|
| subroutine, public | orthogonalization_gbl::GS_ortho_routine (no_cf, p_s, p_e, a_s, a_e, vecs, ao_overlaps, symmetry, mo2so_range, del_thrs) |
| | Performs orthogonalization of several vectors (vecs(1:no_cf,a_s:a_e)) against the set of fixed vectors (vecs(1:no_cf,p_s:p_e)) which are assumed to be orthogonal with respect to the overlap matrix ao_overlaps. The number of AO coefficients is no_cf and the range of active vectors (columns of vecs) is given by p_s,p_e. The active and passive sets of vectors are assumed to be disjunct. mo2so_range gives the index of the first and the last AO which contributes to a given MO (given in the column index). Note that the use of this array ensures that symmetry adaptaion of the AO basis is handled with maximum efficiency. Note that we cannot use mo2so_range in the loops involving the orbitals to orthogonalize since coefficients of these orbitals change. However, these loops can be made more efficient if I supply the routine with the information on the range of the symmetry adapted basis set in each symmetry (once symmetry adaptation is implemented). More...
|
| |
| subroutine, public | orthogonalization_gbl::SYM_ortho_routine (no_cf, a_s, a_e, vecs, mo_overlaps, del_thr, del_thr_ao, to_delete) |
| | Performs symmetric orthogonalization of orbitals given in columns of vecs(:,:). The range of orbitals (columns) to orthogonalize is given by a_s, a_e. The overlap matrix in the basis of the orbitals being orthogonalized is given in mo_overlaps. The number of coefficients (rows of vecs) each orbital has is no_cf. The deletion threshold is given by del_thr. The logical array marking the orbitals for deletion is to_delete. More...
|
| |
