About G-vectors

[Daniele Varsano]

Dear Marina,
I can see the mirror of the input file from the report,
can you please also post the input files you used for the setup and the GW calculation?
Did you use the MaxGves variable in any of your input files?

Best,
Daniele

[mfilip]

Dear Daniele,

Many thanks for your quick reply. I did not use the MaxGvecs variable for these calculations, I was actually trying to test what the effect of lowering this variable, so for my first calculation I wanted to use all of them.

For the setup I simply used the default input from yambo -i.
Please see below the input file I have been playing with for the GW calculation, but where I really got stuck at is the initialization point where the assigned Gvectors were too few.

Many thanks for your help,

Marina

Marina Filip
Postdoctoral Research Assistant
University of Oxford,
Department of Materials

gw0 # [R GW] GoWo Quasiparticle energy levels
ppa # [R Xp] Plasmon Pole Approximation
HF_and_locXC # [R XX] Hartree-Fock Self-energy and Vxc
em1d # [R Xd] Dynamical Inverse Dielectric Matrix
EXXRLvcs= 30 Ry # [XX] Exchange RL components
Chimod= "Hartree" # [X] IP/Hartree/ALDA/LRC/BSfxc
% BndsRnXp
1 | 200 | # [Xp] Polarization function bands
%
NGsBlkXp= 6.0 Ry # [Xp] Response block size
% LongDrXp
1.000000 | 0.000000 | 0.000000 | # [Xp] [cc] Electric Field
%
PPAPntXp= 27.21138 eV # [Xp] PPA imaginary energy
% GbndRnge
1 | 200 | # [GW] G[W] bands range
%
GDamping= 0.10000 eV # [GW] G[W] damping
dScStep= 0.10000 eV # [GW] Energy step to evaluate Z factors
DysSolver= "n" # [GW] Dyson Equation solver ("n","s","g")
%QPkrange # [GW] QP generalized Kpoint/Band indices
1| 6| 44| 45|
%

[Davide Sangalli]

Dear Marina,
can you upload the DFT input file and pseudo-potentials ?
We will try to reproduce the problem.

Best,
D.

[mfilip]

Dear Davide,

Once again, many thanks for looking into this - I really appreciate it.
Please find attached the dft input (scf and nscf) and the pseudopotentials (PBE, oncv).

Please let me know if you need any other details.

Kind Regards
Marina

Marina Filip
Postdoctoral Research Assistant
University of Oxford
Department of Materials

[Davide Sangalli]

Dear Marina,
just a couple of observation before I try to reproduce the behaviour, related to the inpus you just uploaded

1 - in the nscf calculation the cut-off on the WFs is different from the scf calculation.
This is not possible, since the density needs to be on the same grid in G-space

2 - Moreover the restart_mode should be "restart", not from scratch.
This is why you did not notice the problem with the cut-off.

Do you still get the same behaviour once changing these to parameters in the nscf calculation ?

D.

[mfilip]

Dear Davide,

My apologies, I accidentally copied another version of the nscf input file. Just to be clear, the cutoff I used for both scf and nscf is of 100 Ry.
For the restart option, the output renders identical eigenvalues either way. However, I just checked if the number of G-vectors changes if I use 'restart' instead of 'from scratch' and I get the same number in the initialization, 749 RL.

Thanks again,

Marina

Marina Filip
Postdoctoral Research Assistant
University of Oxford,
Department of Materials

[Davide Sangalli]

Ok.
I tried and this is what I get

Code: Select all

 [RD./SAVE//ns.db1]------------------------------------------
  Bands                           :  50
  K-points                        : 24
  G-vectors             [RL space]:  54601
  Components       [wavefunctions]:  6855
  Symmetries       [spatial+T-rev]:  4
  Spinor components               : 2
  Spin polarizations              : 1
  Temperature                 [ev]: 0.000000
  Electrons                       : 44.00000
  WF G-vectors                    :  8355
  Max atoms/species               : 1
  No. of atom species             : 2
  Magnetic symmetries             : no
 - S/N 004506 -------------------------- v.04.03.00 r.15247 -

I reproduce the problem if I compile yambo in single precision.
You can just recompile everything in double precision (--enable-dp at configure time), and then regenerate the SAVE.
This should solve the issue.

D.