[CPMD-list] WANNIER functions

[Juerg Hutter]

Hi

the short answer to your question is:
You shouldn't include empty states in the localization
(you ask for 10 States, 4 are occupied). In the CPMD
implementation you can only localize states with the
same ocuupation number.

This leads to more general problems. You have a mix of
options in your input file that don't make sense.
You ask for diagonalization and ODIIS optimisation,
you have empty states and you ask for all kind of
density mixings at the same time.
At least for a simple water molecule, the defaults
would work just fine.

As a general rule: CPMD input is not always predictable
if you combine many options that are exclusive.

Another point: we are at version 3.9.2, this includes
many improvements, also in the Wannier function code.

regards

Juerg Hutter


----------------------------------------------------------
Juerg Hutter                   Phone : ++41 44 635 4491
Physical Chemistry Institute   FAX   : ++41 44 635 6838
University of Zurich           E-mail: hutter at pci.unizh.ch
Winterthurerstrasse 190
CH-8057 Zurich, Switzerland
----------------------------------------------------------


On Tue, 17 May 2005, Christian Krekeler wrote:

> Hi,
> I try to calculate the Wannier functions for water and with these the
> dipole moment of water. This property is always too big.
> Can someone help me find the mistake?
>
> Regards,
> Christian
>
>