mirror of https://gitlab.com/QEF/q-e.git
![]() outlined in Macke et al., arXiv:2312.13580 (2023). Features: * Perform orbital-resolved DFT+U calculations in which a Hubbard U is applied only to certain, pre-selected eigenstates (magnetic quantum orbitals or spin-orbitals) of a given nl-subshell * Compute forces and stresses in orbital-resolved DFT+U with no limitations * Apply perturbations to the eigenstates in order to compute orbital-resolved Hubbard U parameters from first principles * Use the (recently introduced) HUBBARD card to provide the necesary input data for orbital-resolved DFT+U calculations in a user-friendly way * Updated documentation and two new examples (15 and 16) Add unit test for orbital-resolved DFT+U Small improvements for using orbital-resolved Hubbard ALPHA First test of orbital-resolved U on GPU First implementation of orbital-resolved DFT+U for SOC/NC Add documentation of NC orbital-resolved DFT+U Small fix in orbital-resolved DFT+U unit test More orbital-resolved ALPHA bugfixes Better output of perturbed occupations in DFT+U (orbital-resolved) Fix little typo Preparation for implementation Deleted warning messages, made orbital_resolved DFT+U lda_plus_u_type=0 instead of 3 (previously) |
||
---|---|---|
.. | ||
INPUT_OSCDFT.def | ||
INPUT_PW.def | ||
Makefile | ||
user_guide.tex |