Wrong input file generation with yambo_ph

[jasonhan0710]

Dear Yambo developers,

I have tried to use yambo_ph to generate a BSE input file. I have used the yambo_ph to generate the GW input file (as blow) and run it successfully,

Code: Select all

dyson                            # [R] Dyson Equation solver
gw0                              # [R] GW approximation
el_ph_corr                       # [R] Electron-Phonon Correlation
ElecTemp= 300.0000         Kn    # Electronic Temperature
BoseTemp= 300.0000         Kn    # Bosonic Temperature
NLogCPUs=0                       # [PARALLEL] Live-timing CPU`s (0 for all)
PAR_def_mode= "balanced"         # [PARALLEL] Default distribution mode ("balanced"/"memory"/"workload")
SE_CPU= "1 7 4"                       # [PARALLEL] CPUs for each role
SE_ROLEs= "q qp b"                     # [PARALLEL] CPUs roles (q,qp,b)
SE_Threads=  4                   # [OPENMP/GW] Number of threads for self-energy
DysSolver= "n"                   # [GW] Dyson Equation solver ("n","s","g")
% GphBRnge
  1 | 12 |                           # [ELPH] G[W] bands range
%
% ElPhModes
  1 |  6 |                           # [ELPH] Phonon modes included
%
RandQpts=0                       # [RIM] Number of random q-points in the BZ
#WRgFsq                        # [ELPH] Dump on file gFsq coefficients
%QPkrange                        # [GW] QP generalized Kpoint/Band indices
1|8|1|12|
%

I got the quasi-particle database file "ndb.QP". Then I tried to get the BSE input file. However, the command

Code: Select all

yambo_ph -X s -o b -k sex -y d -V qp

generate the input file as blow

Code: Select all

#                http://www.yambo-code.org                            
#
optics                           # [R] Linear Response optical properties
bss                              # [R] BSE solver
bse                              # [R][BSE] Bethe Salpeter Equation.
X_Threads=0                      # [OPENMP/X] Number of threads for response functions
DIP_Threads=0                    # [OPENMP/X] Number of threads for dipoles
K_Threads=0                      # [OPENMP/BSK] Number of threads for response functions
BSEmod= "resonant"               # [BSE] resonant/retarded/coupling
BSKmod= "HARTREE"                # [BSE] IP/Hartree/HF/ALDA/SEX/BSfxc
BSSmod= "d"                      # [BSS] (h)aydock/(d)iagonalization/(s)lepc/(i)nversion/(t)ddft`
BSENGexx=  9257            RL    # [BSK] Exchange components
KfnQPdb= "none"                  # [EXTQP BSK BSS] Database action
KfnQP_INTERP_NN= 1               # [EXTQP BSK BSS] Interpolation neighbours (NN mode)
KfnQP_INTERP_shells= 20.00000    # [EXTQP BSK BSS] Interpolation shells (BOLTZ mode)
KfnQP_DbGd_INTERP_mode= "NN"     # [EXTQP BSK BSS] Interpolation DbGd mode
% KfnQP_E
 0.000000 | 1.000000 | 1.000000 |        # [EXTQP BSK BSS] E parameters  (c/v) eV|adim|adim
%
KfnQP_Z= ( 1.000000 , 0.000000 )         # [EXTQP BSK BSS] Z factor  (c/v)
KfnQP_Wv_E= 0.000000       eV    # [EXTQP BSK BSS] W Energy reference  (valence)
% KfnQP_Wv
 0.000000 | 0.000000 | 0.000000 |        # [EXTQP BSK BSS] W parameters  (valence) eV| 1|eV^-1
%
KfnQP_Wv_dos= 0.000000     eV    # [EXTQP BSK BSS] W dos pre-factor  (valence)
...

Obviously, it didn't generate the dielectric function variable "em1s" and the BSE calculation stop because it cannot find the "em1s" file. the command

Code: Select all

yambo_ph -X s

cannot generate the correct input file either.

Would you please help me to take a look?

Best regards,
Jason

[Davide Sangalli]

Dear Jason,
which version of yambo are you using?

With the latest version I'm able to correctly produce the input file.
Can you attach the r_setup file for your system?

Best,
D.

[jasonhan0710]

Dear Davide,

Thank you for your reply.

I am now using the latest version (5.0.4) of yambo, however, the dielectric function related parameters still cannot be generated by "yambo -X s" or "yambo -X d". The same error happened in 5.0.3 or older version. Please find the report file as attached.

Best,
Jason

[Davide Sangalli]

Dear Jason,
sorry for the late reply.
The issue is that the k-points grid is detected as "not uniform".
See this line in the report

Code: Select all

  X grid is uniform                                : no

As a consequence yambo cannot compute the screening.
I'm not sure why this is happening, since indeed the k-points grid looks uniform to me.
Can you provide the input file you used with quantum-espresso.

Best,
D.

[jasonhan0710]

Dear Davide,

Thanks for your reply. The input files were download at the tutorial page http://www.yambo-code.org/educational/t ... si.epc.tgz (also see the attachment). I used QE-6.6 version and did not change the input files.

If I use "yambo" command to generate the BSE input file, it gives the "NaN" in the absorption output file (the input file and output files were in the "BSE" attachment). Would you please help me figure it out?

Thank you!

Best,
Jason

[Davide Sangalli]

Dear Jason,
I think something is wrong in the SAVE you are using.

I used the same input files and I correctly got that the k-grid is uniform.

Code: Select all

 [03] Transferred momenta grid and indexing
 ==========================================


 IBZ Q-points :   8
 BZ  Q-points :   64

 K/Q-points units:
 rlu = crystal or reduced units; cc = cartesian coordinates; iku = interal k-units

 Q [1]:  0.000000  0.000000  0.000000 [rlu]
 Q [2]:  0.000000  0.000000  0.250000 [rlu]
 Q [3]:  0.000000  0.000000 -0.500000 [rlu]
 Q [4]:  0.000000  0.250000  0.250000 [rlu]
 Q [5]:  0.000000  0.250000  0.500000 [rlu]
 Q [6]:  0.000000  0.250000 -0.250000 [rlu]
 Q [7]:  0.000000 -0.500000 -0.500000 [rlu]
 Q [8]:  0.250000 -0.500000 -0.250000 [rlu]

Not sure why. Maybe something went wrong in the pwscf calculation or in the p2y conversion.

Can you write in details all the steps you did up to the inizialization to obtaine the r_setup file

Best,
D.

[jasonhan0710]

Dear Davide,

The whole process was done as follows:

1. In 'scf' folder, run 'pw.x' to get the 'si.save' folder, and then copy it into 'nscf' folder;
2. In 'nscf' folder, run 'pw.x', after that, run 'p2y' in 'si.save' folder and get the 'SAVE' folder. Next, run 'yambo_ph' to set up and run 'ypp_ph' to get the q-points;
3. In 'nscf' folder, copy the 'si.save' folder to 'phonon' folder and 'dvscf' folder respectively;
4. run 'ph.x' in the 'phonon' folder, after get '_ph0' and dynamic matrix files, move them into 'dvscf' folder;
5. run 'ph.x' in 'dvscf' folder and get the 'elph_dir' folder. Then, run 'ypp_ph'. Move 'SAVE' folder into a new directory and run 'yambo_ph'.

Is there something wrong in my calculation?

Thank you!

Best,
Jason

[jasonhan0710]

Dear Davide,

I noticed that I copy a wrong 'is.save' folder in step 3, which makes the k grid not uniform. After I fixed the mistake, another error occurred, which reads

Code: Select all

Intel MKL ERROR: Parameter 3 was incorrect on entry to CGEBAL.

And the calculation stops at

Code: Select all

   [03.04.01] Diago Solver @q1
   ===========================


   [WARNING]Parallel non-herm diag not yet implemented. Switching to LK

Which configuration options should I use to avoid linking the intel mkl package?

Thank you!

Best,
Jason