DFTTK: Density Functional Theory Tool Kit

Ultimate goals: For a given structure and elements, calculate the free energy with respect to possible internal degree of freedoms.

Features

High-throughput. It can run plenty of structures with one simple command.

Simple. Only the structure file is required.

The following workflows are currently implemented:

Gibbs energy workflow for stable structures

Minimum volume finding workflow

Note: This repo contains the custom workflows developed by the Phases Research Lab that do not fit into the scope of the public atomate repository.

Installation

DFTTK requires Python 3. Python 2 support for NumPy ends 2019-01-01.

Create virtual environment (optional)

Anaconda or Miniconda is required. (Another option is using virtualenv)

#conda create -n ENV_NAME python=VERSION
conda create -n dfttk python=3.6
#Activate
conda activate dfttk
#Deactivate
conda deactivate

Install dfttk

To install dfttk, there are several ways.

From the command line, run pip install dfttk

conda

Anaconda packages of DFTTK are currently not supported. If you are using Anaconda, you should be able to install with pip.

development versions

git clone https://github.com/phasesresearchlab/dfttk
cd dfttk
pip install -e .

For more details, ref. Install_link

Configuration

Preparation

Prepare following files.

current_folder
├── psp                          [specified by -psp]
│   ├── pseudopotential_content  [required if you didnot configurate pymatgen]
│   └── ...
├── config                       [specified by -c parameter]
│   ├── db.json                  [required]
│   ├── my_launchpad.yaml        [required]
│   ├── FW_config.yaml           [optional]
│   ├── my_fworker.yaml          [optional]
│   └── my_qadapter.yaml         [optional]
└── vaspjob.pbs                  [optional, specified by -q parameter]

Config

After prepared above files, simply run

dfttk config -all

PRL GROUP NOTE: If you use ACI cluster at PRL group, you can get the pseudopotentials from ACI

dfttk config -all -aci

For more details, ref. Configuration

Use

Prepare structure file(s), e.g. POSCAR
Simply run dfttk run -f POSCAR

For more details, ref. Get_Started

Contributing

See CONTRIBUTING.rst

License

DFTTK is MIT licensed. See LICENSE

hitliaomq / dfttk-old