To answer my own question, both Vasp and Wien2k use the Harris functional ( https://en.wikipedia.org/wiki/Harris_functional) which is not variational, bounded by (e.g. Doi 10.1088/0953-8984/1/41/007) (rho_in-rho_true)*(rho_true-rho_out) where rho_in is before the SCF mapping, rho_out is after (i.e. lapw2) and rho_true is the fixed-point solution. This compares to the KS functional which is variational and bounded by (rho_true-rho_out)**2. The literature indicates that the Harris functional energy is in general closer (for non-converged SCF) to the energy than the KS energy when the density is not converged. (This follows because rho_out is often quite bad.)
When the density is converged the KS functional and the Harris functional give the same energy. On Mon, Jan 31, 2022 at 8:23 AM Laurence Marks <laurence.ma...@gmail.com> wrote: > In Wien2k the total energy is calculated in part from the current density, > in part from the density of the orbitals that solve the KS equation. As > such it is not a true variational energy except when the density is > converged. [1] > > For other dft codes (not wavefunction codes [2]), is it the same, e.g. > Vasp, QE, ab-init? Responses welcome, either via the list of directly. > > [1] There are technically ways one could calculate some form of > variational energy, but it would require extra steps and to my knowledge > has never looked useful so is not in the code. > [2] In wavefunction codes where one varies the occupancy the energy is > typically variational I think. > > -- > Professor Laurence Marks > Department of Materials Science and Engineering > Northwestern University > www.numis.northwestern.edu > "Research is to see what everybody else has seen, and to think what nobody > else has thought" Albert Szent-Györgyi > -- Professor Laurence Marks Department of Materials Science and Engineering Northwestern University www.numis.northwestern.edu "Research is to see what everybody else has seen, and to think what nobody else has thought" Albert Szent-Györgyi
_______________________________________________ Wien mailing list Wien@zeus.theochem.tuwien.ac.at http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien SEARCH the MAILING-LIST at: http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html