To get rid of the underflow flag output (which is essentially a debugging
output) compile with -ffpe-summary=none
See https://gcc.gnu.org/onlinedocs/gfortran/Debugging-Options.html
From: users on behalf of
tanghong...@outlook.com
Sent: Tuesday, 24 August
I fear your enterprise is destined for failure.
Performing periodic calculations on charged materials does not converge. They
can be made to converge by adding a jellium neutralising background, but I do
not think that it has been proven that that is an accurate approach. It's
worse for
While we are talking about interesting ways to include anharmonic
effects, it's worth taking a look at SSCHA: http://sscha.eu/
Not necessarily computationally inexpensive.
On Wed, 2021-03-31 at 07:20 +0900, Kazume NISHIDATE wrote:
> > 2021/03/31 1:00、Lorenzo Paulatto のメール:
> >
> > Hello you
Hi Oukahou & others
An approach we have had some success with, depending on the possible
states you are trying to distinguish, is to look at the spin density
around a nucleus. That is much more sensitive to formal oxidation than
the charge density. But YMMV.
Ciao
Terry
On Tue, 2021-02-09 at
Hi Mostaffa
To close the loop here, I appreciate your response, though I am familiar with
the Modern Theory and agree that the dipole from any particular cell is not
useful in isolation. However, the calculated dipole numbers should not jump
from e.g. -10 to +5 in the last iteration as the
Hi folks
I am trying to perform some finite field calculations on a
moderate-sized cell (~1000 Ang^3), using PW example 10 as a guide.
When doing a zero field calculation (all efield_cart zero) the total
energy converges happily to whatever level I want (e.g. conv_thr=1.d-9).
However, the
