β Scribed by Tai-Sung Lee; Weitao Yang
No coin nor oath required. For personal study only.
A frozen density matrix approach for determining local geometry changes of large molecules is proposed based on the density matrix divide-and-conquer method. This approach divides a large molecule into a small active part and the rest as Ε½ . the frozen part. After a single-point self-consistent field SCF calculation for the whole molecule, only the local molecular orbitals of the active part and its neighbor are updated when the geometry of the active part changes. The updated density matrix is calculated from the new local molecular orbitals of the active part and the stored local molecular orbitals of the frozen part. The electron transfer between the active and the frozen part is allowed via a common and variable chemical potential. The preliminary tests using semiempirical quantum chemical methods show this approach predicts the local geometry change well. The error in geometric parameters is less than 0.002 A and less than 0.5Π for bond length and bond angle, respectively.
π SIMILAR VOLUMES
## Abstract Efficient truncation criteria used in multiatom blocked sparse matrix operations for __ab initio__ calculations are proposed. As system size increases, so does the need to stay on top of errors and still achieve high performance. A variant of a blocked sparse matrix algebra to achieve s
The performance of linear-scaling electronic structure calculations depends critically on matrix sparsity. This article gives an overview of different strategies for removal of small matrix elements, with emphasis on schemes that allow for rigorous control of errors. In particular, a novel scheme is
Electronic structure calculations which are based on Wannier, like localized orbitals or the related density matrix, are an alternative to conventional calculations based on extended orbitals. For large systems this approach is potentially faster since it offers \(O(N)\) scaling with respect to the
The performance of conjugate gradient schemes for minimizing unconstrained energy functionals in the context of condensed matter electronic structure density functional calculations is studied. The unconstrained functionals allow a straightforward application of conjugate gradients by removing the e