Skip to content
Density-functional toolkit
Julia Shell
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.
.github/workflows Add runexamples action Feb 2, 2020
deps Version check for spglib Feb 13, 2020
docs Add latex documentation for symmetry Feb 13, 2020
examples Fix examples Feb 7, 2020
.bumpversion.cfg Bump version: 0.0.3 → 0.0.4 Feb 4, 2020
.travis.yml Refactor travis file Feb 2, 2020
LICENSE Initial commit May 2, 2019
Project.toml Run ABINIT directly from DFTK datastructures Feb 7, 2020 Amend README Feb 4, 2020

dftk logo

Density-functional toolkit

license Build Status on Linux Coverage Status DOI

The density-functional toolkit, or short DFTK is a library of Julia routines for experimentation with plane-wave-based density-functional theory (DFT), as implemented in much larger production codes such as Abinit, Quantum Espresso and VASP. The main aim at the moment is to provide a platform to facilitate numerical analysis of algorithms and techniques related to DFT. For this we want to leverage as much of the existing developments in plane-wave DFT and the related ecosystems of Julia python or C codes as possible.


The library is at an early stage and the supported feature set is still limited. An overview:

  • Methods and models:
    • Periodic Hamiltonians, such as reduced Hartree-Fock, Gross-Pitaevskii, density-functional theory, analytic potentials
    • All LDA and GGA functionals from libxc supported.
    • Godecker pseudopotentials (GTH, HGH)
    • Exploit Brillouin zone symmetry for k-Point sampling
    • Fermi-Dirac or Methfessel-Paxton smearing to treat metals
    • Self-consistent field approaches: Damping, Kerker mixing, Anderson mixing (Pulay DIIS), all solvers from NLsolve.jl
    • Direct minimization
  • Ground-state properties and post-processing:
    • Total energy
    • Forces
    • Density of states (DOS) and local density of states (LDOS)
    • Band structures
    • Full access to all intermediate quantities (e.g. density, Bloch wave)
  • Support for arbitrary floating point types, including Float32 (single precision) or Double64 (from DoubleFloats.jl). For DFT this is currently restricted to LDA (with Slater exchange and VWN correlation).

Note: DFTK has only been compared against standard packages for a small number of test cases and might still contain bugs.


The package is not yet registered in the General registry of Julia. Instead you can obtain it from the MolSim registry, which contains a bunch of packages related to performing molecular simulations in Julia. Note that at least Julia 1.2 is required.

First add MolSim to your installed registries. For this use

] registry add

for a Julia command line. Afterwards you can install DFTK like any other package in Julia:

] add DFTK

or if you like the bleeding edge:

] add DFTK#master

Some parts of the code require a working Python installation with the libraries scipy, pymatgen and spglib. The examples require matplotlib as well. Check out which version of python is used by the PyCall.jl package. You can do this for example with the Julia commands

using PyCall

Then use the corresponding package manager (usually apt, pip or conda) to install aforementioned libraries, for example

pip install scipy spglib matplotlib pymatgen


conda install -c conda-forge scipy spglib matplotlib pymatgen

You can then run the code in the examples/ directory.


Despite the current focus on numerics, the intention is to keep the project rather general, so that this platform is useful for general research in materials science.




Feel free to contact us (@mfherbst and @antoine-levitt) directly, open issues or submit pull requests. Any contribution or discussion is welcome!

You can’t perform that action at this time.