pyfhiaims – an FHI-aims Python suite

For a more complete set of documentation see: https://fhi-aims-club.gitlab.io/pyfhiaims/

Installation

You can install pyfhiaims from PyPI via pip:

pip install pyfhiaims

Or, alternatively, you can install it from the main branch of this Git repository:

git clone https://gitlab.com/FHI-aims-club/pyfhiaims.git
cd pyfhiaims
pip install .

Usage

The main purpose of the package is to provide an interface to build input files for FHI-aims and to parse its outputs.

The input files can be created in the following way:

• geometry.in - build from ASE or pymatgen objects:

from ase.build import bulk
from pyfhiaims.geometry import AimsGeometry

cu = bulk('Cu', 'fcc', a=3.6)
aims_cu = AimsGeometry.from_atoms(cu)
aims_cu.write_file("geometry.in")

• control.in - build from Python dictionary:

import os
from pyfhiaims.control import AimsControl

# create the dictionary of input parameters
params = {"xc": "pbe",
          "species_dir": os.path.join(os.getenv("AIMS_SPECIES_DIR"), "light"),
          "k_grid": (7, 7, 7)
          }

# create the AimsControl object
control_in = AimsControl(parameters=params)
control_in.write_file(aims_cu)

The FHI-aims output file can be parsed in the following way:

from pyfhiaims import AimsStdout

stdout = AimsStdout("aims.out")

Then you can access all the parsed results (stdout.results), run metadata (runtime choices and some geometry statistics — stdout.metadata), warnings (stdout.warnings), and errors. Also, stdout.is_finished_ok tells if FHI-aims run has been finished without any errors.

There are several properties defined that make access to different widely used values easier, like energy and forces. Also, the top level keys of stdout.results dictionary can be accessed using dot notation (so run times can be accessed with stdout.final["time"]).

There are many values that are parsed from the aims.out file; you are welcome to explore stdout.results dictionary.

Support

Just write us an issue in the issue tracker!

Roadmap

To be written...

Contributing

Contributions are extremely welcome!

Authors and acknowledgment

The package was written by:

• Tom Purcell
• Andrei Sobolev

License

The project is licensed under MIT license.