~nickpapior/siesta/trunk-kpoint-dos : changes

~nickpapior/siesta/trunk-kpoint-dos » Changes from revision 710

From Revision 710 to 691

Rev	Summary	Authors	Date
710	Added Alberto's suggestions for the documentation Added Alberto's suggestions for the documentation I have moved the note into the DOS.kgrid specification because I think it will be more highlighted.	Nick Papior	5 years ago
709	Merged trunk-708 Merged trunk-708	Nick Papior	5 years ago
708	Enabled LDOS and PDOS energy specification with respect to E... Enabled LDOS and PDOS energy specification with respect to EF Converted projected_DOS to F90. Documented the new DOS.kgrid.* and L/PDOS.kgrid.* input options. Now LDOS and PDOS may be prefixed with EF to specify that the energies are with respect to the Fermi level. This may help users who want to use the same input for various systems.	Nick Papior	5 years ago
707	Enabled an LDOS kgrid instead of the SCF one Enabled an LDOS kgrid instead of the SCF one This enables one to fine-tune the LDOS calculation although it should be less sensitive it may be useful. This commit also changes the following: 1) All kpoints_* are renamed to kpoint_* for clarity. 2) The LDOS and PDOS k-points are defaulted to the DOS k-points and thus have a common reference. This means that one can specify: DOS.kgrid.Cutoff which will act as k-grid for LDOS and PDOS calculations. 3) reset_kpoint* routines have been added which cleans the memory upon siesta exit. 4) Changed local_DOS.F to local_DOS.F90	Nick Papior	5 years ago
706	Merged r927-932, ion table precision, transiesta fixes and F... Merged r927-932, ion table precision, transiesta fixes and FA/FC precision	Nick Papior	5 years ago
705	Enabled kgrid.MonkhorstPack as list input (only if block is ... Enabled kgrid.MonkhorstPack as list input (only if block is not present) When the MonkhorstPack block is not present the kgrid can be read from the kgrid.MonkhorstPack list, i.e. kgrid.MonkhorstPack [10 10 10] which will create a 10 x 10 x 10 MP grid.	Nick Papior	5 years ago
704	Merged 4.1, r926, fix for overfsm conversion Merged 4.1, r926, fix for overfsm conversion	Alberto Garcia	5 years ago
703	Improve and document the handling of the SOC SR-SO split. Ne... Improve and document the handling of the SOC SR-SO split. New ref. outputs The splitting procedure for the off-site flavor of SOC (involving full lj projectors) can sometimes be ill-defined, and in those cases the program relies on a heuristic to compute the two contributions. A (single) warning is printed (this is a fix from a previous patch). The use of soc-split-sr-so F will prevent the program from attempting the splitting (but it still will be able to detect a possible problem and report an informational message). The reference outputs for the 'SOC' tests had been generated by a non-tracked version during merging, and showed irreproducible differences with other versions. They have been re-generated. (+ some easy optimizations in nlefsm_SO_off)	Alberto Garcia	5 years ago
702	Merged r924-925, overfsm removes S, wave-io and iosockets fi... Merged r924-925, overfsm removes S, wave-io and iosockets fixes	Nick Papior	5 years ago
701	Merged r922-923, pivoting errors in electrodes and numberofe... Merged r922-923, pivoting errors in electrodes and numberofeigenstates	Nick Papior	5 years ago
700	Allow the option of not separating SR and SO in SOC Allow the option of not separating SR and SO in SOC In a SO calculation, the use of the fdf line: soc-split-sr-so F will prevent the program from splitting the lj NL contributions to the Hamiltonian and energies into SR (ion) and SO parts. The splitting procedure for the 'offsite' flavor of SOC can be ill-defined. For the 'onsite' flavor of SOC this problem does not appear, but the option is also available for generality. When the SO contribution is not split, the Enl energy is tagged 'Enl(+so)' and the Eso energy is tagged 'Eso(nil)' in the output file. The CML file is not thus changed (but there is a new parameter 'Split-SR-SO'). Note that this is only a cosmetic change affecting the reporting of some components of the energy. Internally, the change affects the H_vkb and H_so components of the Hamiltonian, which are currently only used to compute the Enl and Eso energies.	Alberto Garcia	5 years ago
699	Merged 4.1, r920-921: YAML infrastructure; TS tests fixes Merged 4.1, r920-921: YAML infrastructure; TS tests fixes	Alberto Garcia	5 years ago
698	Restructured k-point sampling Restructured k-point sampling Now k-points may be user-supplied for further finetuning. This may be extremely useful in various situations: 1) Detailed analysis on PDOS for spin-orbit calculations 2) Faster convergence with fewer k-points 3) Improved convergence for bias calculations by adding zoom-in k-points in the bias-window. 4) And others. Now the kpoints are associated with a type: kpoint_t this will allow one to read k-points using kgrid.File flag. The precedence is much like the older implementation: 1) kgrid.MonkhorstPack 2) kgrid.Cutoff 3) kgrid.File in that order. If none are specified, then the Gamma-point will be used. This meant that several places the code needed updating. - Projected DOS k-points - TranSiesta k-points - SCF k-points Now all references to the kpoint_grid module is using kpoint_scf_m module which uses the type. This not only reduces lots of duplicated code, but it also allows one implementation to benefit across all code segments. Also added a test that reads in the k-points from a file (si_coop_kp_file).	Nick Papior	6 years ago
697	Merged 4.1, r919: io_close placement Merged 4.1, r919: io_close placement	Alberto Garcia	6 years ago
696	Merged 4.1, r910-918 Merged 4.1, r910-918	Alberto Garcia	6 years ago
695	Merged r905-909, small fixes here and there, improved print-... Merged r905-909, small fixes here and there, improved print-out for ts	Nick Papior	6 years ago
694	Merged r903, pivoting fixes, DM in tbtrans and custom energy... Merged r903, pivoting fixes, DM in tbtrans and custom energy-points	Nick Papior	6 years ago
693	Merge the 'offsite spin-orbit' implementation by Ramon Cuadr... Merge the 'offsite spin-orbit' implementation by Ramon Cuadrado. Reference: R. Cuadrado and J. I. Cerda, "Fully relativistic pseudopotential formalism under an atomic orbital basis: spin–orbit splittings and magnetic anisotropies", J. Phys.: Condens. Matter 24, 086005, (2012) (DOI:10.1088/0953-8984/24/8/086005) The new 'offsite' implementation is now the default when Spin { SO, SOC, S+O } SpinOrbit T (deprecated) is specified in the input fdf file. To request the 'onsite' approximation, use Spin { SO+onsite, SOC+onsite, S+O+onsite} In this 'offsite' implementation the introduction of a fully relativistic Hamiltonian is done by using fully non-local operators for the pseudopotentials. In this way it is possible to avoid the 'onsite' approximation needed to reduce the computational effort required when explicitly computing the LS term. The construction of a fully relativistic Hamiltonian involves the use of a full set of lj KB projectors. These are constructed by new code in 'atom', and processed in the new routine 'nlefsm_SO_off', which has roughly the same structure as 'nlefsm', but constructs at the same time the 'ion' and 'SO' pieces from the relativistic projectors. This routine calls 'calc_Vj_offsiteSO', where VSO and Vion and the corresponding forces are computed using the Clebsch–Gordan coefficients needed to change from the basis \|l,m,sigma> to \|j,mj>. The conventions for structure and signs in H and the DM are the same as in the existing 'onsite' implementation, so there are no changes in the diagonalization routines, or in the analysis routines and tools. Eventually, the 'offsite' qualifier might be removed, as this is a full spin-orbit implementation which involves a similar computational effort using less drastic approximations. Note that this merge focuses on the core electronic-structure functionality of full spin-orbit coupling, and does not provide any spin-orbit enhancements to the analysis tools.	Ramon Cuadrado	6 years ago
692	Merged 4.1, r903: update imports in pexsi_local_dos Merged 4.1, r903: update imports in pexsi_local_dos	Alberto Garcia	6 years ago
691	Merged 4.1, r902: catch charge inconsistencies in basis stat... Merged 4.1, r902: catch charge inconsistencies in basis states	Alberto Garcia	6 years ago

Older »