yyp for DOS

[kazemoor]

Dear All
After obtaining QP spectra for my system I used ypp to get corrected DOS as follows
first I run ypp -s s and the run ypp, but the obtained dos is very little relative to GGA DOS ? Do I perform wrong with ypp option ?
best
Ali Kazempour ,Isfahan Univ .of Tech.,Iran

[claudio]

Dear Ali Kazempour

in order to obtain the QP density of states you have to use the comand

ypp -s s -V qp

and then read the QP database changing the line GfnQPdb= "none" in GfnQPdb= "E < SAVE/ndb.QP"

[kazemoor]

Dear Claudio
I did ypp -s s -V qp and change the line GfnQPdb= "none" in GfnQPdb= "E < SAVE/ndb.QP" and then run ypp , but it is similar to previous calculation and nothing is changed what can I do to get correct qp dos. and another qustion is why the hight of the qp dos is much less than GGA one?
best

Ali Kazempour ,Isfahan Univ .of Tech.,Iran

[claudio]

Dear Ali

do you have the file "SAVE/ndb.QP"?
if not it is normal that you get the same result,
moreover if you did not compile yambo with NetCDF (http://www.unidata.ucar.edu/software/netcdf/)
the name of the database will be "db.QP" and not "ndb.QP".

let me know

Claudio

[kazemoor]

Dear Claudio
I have db.QP in SAVE but again I got the wrong dos. Should I multiply qp DOS with somthing? for example for qp dos I have:
E qp-DOS
4.70373 0.08592
4.713729 0.103889
4.723730 0.102235
4.73373 0.09615
4.74373 0.08110
4.75373 0.06342
4.76373 0.06792
4.77373 0.09681
4.783735 0.121704
4.793736 0.128094
4.803736 0.141723

while for GGA DOS I have
4.700 0.7126E+01
4.710 0.7114E+01
4.720 0.7096E+01
4.730 0.7070E+01
4.740 0.7035E+01
4.750 0.6989E+01
4.760 0.6931E+01
4.770 0.6862E+01
4.780 0.6780E+01
4.790 0.6686E+01
4.800 0.6582E+01


Why Does these two differ so much??
best
Ali

[claudio]

Dear Ali

it is not so clear to me your problem, the qp-dos and cga-dos are different, this is normal.

Try a simpler case, a calculation with only one k-point (gamma point), and you will see immediatly the difference.

Moreover remember that if you have calculated the qp correction for a given set of bands, and then
you want the qp-dos for a larger number of bands, yambo uses an internal interpolation to generate the missing qp corrections.

Ohterwise put in the forum your input files, ground state and yambo input.

[kazemoor]

dear claudio
note that for example in energy 4.7 eV the dos value for qp is 0.08 and for gga is 7.1 . this difference is so much and it is seen for all value of the range.


yambo.in:
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= 2969 RL # [XX] Exchange RL components
% QpntsRXp
1 | 148 | # [Xp] Transferred momenta
%
% BndsRnXp
1 | 100 | # [Xp] Polarization function bands
%
NGsBlkXp= 1 RL # [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 | 100 | # [GW] G[W] bands range
%
GDamping= 0.10000 eV # [GW] G[W] damping
dScStep= 0.10000 eV # [GW] Energy step to evalute Z factors
%QPkrange # [GW] QP generalized Kpoint/Band indices
1|148| 1|100|
%
%QPerange # [GW] QP generalized Kpoint/Energy indices
1|148| 0.0|-1.0|


and ypp.in is

dos # [R] DOS
electrons # [R] Electrons (and holes)
GfnQPdb= "E < SAVE/db.QP" # [EXTQP G] Database
GfnQP_N= 1 # [EXTQP G] Interpolation neighbours
% GfnQP_E
0.000000 | 1.000000 | 1.000000 | # [EXTQP G] E parameters (c/v)
%
% GfnQP_W
0.000 | 0.000 | 0.000 | 0.000 | # [EXTQP G] W parameters (c/v)
%
GfnQP_Z= ( 1.000000 , 0.000000 ) # [EXTQP G] Z factor (c/v)
DOS_broad= 0.10000 eV # Broadening of the DOS
DOS_bands= 100 # DOS bands

thanks a lot

[claudio]

Dear Ali

you are right.

I checked your input file, but I didn't notice anything strange, except

NGsBlkXp= 1 RL # [Xp] Response block size

remember that you have to converge your calculations respect to the block size,
and 1 is not a converged result, so increase it until the quasi-particle energies do not change anymore.

Ali send also the o.qp file and the ground state ones,
at the moment I have no idea what went wrong.

Claudio

[kazemoor]

Dear claudio
I increased the NGsBlkXp to 150 ,but the problem didn't solve.
yambo.in:
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= 2969 RL # [XX] Exchange RL components
% QpntsRXp
1 | 148 | # [Xp] Transferred momenta
%
% BndsRnXp
1 | 100 | # [Xp] Polarization function bands
%
NGsBlkXp= 150 RL # [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 | 100 | # [GW] G[W] bands range
%
GDamping= 0.10000 eV # [GW] G[W] damping
dScStep= 0.10000 eV # [GW] Energy step to evalute Z factors
%QPkrange # [GW] QP generalized Kpoint/Band indices
1|148| 1|100|
%
%QPerange # [GW] QP generalized Kpoint/Energy indices
1|148| 0.0|-1.0|


ypp.in:

dos # [R] DOS
electrons # [R] Electrons (and holes)
GfnQPdb= "E < SAVE/db.QP" # [EXTQP G] Database
GfnQP_N= 1 # [EXTQP G] Interpolation neighbours
% GfnQP_E
0.000000 | 1.000000 | 1.000000 | # [EXTQP G] E parameters (c/v)
%
% GfnQP_W
0.000 | 0.000 | 0.000 | 0.000 | # [EXTQP G] W parameters (c/v)
%
GfnQP_Z= ( 1.000000 , 0.000000 ) # [EXTQP G] Z factor (c/v)
DOS_broad= 0.10000 eV # Broadening of the DOS
DOS_bands= 100 # DOS bands

I also get converged qp states for si as a example but i cannot obtain correct dos. WHY? may be I should multiply with somthing?
BEST

[claudio]

Dear Ali

some comment about your input files:

1) scf.in
remove nosym=.true. why you don't want to use the symmetries?
if you use them the calculations will be faster

2) bands.in
please do not use

calculation = 'bands'
but
calculation = 'nscf'

'bands' is not well supported in PWSCF.
Then remove nosym=.true. and put force_symmorphic=.true.

see discussion here: viewtopic.php?f=7&t=25

for the GW calculation only a restricted set of symmetries can be used.

Then remember that when you perform a 'bands' or 'nscf' calculation with ESPRESSO
the cutoff is "diago_thr_init = 1.0d-8" and not "conv_thr = 1.0d-8" !!!!
conv_thr has to be used only in 'scf' self-consistent calculation, see the ESPRESSO
input file /espresso-4.1.3/Doc/INPUT_PW.txt

3) please use a standard pseudopotential for example Si.pz-vbc.UPF
that you can find here: http://www.quantum-espresso.org/pseudo/1.3/html/Si.html

I checked the calculation and now everything works fine, I put in the attachment
the correct input files and a small calculation with only 10 bands,
remember to check the convergence on number of bands, block-size, cutoff etc..

Claudio