| TRUSTE | ||
|---|---|---|
|
Appendix A. Keyword Reference |  |
| TRUSTE | ||
|---|---|---|
|
|
Appendix A. Keyword Reference | |
TRUSTE — Default method for geometry optimization using trust radii.
TRUSTE
TRUSTE=LARGE
Minimize energy using the TRUST algorithm. (This method is now AMPAC’s default approach.)
The introduction of SPARSE and PSOLVE allows AMPAC to compute energies and properties of molecules considerably larger than in previous versions. This keyword represents a variant of TRUSTE intended to extend the effectiveness of this method to these very large structures. It differs from regular TRUSTE in the following ways. (1) The standard 2-body initial guess Hessian is replaced by Lindh's approximate Hessian (LINDH). (2) The initial Hessian is not diagonalized (to reduce CPU time) although this results in a small degridation of optimzation permformce. (3) The trust radius step is solved for iteratively (thus avoiding a costly diagonalization).
 |
Copyright © 1992-2013 Semichem, Inc. All rights reserved. |
|
|
 |
|
| TRUES |  |
TRUSTG |