WARNING: missing poles for contour deformation

GW, Bethe-Salpeter …

Moderators: maryam.azizi, bruneval

Locked
lorenzo.sponza
Posts: 9
Joined: Wed Nov 07, 2012 11:40 am

WARNING: missing poles for contour deformation

Post by lorenzo.sponza » Tue Nov 13, 2012 3:34 pm

Hi.
I am making the GW tutorials with version 6.12.3.
In running 'tgw2_3.in' I had some WARNING I can't understand:
calc_sigc_me.F90:1363:WARNING
Total number of missing poles for contour deformation method: 4003
For band 1 there are 4003 missing poles
Max omega value used in W(omega): 50.31635 eV

What does it mean?
I know that caring about warnings is somewhat masochism, but I was curious.

By the way I got a look to this warning because the results of the tutorial does not coincide with those in the website which are much closer to previous versions of ABINIT (as 5.6). And previous versions calculations didn't give any warning.
Thanks, cheers.

gabriel.antonius
Posts: 58
Joined: Mon May 03, 2010 10:34 pm

Re: WARNING: missing poles for contour deformation

Post by gabriel.antonius » Tue Nov 13, 2012 6:32 pm

Hi Lorenzo,

What this means, is that the real-axis frequency grid used for the screening does not go far enough to compute the corrections to *all* bands.
This is ok though, if you wish to compute only the corrections on the first few bands.
With the contour deformation method, the maximum frequency needed to obtain the corrections on a band of energy E_i is ( E_i - E_f ), where E_f is the fermi energy.
Hence, if you only want corrections around the gap, you don't need a big frequency grid on the real axis.

Cheers!
Gabriel Antonius
Université du Québec à Trois-Rivières

Locked