Page 1 of 1

GW calculations from Quantum Espresso, using ibrav=5

Posted: Mon Jan 30, 2023 5:03 pm
by Joe_Page_93
Hi All,

I am calculating the electronic structure of a material using Quantum Espresso. While defining the symmetry of the system I am using ibrav = 5, which corresponds to a rhombohedral lattice type.

The SCF and NSCF calculations both run and finish, with no errors (I am using an automatic k-point grid). After performing p2y, and the yambo initialization command I want to get the HF convergences, however when I do the command 'yambo -x -F hf.in' the hf.in file is empty (apart from the yambo header).

I have assessed the r_setup file (and r_setup_1, as i had 80 k-points, so had to use the yambo -fatlog command), and there doesn't seem to be any errors in it (see attached).

Is there something I am missing about the symmetry, as it works great when I use ibrav=4. Or have I done something else wrong that causing the problem.

Thanks you all,
Joe Page

Re: GW calculations from Quantum Espresso, using ibrav=5

Posted: Wed Feb 01, 2023 9:21 am
by Daniele Varsano
Dear Joe,
the problem here is due to the fact that Yambo does not recognize your k point grid as uniform:
see at the end of the setup report:

Code: Select all

X grid is uniform                                : no
You can try to remove all your ndb.* databases and run again the setup adding the keyword (NoDiagSC ) in the setup input file, or generate the input file using:

Code: Select all

yambo -i -V all 
and uncomment that variable.

In this way, Yambo use a different algorithm to make the set-up for non-diagonal cell. See if this solves the problem.

Best,
Daniele

Re: GW calculations from Quantum Espresso, using ibrav=5

Posted: Thu Feb 02, 2023 12:03 pm
by Joe_Page_93
Dear Daniele,

I have performed a number of different tests based on what you sent. The first was to reduce the nscf k-point grid to 3x3x3, which resulted in the HF.in file working, but the overall GW band structure is not quite correct.

So I increased the nscf k-point grid, generated the save file, then used what you sent me. So I used the command 'yambo -i -V all', and put 'NoDiagSC' in the file. Ran yambo using 'yambo -F yambo.in', which seemed to solve the X grid is uniform issue, finally removing the ndb.* files from the SAVE folder.

I then ran the 'yambo -x -F hf.in' command and the final generated only has the 'setup' word in it.

I have attached the r_setup file and the new HF.in file.

Have I gone wrong somewhere along the way?

Re: GW calculations from Quantum Espresso, using ibrav=5

Posted: Mon Feb 06, 2023 10:13 am
by Daniele Varsano
Dear Joseph,

looking at the setup report it seems the issue it is not solved:

Code: Select all

X grid is uniform                                : no
My suggestion here is:
1. remove the ndb* in the SAVE directory
2. Build a yambo_setup.in input file containing only:

Code: Select all

setup                            # [R] Initialization
NoDiagSC 
3. run yambo -F yambo_setup.in

and see id you finally get a "yes" in the X grid is uniform.

Best,
Daniele