~nickpapior/siesta/low-D-stress : changes

~nickpapior/siesta/low-D-stress » Changes from revision 1072

From Revision 1049 to 1030

Rev	Summary	Authors	Tags	Date
1049	Updated documentation for bond-currents in tbtrans Updated documentation for bond-currents in tbtrans	Nick Papior		5 years ago
1048	Fix typo in molecularmechanics Fix typo in molecularmechanics A variable was mistyped in the C10 sections.	Alberto Garcia		5 years ago
1047	Updated documentation for linking MKL with GNU compilers Updated documentation for linking MKL with GNU compilers	Nick Papior		5 years ago
1046	Fixed nullification of Elecs(:)%Sigma variables Fixed nullification of Elecs(:)%Sigma variables They are only used in the calculation subroutines but adding warnings to flags will show them as "possibly" being misused.	Nick Papior		5 years ago
1045	Updated contributions documentation Updated contributions documentation	Nick Papior		5 years ago
1044	Fixed realloc-lhs for some low-level routines Fixed realloc-lhs for some low-level routines In particular the interpolation.f90 code has been revised with two major changes: 1. The codes are now skipping LHS allocation checks 2. Calling spline did a) spline calculation, b) storing spline intermediates to derived type, c) copied result from type Now it does not store the intermediate data since it is not used/needed. The result is copied directly in the method. I tested this on a test and results are the same. Other smaller modules now also tries to circumvent the reallocation statements.	Nick Papior		5 years ago
1043	Added contribution documentation Added contribution documentation A simple contribution document has been created.	Nick Papior		5 years ago
1042	Ensured transiesta crashes for non-collinear and spin-orbit Ensured transiesta crashes for non-collinear and spin-orbit	Nick Papior		5 years ago
1041	Made the right-part of the equilibrium contour the default Made the right-part of the equilibrium contour the default Indications are that it is better.	Nick Papior		5 years ago
1040	Updated documentation to reflect the device region specifica... Updated documentation to reflect the device region specification Also allowed not-atom specification for buffer atoms.	Nick Papior		5 years ago
1039	Removed unnecessary memory in tbtrans Removed unnecessary memory in tbtrans There are certain lists that may easily be removed and thus reducing memory requirements. This is in particular lists of indices that refer to electrode regions and downfolding regions. Now, I have removed some of them which is only minimally affecting the memory requirement, none theless they have some effect.	Nick Papior		5 years ago
1038	Fixed region writeout in tbtrans Fixed region writeout in tbtrans This was merely a cosmetic thing were the number of orbitals was wrong in the std-out.	Nick Papior		5 years ago
1037	Converted CG comments to optimization comments, and added de... Converted CG comments to optimization comments, and added description of "errors" in TS There are no code changes. Comments in the code are changed from CG to "optimization" since Broyden or FIRE are also enabled. TranSiesta discussions on the extra errors printed out during the SCF for bias calculations are now discussed thoroughly. Also reverted the documentation details when people compile it themselves.	Nick Papior		5 years ago
1036	Updated release.sh script to add version/title to the .tex ... Updated release.sh script to add version/title to the .tex files Since we have added PDF-metadata we need a bit more information in the files.	Nick Papior	v4.1-b4	5 years ago
1035	Updated installation script for NetCDF-4 support (library up... Updated installation script for NetCDF-4 support (library updates)	Nick Papior		5 years ago
1034	Updated all test output and added the ones that were missing Updated all test output and added the ones that were missing Since this run also has the new NRTAB implementation nearly all things change (to the X.XY Y digit in energies). Also amended the RELEASE_NOTES	Nick Papior		5 years ago
1033	Added TS tests for the run_tests.sh script, also fixed a cp ... Added TS tests for the run_tests.sh script, also fixed a cp statement	Nick Papior		5 years ago
1032	Added 3 tests such that we can compare with DirectPhi calcul... Added 3 tests such that we can compare with DirectPhi calculations For now we have never had tests using DirectPhi, I have now added 2 tests (and 1 without DirectPhi) such that we can compare: The si111-spinpol test has been split into 4 test cases: - si111 (no polarization, no direct-phi) - si111-directphi (no polarization, direct-phi) - si111-spinpol (polarization, no direct-phi) - si111-spinpol-directphi (polarization, direct-phi) Running the tests for the 4 systems gives the same output (by symmetry).	Nick Papior		5 years ago
1031	Final fix for OMM with high debug flags Final fix for OMM with high debug flags Now the psi array is minimally 1 element big such that we can pass arrays as dummy-arguments.	Nick Papior		5 years ago
1030	Fix OMM reuse of the psi array from 'densematrix' module Fix OMM reuse of the psi array from 'densematrix' module The OMM method can use eigenvectors from diagonalization as seeds. In that case the first few nominal OMM iterations are actually calls to 'diagon'. In 'compute_dm', 'diagon' is now passed an extra (optional) argument to signal the need to keep the eigenvectors. The array 'psi' will not be deallocated in diagon, but after the first true OMM step. (Thanks to Nick Papior for finding the bug)	Alberto Garcia		5 years ago