Dear all,
In Yambo one can use plasmon pole model to calculate GW gap, and PPA value can been set in the Input file. However, I am wondering when we should consider if the PPA fails and we have to go beyond PPA?
Thanks!
PPA
Moderators: Davide Sangalli, andrea.ferretti, myrta gruning, andrea marini, Daniele Varsano
- Daniele Varsano
- Posts: 3816
- Joined: Tue Mar 17, 2009 2:23 pm
- Contact:
Re: PPA
Dear Shudong,
as a first check you can have a look to the electron energy loss spectrum (EELS) in linear response and see if the dielectric matrix (at least the 00 component) present a clear plasmon excitation and it can be fitted by a one-pole function. If it present a more sophisticated structure, it can be an hint that the PPA could fail. To have an idea, you can have a look to Fig.2 of Hybertsen and Louie PRB 34 5390 (1986).
Other option, is to perform a full real axis and compare with the PPA, of course the real axis calculation much harder to converge.
Cheers,
Daniele
as a first check you can have a look to the electron energy loss spectrum (EELS) in linear response and see if the dielectric matrix (at least the 00 component) present a clear plasmon excitation and it can be fitted by a one-pole function. If it present a more sophisticated structure, it can be an hint that the PPA could fail. To have an idea, you can have a look to Fig.2 of Hybertsen and Louie PRB 34 5390 (1986).
Other option, is to perform a full real axis and compare with the PPA, of course the real axis calculation much harder to converge.
Cheers,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/
-
- Posts: 287
- Joined: Fri Apr 09, 2010 12:30 pm
-
- Posts: 287
- Joined: Fri Apr 09, 2010 12:30 pm
Re: PPA
Dear Daniele,
1). I am now trying to test real axis calculation, when I used./yambo -g n -d, the calculation is very quick than ./yambo -g n, and what is the difference between these two methods,or which one is more reasionable? (I used version yambo-3.3.0-rev41).
2). when I use -g n -d, I am confused about parameters as following:
% EnRngeXd
0.00000 | 20.00000 | eV # [Xd] Energy range (how to set this? I found in output file, it is changed, for Silicon is 0.000| 48.88186)
%
% DmRngeXd
0.1000 | 0.10000 | eV # [Xd] Damping range (this also changes gap with different value?)
%
DmERefXd= 0.000000 eV # [Xd] Damping reference energy
CGrdSpXd= 100.0000 # [Xd] [o/o] Coarse grid controller
ETStpsXd= 500 ( and this value need to test convergence? I found 500 the gap as same as 1000 even 2000)
The GW ppa gap is 1.24 eV ,but why real axis gap is 2.24eV?
Rela axis input is:
rim_cut # [R RIM CUT] Coulomb interaction
gw0 # [R GW] GoWo Quasiparticle energy levels
em1d # [R Xd] Dynamical Inverse Dielectric Matrix
HF_and_locXC # [R XX] Hartree-Fock Self-energy and Vxc
StdoHash= 10 # [IO] Live-timing Hashes
Nelectro= 8.000000 # Electrons number
ElecTemp= 0.000000 eV # Electronic Temperature
BoseTemp=-1.000000 eV # Bosonic Temperature
OccTresh=0.1000E-4 # Occupation treshold (metallic bands)
FFTGvecs= 411 RL # [FFT] Plane-waves
RandQpts= 1000000 # [RIM] Number of random q-points in the BZ
RandGvec= 1 RL # [RIM] Coulomb interaction RS components
#QpgFull # [F RIM] Coulomb interaction: Full matrix
% Em1Anys
0.00 | 0.00 | 0.00 | # [RIM] X Y Z Static Inverse dielectric matrix
%
IDEm1Ref=0 # [RIM] Dielectric matrix reference component 1(x)/2(y)/3(z)
CUTGeo= "none" # [CUT] Coulomb Cutoff geometry: box/cylinder/sphere
% CUTBox
0.00 | 0.00 | 0.00 | # [CUT] [au] Box sides
%
CUTRadius= 0.000000 # [CUT] [au] Sphere/Cylinder radius
CUTCylLen= 0.000000 # [CUT] [au] Cylinder length
#CUTCol_test # [CUT] Perform a cutoff test in R-space
EXXRLvcs= 7 Ry # [XX] Exchange RL components
XfnQPdb= "none" # [EXTQP Xd] Database
XfnQP_N= 1 # [EXTQP Xd] Interpolation neighbours
% XfnQP_E
0.000000 | 1.000000 | 1.000000 | # [EXTQP Xd] E parameters (c/v) eV|adim|adim
%
% XfnQP_Wv
0.00 | 0.00 | 0.00 | # [EXTQP Xd] W parameters (valence) eV|adim|eV^-1
%
% XfnQP_Wc
0.00 | 0.00 | 0.00 | # [EXTQP Xd] W parameters (conduction) eV|adim|eV^-1
%
XfnQP_Z= ( 1.000000 , 0.000000 ) # [EXTQP Xd] Z factor (c/v)
% Qdirection
0.00 | 0.00 | 0.00 | # [Xd] Transferred momentum direction (iku)
%
QShiftOrder= 1 # [Xd] Pick-up the (QShiftOrder)th q+G vector
% QpntsRXd
1 | 8 | # [Xd] Transferred momenta
%
% BndsRnXd
1 | 30 | # [Xd] Polarization function bands
%
NGsBlkXd= 7 Ry # [Xd] Response block size
GrFnTpXd= "t" # [Xd] Green`s function t/c/r/a
% EnRngeXd
0.00000 | 20.00000 | eV # [Xd] Energy range
%
% DmRngeXd
0.100 | 0.1000 | eV # [Xd] Damping range
%
DmERefXd= 0.000000 eV # [Xd] Damping reference energy
CGrdSpXd= 100.0000 # [Xd] [o/o] Coarse grid controller
ETStpsXd= 1000 # [Xd] Total Energy steps
EMStpsXd= 100.0000 # [Xd] [o/o] Memory Energy steps
DrudeWXd= ( 0.00 , 0.00 ) eV # [Xd] Drude plasmon
% EhEngyXd
-1.000000 |-1.000000 | eV # [Xd] Electron-hole energy range
%
% LongDrXd
1.000000 | 0.000000 | 0.000000 | # [Xd] [cc] Electric Field
%
GfnQPdb= "none" # [EXTQP G] Database
GfnQP_N= 1 # [EXTQP G] Interpolation neighbours
% GfnQP_E
0.000000 | 1.000000 | 1.000000 | # [EXTQP G] E parameters (c/v) eV|adim|adim
%
% GfnQP_Wv
0.00 | 0.00 | 0.00 | # [EXTQP G] W parameters (valence) eV|adim|eV^-1
%
% GfnQP_Wc
0.00 | 0.00 | 0.00 | # [EXTQP G] W parameters (conduction) eV|adim|eV^-1
%
GfnQP_Z= ( 1.000000 , 0.000000 ) # [EXTQP G] Z factor (c/v)
BoseCut= 0.10000 # [BOSE] Finite T Bose function cutoff
% GbndRnge
1 | 30 | # [GW] G[W] bands range
%
GDamping= 0.10000 eV # [GW] G[W] damping
dScStep= 0.10000 eV # [GW] Energy step to evalute Z factors
#GWTerm # [GW] Use a terminator for the self-energy
GwEnComp=-27.21138 eV # [GW] If <= 0 use 0-order EET ; > 0 the Bruneval-Gonze terminator
DysSolver= "n" # [GW] Dyson Equation solver (`n`,`s`,`g`)
#NewtDchk # [F GW] Test dSc/dw convergence
#ExtendOut # [F GW] Print all variables in the output file
%QPkrange # [GW] QP generalized Kpoint/Band indices
1| 8| 1| 8|
%
%QPerange # [GW] QP generalized Kpoint/Energy indices
1| 8| 0.0|-1.0|
%
Thanks!
1). I am now trying to test real axis calculation, when I used./yambo -g n -d, the calculation is very quick than ./yambo -g n, and what is the difference between these two methods,or which one is more reasionable? (I used version yambo-3.3.0-rev41).
2). when I use -g n -d, I am confused about parameters as following:
% EnRngeXd
0.00000 | 20.00000 | eV # [Xd] Energy range (how to set this? I found in output file, it is changed, for Silicon is 0.000| 48.88186)
%
% DmRngeXd
0.1000 | 0.10000 | eV # [Xd] Damping range (this also changes gap with different value?)
%
DmERefXd= 0.000000 eV # [Xd] Damping reference energy
CGrdSpXd= 100.0000 # [Xd] [o/o] Coarse grid controller
ETStpsXd= 500 ( and this value need to test convergence? I found 500 the gap as same as 1000 even 2000)
The GW ppa gap is 1.24 eV ,but why real axis gap is 2.24eV?
Rela axis input is:
rim_cut # [R RIM CUT] Coulomb interaction
gw0 # [R GW] GoWo Quasiparticle energy levels
em1d # [R Xd] Dynamical Inverse Dielectric Matrix
HF_and_locXC # [R XX] Hartree-Fock Self-energy and Vxc
StdoHash= 10 # [IO] Live-timing Hashes
Nelectro= 8.000000 # Electrons number
ElecTemp= 0.000000 eV # Electronic Temperature
BoseTemp=-1.000000 eV # Bosonic Temperature
OccTresh=0.1000E-4 # Occupation treshold (metallic bands)
FFTGvecs= 411 RL # [FFT] Plane-waves
RandQpts= 1000000 # [RIM] Number of random q-points in the BZ
RandGvec= 1 RL # [RIM] Coulomb interaction RS components
#QpgFull # [F RIM] Coulomb interaction: Full matrix
% Em1Anys
0.00 | 0.00 | 0.00 | # [RIM] X Y Z Static Inverse dielectric matrix
%
IDEm1Ref=0 # [RIM] Dielectric matrix reference component 1(x)/2(y)/3(z)
CUTGeo= "none" # [CUT] Coulomb Cutoff geometry: box/cylinder/sphere
% CUTBox
0.00 | 0.00 | 0.00 | # [CUT] [au] Box sides
%
CUTRadius= 0.000000 # [CUT] [au] Sphere/Cylinder radius
CUTCylLen= 0.000000 # [CUT] [au] Cylinder length
#CUTCol_test # [CUT] Perform a cutoff test in R-space
EXXRLvcs= 7 Ry # [XX] Exchange RL components
XfnQPdb= "none" # [EXTQP Xd] Database
XfnQP_N= 1 # [EXTQP Xd] Interpolation neighbours
% XfnQP_E
0.000000 | 1.000000 | 1.000000 | # [EXTQP Xd] E parameters (c/v) eV|adim|adim
%
% XfnQP_Wv
0.00 | 0.00 | 0.00 | # [EXTQP Xd] W parameters (valence) eV|adim|eV^-1
%
% XfnQP_Wc
0.00 | 0.00 | 0.00 | # [EXTQP Xd] W parameters (conduction) eV|adim|eV^-1
%
XfnQP_Z= ( 1.000000 , 0.000000 ) # [EXTQP Xd] Z factor (c/v)
% Qdirection
0.00 | 0.00 | 0.00 | # [Xd] Transferred momentum direction (iku)
%
QShiftOrder= 1 # [Xd] Pick-up the (QShiftOrder)th q+G vector
% QpntsRXd
1 | 8 | # [Xd] Transferred momenta
%
% BndsRnXd
1 | 30 | # [Xd] Polarization function bands
%
NGsBlkXd= 7 Ry # [Xd] Response block size
GrFnTpXd= "t" # [Xd] Green`s function t/c/r/a
% EnRngeXd
0.00000 | 20.00000 | eV # [Xd] Energy range
%
% DmRngeXd
0.100 | 0.1000 | eV # [Xd] Damping range
%
DmERefXd= 0.000000 eV # [Xd] Damping reference energy
CGrdSpXd= 100.0000 # [Xd] [o/o] Coarse grid controller
ETStpsXd= 1000 # [Xd] Total Energy steps
EMStpsXd= 100.0000 # [Xd] [o/o] Memory Energy steps
DrudeWXd= ( 0.00 , 0.00 ) eV # [Xd] Drude plasmon
% EhEngyXd
-1.000000 |-1.000000 | eV # [Xd] Electron-hole energy range
%
% LongDrXd
1.000000 | 0.000000 | 0.000000 | # [Xd] [cc] Electric Field
%
GfnQPdb= "none" # [EXTQP G] Database
GfnQP_N= 1 # [EXTQP G] Interpolation neighbours
% GfnQP_E
0.000000 | 1.000000 | 1.000000 | # [EXTQP G] E parameters (c/v) eV|adim|adim
%
% GfnQP_Wv
0.00 | 0.00 | 0.00 | # [EXTQP G] W parameters (valence) eV|adim|eV^-1
%
% GfnQP_Wc
0.00 | 0.00 | 0.00 | # [EXTQP G] W parameters (conduction) eV|adim|eV^-1
%
GfnQP_Z= ( 1.000000 , 0.000000 ) # [EXTQP G] Z factor (c/v)
BoseCut= 0.10000 # [BOSE] Finite T Bose function cutoff
% GbndRnge
1 | 30 | # [GW] G[W] bands range
%
GDamping= 0.10000 eV # [GW] G[W] damping
dScStep= 0.10000 eV # [GW] Energy step to evalute Z factors
#GWTerm # [GW] Use a terminator for the self-energy
GwEnComp=-27.21138 eV # [GW] If <= 0 use 0-order EET ; > 0 the Bruneval-Gonze terminator
DysSolver= "n" # [GW] Dyson Equation solver (`n`,`s`,`g`)
#NewtDchk # [F GW] Test dSc/dw convergence
#ExtendOut # [F GW] Print all variables in the output file
%QPkrange # [GW] QP generalized Kpoint/Band indices
1| 8| 1| 8|
%
%QPerange # [GW] QP generalized Kpoint/Energy indices
1| 8| 0.0|-1.0|
%
Thanks!
- myrta gruning
- Posts: 240
- Joined: Tue Mar 17, 2009 11:38 am
- Contact:
Re: PPA
I suppose you used the same parameters in the two calculations.sdwang wrote:1). I am now trying to test real axis calculation, when I used./yambo -g n -d, the calculation is very quick than ./yambo -g n, and what is the difference between these two methods,or which one is more reasionable? (I used version yambo-3.3.0-rev41).
Just a question. Did you run the calculations on the same directory using the same SAVE? Did you check the report/log?
It may be that the calculation was faster just because yambo found it could re-use one or more of the databases from the previous calculation, and did not repeat that/those step.
Dr Myrta Grüning
School of Mathematics and Physics
Queen's University Belfast - Northern Ireland
http://www.researcherid.com/rid/B-1515-2009
School of Mathematics and Physics
Queen's University Belfast - Northern Ireland
http://www.researcherid.com/rid/B-1515-2009
- Daniele Varsano
- Posts: 3816
- Joined: Tue Mar 17, 2009 2:23 pm
- Contact:
Re: PPA
Dear Shudong,
RD or WR label in front of the name of the database.
Next:
1. Yes, ETStpsXd, these are the number of the steps of your integration and has to be converged.
2. EnRngeXd, this is changed by Yambo, set to the maximum interval energy of the band you included.
3. The differences you find is very big, and unexpected, are you using the same number of bands, NGsBlkXd, EXXRLvcs, bands in GbndRnge?
4. As a general remark I suggest you to switch to the new version.
Best,
Daniele
There is only one way to perform the real axis calculations. If you obtain different execution time, I wonder if in the one case you are reading databases calculated before, and in other care you are producing that database. Having a look to the report check for the databases (at least the screening) if you have1). I am now trying to test real axis calculation, when I used./yambo -g n -d, the calculation is very quick than ./yambo -g n, and what is the difference between these two methods,or which one is more reasionable? (I used version yambo-3.3.0-rev41).
RD or WR label in front of the name of the database.
Next:
1. Yes, ETStpsXd, these are the number of the steps of your integration and has to be converged.
2. EnRngeXd, this is changed by Yambo, set to the maximum interval energy of the band you included.
3. The differences you find is very big, and unexpected, are you using the same number of bands, NGsBlkXd, EXXRLvcs, bands in GbndRnge?
4. As a general remark I suggest you to switch to the new version.
Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/