initialization for ELPH database

[ymzhang]

Dear sir,
I want to calculate electron phonon interaction in Silicon regime and I make the following preparations as the yambo-py codes suggest:
1) scf calculation on a shifted 8*8*8 kgrid (scf.in)
2) nscf calculation on an unshifted 4*4*4 kgrid (nscf_4x4x4.in) using the save from step 1
3) use p2y to generate a SAVE file from step 2 to generate 20 random qpoints and substitute the first one with gamma point for the phonon calculation (ypp -k r -F kgrid.in)
4) use the save in step 2 (coarse grid) to calculate a elph matrix(02ph.in) on the qpoints in step 3
5) nscf calculation on an unshifted 8*8*8 kgrid (nscf_8x8x8.in) using the save from step 1
6) use the save in step 5 (dense grid) and elph matrix in step 4 to calculate the energy on the 20 random qpoints (04elph.in)
7) compile the s.dbph_0000* from elph_dir and SAVE generated from step 5 (dense grid) together to a new folder for further calculation
then I initialize the database and get the following results:
First, I use yambo_ph -i -V kpt -F yambo.in to generate the input and manually append MinusQ to the file as follow:

Code: Select all

setup                          # [R INI] Initialization
% IkSigLim
  1 | 29 |                     # [KPT] QP K-points indices range
%
IkXLim= 29                     # [KPT] X grid last k-point index
#BSEscatt                      # [KPT] Compute extended k/q scatering
MinusQ

The report reads like that:

Code: Select all

 <---> [01] CPU structure, Files & I/O Directories
 <---> [02] CORE Variables Setup
 <---> [02.01] Unit cells
 <---> [02.02] Symmetries
 <---> [02.03] RL shells
 <---> Shells finder |########################################| [100%] --(E) --(X)
 <---> [02.04] K-grid lattice
 <---> Grid dimensions               :  8   8   8
 <---> [02.05] Energies [ev] & Occupations
 <---> [03] Transferred momenta grid
 <---> [RL indx] Q-grid:Electron-Phonon databases
 <---> [RL indx] Q BZ pts are /= from X grid BS pts. Gamma point only.
 <---> [04] Timing Overview
 <---> [05] Game Over & Game summary

It seems there is some mismatch in the Qpoints and only the gamma point is successfully. I use the 'alat' form in cooOut tag as the yambopy do, I'm not sure whether it's suitable. Then I use ypp -g -F ypp.in to get the following inputs:

Code: Select all

gkkp                           # [R] Electron-Phonon databases
DBsPATH= "."                   # Path to the PW el-ph databases
PHfreqF= "none"                # PWscf format file containing the phonon frequencies
PHmodeF= "none"                # PWscf format file containing the phonon modes
#GkkpExpand                    # Expand the gkkp in the whole BZ
#GkkpConvert                   # Convert the gkkp to new I/O format

Code: Select all

<---> [01] CPU structure, Files & I/O Directories
 <---> [02] Y(ambo) P(ost)/(re) P(rocessor)
 <---> [03] Core DB
 <---> :: Electrons             : 8.000000
 <---> :: Temperature       [ev]: 0.000000
 <---> :: Lattice factors [a.u.]: 5.091500  5.091500  5.091500
 <---> :: K points              : 29
 <---> :: Bands                 :  50
 <---> :: Symmetries            :  48
 <---> :: RL vectors            : 2085
 <---> [04] K-point grid
 <---> :: Q-points   (IBZ): 1
 <---> :: X K-points (IBZ): 29
 <---> [05] CORE Variables Setup
 <---> [05.01] Unit cells
 <---> [05.02] Symmetries
 <---> [05.03] RL shells
 <---> [05.04] K-grid lattice
 <---> Grid dimensions               :  8   8   8
 <---> [05.05] Energies [ev] & Occupations
 <---> [06] == Electron-Phonon Interface: PW->Yambo Databases ==
 <---> PW(ELPH) databases ...[PHONON] ...found 20 NOT Q-grid compatible
 <58s> ELPH databases (WRITE) |########################################| [100%] 58s(E) 58s(X)
 <58s> :: Modes           :  6
 <58s> :: Bands range     :  50
 <58s> Databases Q-points PW-formatted
       0.000000000  0.000000000  0.000000000 1
      -0.152828380 -0.435727954  0.045498498 1
       0.334549755 -0.009765699 -0.307347476 1
       0.433034539  0.073082991 -0.131813496 1
       0.215498969  0.171638772  0.350320786 1
       0.187392876 -0.167791083  0.041487098 1
      -0.020649195 -0.418981194  0.134727001 1
      -0.144443765 -0.406561524 -0.099925995 1
      -0.082937986 -0.278221965  0.101857394 1
      -0.054322589 -0.211238682 -0.259933800 1
       0.212642178 -0.187416583  0.234584585 1
      -0.480074733  0.088339582 -0.141250789 1
      -0.283740252 -0.046697296 -0.289366126 1
       0.185389072  0.487134457 -0.068466999 1
      -0.207159266  0.305032223  0.176419392 1
      -0.182985485  0.001356000  0.194662094 1
      -0.185574368  0.306555659 -0.153673396 1
      -0.208446771 -0.026296096  0.118550502 1
       0.183740675 -0.151414290  0.083759591 1
      -0.491928846 -0.079484589  0.173904687 1
 <58s> [07] Game Over & Game summary

It seems things are working well except for the "NOT Q-grid compatible" warning, then I try to use 'yambo_ph -g g -F green.in' for a full frequency calculation and get an enpty input file. There must be something wrong in the above steps. I have tried to use a uniformly distributed qpoints and the input is still empty. Then I try to exclude the ELPH database and calculate only the frozen atom green's function and everything is fine. Hopefully you can help me to check out the problem.
All the above calculation was performed on Yambo 4.5.1 and QE 6.4.1
ps: I think it's much more proper to post this question on the electron phonon interaction subforum but I didn't find a new topic bottom there. So I can only post it here
Thanks,

[palful]

Dear ymzhang,

First of all, may I ask you to add a digital signature including affiliation to your posts?

Second: can you prepare a small example that we can easily run, including all the steps leading to your issue? In this way, we can check it and quickly understand what the problem is.

Best,
Fulvio