Data for: Ab initio machine-learning unveils strong anharmonicity in non-Arrhenius self-diffusion of tungsten

The dataset contains key files to reproduce the results presented in the article " Ab initio machine-learning unveils strong anharmonicity in non-Arrhenius self-diffusion of tungsten":<br> <ul> <li>DFT input files: INCAR, KPOINTS.</li> <li>All POSCAR files for DFT and thermodynamic integration</li> <li>Moment tensor potential (MTP) file</li> <li>Training dataset for MTP</li> <li>All Hessian Matrix files for thermodynamic integration. For the transition state, the stabilized Hessian Matrix is provided.</li> <li>All imaginary mode files for transition state</li> <li>Lattice expansion used in all calculations.</li> </ul>