pwscf. If there are multiple runs in a single file, their stdout/stderr are directed to sequentially numbered run.out files.

vasp-run.py

This script is a thin wrapper around the command:

mpirun -np N BINDIR/vasp 2> PREFIX.err | tee PREFIX.out

which pulls N from the option -n, the BINDIR from the option -e, and the PREFIX from the -p option, as in qe-run.py.

compare_bands.py

This script takes a reference and test band-structure file, and optionally band ranges and Fermi energies, computes the RMS error between the band-structures and plots with gnuplot.