LDIAG: Difference between revisions
No edit summary |
No edit summary |
||
| Line 12: | Line 12: | ||
Loewdin perturbation theory strictly conserves the orbital order, i.e. say the n.th orbital will remain stored in the n.th storage slot and only small rotations into that orbital can occur. For {{TAG|ALGO}} = Damped and All, the final sub space diagonalization is simply skipped if LDIAG=.FALSE. is set. Generally using {{TAG|LDIAG}} = .FALSE. is only advised, if one wants to maintain a certain orbital order, for instance, when reading the orbitals from an existing {{TAG|WAVECAR}} file. | Loewdin perturbation theory strictly conserves the orbital order, i.e. say the n.th orbital will remain stored in the n.th storage slot and only small rotations into that orbital can occur. For {{TAG|ALGO}} = Damped and All, the final sub space diagonalization is simply skipped if LDIAG=.FALSE. is set. Generally using {{TAG|LDIAG}} = .FALSE. is only advised, if one wants to maintain a certain orbital order, for instance, when reading the orbitals from an existing {{TAG|WAVECAR}} file. | ||
For the algorithms {{TAG|ALGO}}=Normal or Fast, by construction | For the algorithms {{TAG|ALGO}}=Normal or Fast, by construction it is not possible to switch off subspace diagonalization, as these algorithms require subspace diagonalizations during the iterative refinement of the orbitals. Furthermore, algorithms that minimize the total energy (ALGO = All) are often too "greedy" and tend to alternate the orbital order in the course of the scf cycle and energy optimization. | ||
In summary, the following combinations are potentially useful: | In summary, the following combinations are potentially useful: | ||
Revision as of 07:29, 10 November 2021
LDIAG = [logical]
Default: LDIAG = .TRUE.
Description: This tag determines whether a subspace diagonalization is performed or not within the main algorithm selected by IALGO.
For ALGO=Normal, Fast, and VeryFast, VASP performs a diagonalization in the subspace spanned by all orbitals. This is often referred to as Rayleigh–Ritz method (https://en.wikipedia.org/wiki/Rayleigh%E2%80%93Ritz_method). The step increases the convergence rate and thus is expedient in most cases. For the direct optimization algorithms (for instance ALGO = All or Damped), a subspace diagonalization is usually not performed, but in order to improve the accuracy of the calculated forces, after convergence has been reached one single diagonalization in the subspace spanned by all orbitals is performed. Note, that the subspace diagonalization sorts the orbital/eigenvalues in ascending order.
For ALGO = VeryFast and Damped it is possible to switch off the subspace diagonalization by specifying LDIAG=.FALSE. in the INCAR file. Specifically, for ALGO = VeryFast, LDIAG= .FALSE. changes from an exact Rayleigh–Ritz diagonalization to Loewdin perturbation theory. Loewdin perturbation theory strictly conserves the orbital order, i.e. say the n.th orbital will remain stored in the n.th storage slot and only small rotations into that orbital can occur. For ALGO = Damped and All, the final sub space diagonalization is simply skipped if LDIAG=.FALSE. is set. Generally using LDIAG = .FALSE. is only advised, if one wants to maintain a certain orbital order, for instance, when reading the orbitals from an existing WAVECAR file.
For the algorithms ALGO=Normal or Fast, by construction it is not possible to switch off subspace diagonalization, as these algorithms require subspace diagonalizations during the iterative refinement of the orbitals. Furthermore, algorithms that minimize the total energy (ALGO = All) are often too "greedy" and tend to alternate the orbital order in the course of the scf cycle and energy optimization.
In summary, the following combinations are potentially useful:
ALGO = VeryFast ; LDIAG = .FALSE
ALGO = Damped ; LDIAG = .FALSE
Other combinations using LDIAG = .FALSE. are likely to result in undesirable results. Finally, note that ALGO = VeryFast is not working for hybrid functionals.