AiiDA plugin for the OpenMX DFT code.
pip install aiida-openmx
To simplify the use of pseudopotentials and orbital bases in aiida-openmx
, aiida-pseudo
and aiida-basis
have support for OpenMX's VPS pseudopotential and PAO pseudoatomic orbital formats respectively.
For high-throughput studies, it is additionally useful to create PseudoPotentialFamily
and BasisSet
AiiDA Group
s, which hold many pseudopotentials and orbital bases as well as various metadata.
As an example, lets create a PseudoPotentialFamily
for OpenMX's soft PBE19 pseudopotentials and a BasisSet
for OpenMX's soft 'standard' recommendation.
First, we collect the appropriate .vps
pseudopotentials in a directory:
export OPENMX_SOURCE=$HOME/openmx3.9/ # Path to the extracted OpenMX source code on your computer
mkdir PBE_19_soft # Directory to hold the .vps files
cp ${OPENMX_SOURCE}/DFT_DATA19/VPS/*PBE19.vps ./PBE_19_soft/ # Copy non-hard/soft .vps files
cp ${OPENMX_SOURCE}/DFT_DATA19/VPS/*PBE19S.vps ./PBE_19_soft/ # Copy soft version of hard/soft .vps files
Next, we'll create a PseudoPotentialFamily
which contains all of the pseudopotentials in our directory.
First, run python
, ipython
, or verdi shell
.
Then, run something like the following code:
from aiida_pseudo.groups.family import PseudoPotentialFamily
from aiida_pseudo.data.pseudo import VpsData
PBE_19_soft_family = PseudoPotentialFamily(
dirpath='./PBE_19_soft/', # Directory where we copied our .vps files
label='OpenMX/19/PBE/S', # Name of the PseudoPotentialFamily in the AiiDA database
description='OpenMX PBE19 (soft)' # Longer description of what the PseudoPotentialFamily contains
pseudo_type=VpsData # AiiDA data type for VPS pseudopotentials
)
We can then check that the PseudoPotentialFamily
is listed in our AiiDA Groups
:
verdi group list -a
PK Label Type string User
---- ----------------- ------------- ---------------------
1 OpenMX/19/PBE/S pseudo.family austin.zadoks@epfl.ch
Great! Now we can set up our pseudoatomic orbitals in a similar manner.
Because it's a bit trickier to collect all the appropriate files in this case, aiida-openmx
provides some PAO tables based on OpenMX's recommendations.
export OPENMX_SOURCE=$HOME/openmx3.9/
export AIIDA_OPENMX=$(python -c "import pathlib; import aiida_openmx; print(pathlib.Path(aiida_openmx.__file__).parent)")
mkdir standard_19_soft
for paofile in $(cat $AIIDA_OPENMX/data/pao/standard_s.txt); do cp $OPENMX_SOURCE/DFT_DATA19/PAO/$paofile ./standard_19_soft/; done
Finally, we'll create a BasisSet
which contains all of the pseudoatomic orbitals in our directory.
First, run python
, ipython
, or verdi shell
.
Then, run something like the following code:
from aiida_basis.groups.set import BasisSet
from aiida_basis.data.basis import PaoData
standard_19_soft_set = BasisSet(
dirpath='./standard_19_soft/', # Directory where we copied our .pao files
label='OpenMX/19/standard/S', # Name of the BasisSet in the AiiDA database
description='OpenMX standard 19 (soft)' # Longer description of what the BasisSet contains
basis_type=PaoData # AiiDA data type for PAO bases
)
We can then check that the BasisSet
is listed in our AiiDA Groups
:
verdi group list -a
PK Label Type string User
---- -------------------- ------------- ---------------------
1 OpenMX/19/PBE/S pseudo.family austin.zadoks@epfl.ch
2 OpenMX/19/standard/S basis.set austin.zadoks@epfl.ch
If you would like to contribute to the development of aiida-openmx
, please clone this Github repository and install aiida-openmx
as follows:
git clone https://github.com/zooks97/aiida-openmx .
cd aiida-openmx
pip install -e .[pre-commit,testing,docs] # Install extra dependencies
pre-commit install # Install pre-commit hooks
pytest -v # Discover and run all tests
See the developer guide for more information.
MIT