---BeginMessage---
Return-Path: [EMAIL PROTECTED]
Received: from listserv (listserv.uam.es [150.244.9.247])
by chumbera.ti.uam.es (8.13.7/8.13.4) with ESMTP id k5KJnBZv019167;
Wed, 21 Jun 2006 07:20:47 +0200 (CEST)
Received: by LISTSERV.UAM.ES (LISTSERV-TCP/IP release 14.3) with
Rudy:
No, the situation has not changed.
Jose
Rudy Coquet wrote:
Dear SIESTA users,
I would like to fix the bond length between two atoms adsorbed on a
surface, so that when the SIESTA calculations are running they remain as
a rigid bond but still able to be displaced with respect to the
Dear Rong:
Thanks for pointing out this problem. As Pablo said, siesta
does consider time-inversion symmetry for k points but,
as you have noticed, it does not do so completely.
Siesta was designed and written for large systems, and k-point
sampling was added later without much care for its
Nichols:
Siesta 1.3 calculates all eigenvectors but next version will
calculate only the occupied ones, since this is considerably
faster, specially with large basis sets (dzp).
To my knowledge, there is no switch to print the occupancies.
Jose
Nichols A. Romero wrote:
Hi,
I have an
Ivan:
Thanks for you bug report. I have included your correction.
Jose
Ivan Rungger wrote:
Dear siesta users,
for noncollinear spin systems it is possible to specify the magnitude
of the initial spin for each atom and the angles theta and phi in the
DM.InitSpin block. The angles are to be
Florent:
That would indeed be very useful. In fact, it is in
our to-do list since some time ago. Bu presently, no,
you cannot do it better that restart from scratch.
Jose
Florent Tournus wrote:
Hi there,
I wonder if there is a way to read a .DM file to restart a calculation
with a different
Mohamad:
The xc potential Vxc is indeed the difference between what is
dumped in the .VT and .VH files. But the xc energy Exc is NOT
the integral of Vxc*rho. It is rather the integral of eps_xc*rho,
where eps_xc is the xc energy density per electron. In the LDA,
Vxc = d(eps_xc*rho)/drho
Jose M
. A simpler version of that routine, that
avoids the complications of parallelization but is
equally operative, is in directory Util/Plrho.
But remember that you cannot redistribute your code
including a siesta routine.
Jose M. Soler
mohamad khazaei wrote:
Dear Friend,
hi,
hope you are fine.
I have
Dear Renato:
Look into program plrho how it handles the polarized rho array.
You will need to code a new option (say 'spin') that treats
the spin polarization (rho_up-rho_dn) as it now does for the
potential. In that way, you will obtain a constant-total-density
isosurface coloured with the spin
Dr. Claire Loison wrote:
Is it possible to use siesta without speudo-potentials,
using all atoms in the calculation ?
No, it is not.
Mads Brandbyge, Associate Professor in
the Technical University of Denmark,
has joined the SIESTA development team.
Mads is a specialist in electron transport
and the main author of the TranSiesta
method and code, a major development of
SIESTA to
Marcos:
The CPU load of matel is relatively small
and increasing NQ will not hurt much.
Increasing it from 1024 to just 2048
should be enough.
Jose
12 matches
Mail list logo