Page 2 of 3
Re: Lkind="full"
Posted: Wed Jul 21, 2021 5:13 pm
by Xiaoming Wang
Hi Daniele,
Daniele Varsano wrote: ↑Wed Jul 21, 2021 4:29 pm
I had a look to your previous report, the only problem I spotted is:
Code: Select all
[WARNING]Wrong serial number for ./SCREEN//ndb.em1s
it seems that the screening was calculated for a different QE run, but if the code did not stop most probably it is compatible.
What's this warning about? Indeed I use the screening file from another run. Same dispersion was observed with that run with no such warning.
Best,
Xiaoming
Re: Lkind="full"
Posted: Wed Jul 21, 2021 5:26 pm
by Daniele Varsano
Dear Xiamong,
here the Q is indeed k-k' and not the q of the perturbation.
In Yambo that expression is calculated exactly as you wrote as Q it is not small and you are not in the q->0 limit.
I cannot see anything wrong in your report besides the warning on the screening that was calculated from a different set of databases.
The warning is emitted as security check, if you previous run had the very same k point grid it is harmless.
I've no clue on what is going wrong here, I will do some tests.
Best,
Daniele
Re: Lkind="full"
Posted: Wed Jul 21, 2021 6:03 pm
by Daniele Varsano
Dear Xiaoming,
another feature I noticed is that you are using a diagonal approximation for the screened interaction.
Probably this is not the reason of the problem, but can you try to comment this variable?
Best,
Daniele
Re: Lkind="full"
Posted: Wed Jul 21, 2021 11:09 pm
by Davide Sangalli
Some of the points.
a) you are right, the linear dependence can be obtained only with the Coulomb cutoff
b) Lfull result depend on Blongdir as you write, not Lbar of course
c) I think I've found an issue with time reversal. LiF has inversion symmetry and t-rev is not used, while your system has only time reversal symmetry
Try the two alternative options:
1) repeat the calculation without time-reversal
2) test the following tentative fix, with time reversal
Open file src/bse/K_exchange_collisions.F replace line 161
Code: Select all
if (i_s>nsym/(i_time_rev+1)) call dev_conjg(BS_T_grp(i_T_grp)%DEV_VAR(O_x)(:,i_T_el))
with
Code: Select all
if (i_s>nsym/(i_time_rev+1)) call dev_conjg(BS_T_grp(i_T_grp)%DEV_VAR(O_x)(:,i_T_el),range1=(/2,BS_n_g_exch/) )
Best,
D.
Re: Lkind="full"
Posted: Thu Jul 22, 2021 6:23 am
by Daniele Varsano
Right,
Lfull depends on the Blongdir for the q=0 case.
when a cut-off coulomb is used i.e. V /= 1/q^2 the G=0 term in the exchange is zero and Lfull and Lbar provide the same result for q=0.
Best,
Daniele
Re: Lkind="full"
Posted: Thu Jul 22, 2021 2:01 pm
by Xiaoming Wang
Thanks Davide and Daniele,
I have tried
1. comment out WehDiag
2. BSEmod= "coupling"
3. The second point of Davide about time-reversal
But none of them work. I'm now trying to repeat the calculation without time reversal (the first point of Davide). Will update you later.
Best,
Xiaoming
Re: Lkind="full"
Posted: Thu Jul 22, 2021 7:25 pm
by Xiaoming Wang
Finally, I get correct exciton dispersion both with LBAR and FULL by disabling the time reversal symmetry according to Davide's suggestion #1. Thanks Davide and Daniele for all your help.
Best,
Xiaoming
Re: Lkind="full"
Posted: Thu Jul 22, 2021 9:54 pm
by Davide Sangalli
Dear Xiaoming,
thank you for reaching out.
I'm a bit puzzled by the fact that the patch did not work.
Can you attach the input files of your scf/nscf simulations.
I'll try to further explore the issue.
Best,
D.
Re: Lkind="full"
Posted: Thu Jul 22, 2021 10:40 pm
by Xiaoming Wang
Hi Davide,
The scf input is:
Code: Select all
&CONTROL
calculation = 'scf'
outdir = './tmp/'
prefix = 'smpi'
pseudo_dir = '~/soft/pseudo/dojo/pbe/fr/'
verbosity = 'high'
/
&SYSTEM
ecutwfc = 80
ibrav = 0
nat = 188
ntyp = 5
occupations = 'fixed'
force_symmorphic=.TRUE.
noncolin=.true.
lspinorb=.true.
!nbnd = 600
/
&ELECTRONS
diago_full_acc = .TRUE.,
conv_thr = 1.0d-10
electron_maxstep = 80
mixing_beta = 4.0000000000d-01
/
ATOMIC_SPECIES
C 12.0107 C.upf
H 1.00794 H.upf
I 126.90447 I.upf
N 14.0067 N.upf
Pb 207.2 Pb.upf
ATOMIC_POSITIONS crystal
Pb 0.8475465178 0.2588860394 0.2487114852
Pb 0.6524534822 0.7588860394 0.7512884998
Pb 0.1524534822 0.2411139606 0.7487115002
Pb 0.3475465178 0.7411139606 0.2512884998
I 0.6686973455 0.5712988105 0.2509738439
I 0.8313026545 0.0712988105 0.7490261861
I 0.3313026545 0.9287011895 0.7509738139
I 0.1686973455 0.4287011895 0.2490261561
I 0.5416844104 0.0539155185 0.2499771565
I 0.9583155896 0.5539155185 0.7500228735
I 0.4583155896 0.4460844815 0.7499771265
I 0.0416844104 0.9460844815 0.2500228435
I 0.8717586865 0.2203576777 0.3608064039
I 0.6282413135 0.7203576777 0.6391935961
I 0.1282413135 0.2796423223 0.8608064039
I 0.3717586865 0.7796423223 0.1391935961
I 0.8668817359 0.2216595585 0.1370739739
I 0.6331182641 0.7216595585 0.8629260411
I 0.1331182641 0.2783404415 0.6370739589
I 0.3668817359 0.7783404415 0.3629260411
C 0.7931398904 0.2612185954 0.0004081331
C 0.7068601096 0.7612185954 0.9995918399
C 0.2068601096 0.2387814046 0.5004081631
C 0.2931398904 0.7387814046 0.4995918669
C 0.2162469836 0.2064471575 0.0024833829
C 0.2837530164 0.7064471575 0.9975165991
C 0.7837530164 0.2935528425 0.5024834009
C 0.7162469836 0.7935528425 0.4975166291
C 0.3165736261 0.1166452145 0.9780063757
C 0.1834263739 0.6166452145 0.0219936253
C 0.6834263739 0.3833547855 0.4780063757
C 0.8165736261 0.8833547855 0.5219936243
C 0.6792689963 0.3557412395 0.0155442407
C 0.8207310037 0.8557412395 0.9844557413
C 0.3207310037 0.1442587605 0.5155442587
C 0.1792689963 0.6442587605 0.4844557713
C 0.3632935513 0.1711966880 0.0734526921
C 0.1367064487 0.6711966880 0.9265472999
C 0.6367064787 0.3288033120 0.5734527001
C 0.8632935213 0.8288033120 0.4265472999
C 0.4399968415 0.0541099881 0.0011053514
C 0.0600031585 0.5541099881 0.9988946776
C 0.5600031585 0.4458900119 0.5011053224
C 0.9399968415 0.9458900119 0.4988946476
C 0.3949109714 0.2063106414 0.1242596258
C 0.1050890286 0.7063106414 0.8757403592
C 0.6050890286 0.2936893586 0.6242596408
C 0.8949109714 0.7936893586 0.3757403592
C 0.5756751764 0.3634358829 0.9373862951
C 0.9243248236 0.8634358829 0.0626137279
C 0.4243248236 0.1365641171 0.4373862651
C 0.0756751764 0.6365641171 0.5626137049
C 0.2392299933 0.2334000206 0.0499793498
C 0.2607700067 0.7334000206 0.9500206802
C 0.7607700367 0.2665999794 0.5499793198
C 0.7392299633 0.7665999794 0.4500206502
C 0.5707775236 0.4073609142 0.9840033188
C 0.9292224764 0.9073609142 0.0159966892
C 0.4292224764 0.0926390858 0.4840033188
C 0.0707775236 0.5926390858 0.5159966812
C 0.4631890134 0.0810151567 0.0485703299
C 0.0368109866 0.5810151567 0.9514296701
C 0.5368110166 0.4189848433 0.5485703299
C 0.9631889834 0.9189848433 0.4514296701
C 0.5457170755 0.1514304811 0.8545949660
C 0.9542829245 0.6514304811 0.1454050490
C 0.4542829245 0.3485695189 0.3545949660
C 0.0457170755 0.8485695189 0.6454050340
C 0.4132953930 0.3682287226 0.1342806520
C 0.0867046070 0.8682287226 0.8657193630
C 0.5867046070 0.1317712774 0.6342806370
C 0.9132953930 0.6317712774 0.3657193630
C 0.6902499782 0.2688188564 0.9217691308
C 0.8097500218 0.7688188564 0.0782308912
C 0.3097500218 0.2311811436 0.4217691018
C 0.1902499782 0.7311811436 0.5782308692
C 0.6943948786 0.2168087372 0.8716247516
C 0.8056051214 0.7168087372 0.1283752634
C 0.3056051214 0.2831912628 0.3716247516
C 0.1943948786 0.7831912628 0.6283752484
C 0.7994814173 0.2193072719 0.9535799914
C 0.7005185827 0.7193072719 0.0464200156
C 0.2005185827 0.2806927281 0.4535799914
C 0.2994814173 0.7806927281 0.5464200086
H 0.8754557589 0.2185495152 0.0254477774
H 0.6245442411 0.7185495152 0.9745522406
H 0.1245442411 0.2814504848 0.5254477594
H 0.3754557589 0.7814504848 0.4745522106
H 0.1222166001 0.2586624189 0.9841215677
H 0.3777833849 0.7586624189 0.0158784173
H 0.8777833849 0.2413375811 0.4841215677
H 0.6222166151 0.7413375811 0.5158784323
H 0.2990994990 0.0986187594 0.9406701876
H 0.2009005010 0.5986187594 0.0593298314
H 0.7009005310 0.4013812406 0.4406701576
H 0.7990994690 0.9013812406 0.5593298124
H 0.6752391104 0.3868987628 0.0523266536
H 0.8247608896 0.8868987628 0.9476733284
H 0.3247608896 0.1131012372 0.5523266716
H 0.1752391104 0.6131012372 0.4476733574
H 0.5210168103 0.9864521619 0.9821148508
H 0.9789831897 0.4864521619 0.0178851372
H 0.4789831897 0.5135478381 0.4821148508
H 0.0210168103 0.0135478381 0.5178851492
H 0.4984054365 0.1483709904 0.1354064026
H 0.0015945635 0.6483709904 0.8645935824
H 0.5015945345 0.3516290096 0.6354064176
H 0.9984054655 0.8516290096 0.3645935824
H 0.4875612906 0.4014873516 0.9133885400
H 0.0124387094 0.9014873516 0.0866114450
H 0.5124387094 0.0985126484 0.4133885400
H 0.9875612906 0.5985126484 0.5866114600
H 0.1602705128 0.3047012387 0.0686151014
H 0.3397294872 0.8047012387 0.9313849136
H 0.8397295172 0.1952987613 0.5686150864
H 0.6602704828 0.6952987613 0.4313849136
H 0.4793171430 0.4788696640 0.9957226616
H 0.0206828570 0.9788696640 0.0042773344
H 0.5206828870 0.0211303360 0.4957226616
H 0.9793171130 0.5211303360 0.5042773384
H 0.5615603633 0.0335621677 0.0662143259
H 0.9384396367 0.5335621677 0.9337856971
H 0.4384396367 0.4664378323 0.5662143029
H 0.0615603633 0.9664378323 0.4337856671
H 0.7214050842 0.3169943013 0.8053347960
H 0.7785949158 0.8169943013 0.1946652040
H 0.2785949158 0.1830056987 0.3053347960
H 0.2214050842 0.6830056987 0.6946652040
H 0.6852138445 0.4393897917 0.8476858381
H 0.8147861555 0.9393897917 0.1523141769
H 0.3147861555 0.0606102083 0.3476858381
H 0.1852138445 0.5606102083 0.6523141619
H 0.8581805847 0.3592283711 0.8437612028
H 0.6418194153 0.8592283711 0.1562388122
H 0.1418194153 0.1407716289 0.3437612028
H 0.3581805847 0.6407716289 0.6562387972
H 0.2720457927 0.0309454701 0.1528855329
H 0.2279542073 0.5309454701 0.8471144371
H 0.7279542073 0.4690545299 0.6528855629
H 0.7720457927 0.9690545299 0.3471144671
H 0.2882142756 0.1725065495 0.1896970059
H 0.2117857244 0.6725065495 0.8103029941
H 0.7117856954 0.3274934505 0.6896970059
H 0.7882143046 0.8274934505 0.3103029941
H 0.1613997905 0.1812595454 0.1459763726
H 0.3386002095 0.6812595454 0.8540236274
H 0.8386002095 0.3187404546 0.6459763726
H 0.6613997905 0.8187404546 0.3540236274
H 0.5548340624 0.1154573637 0.8180669137
H 0.9451659376 0.6154573637 0.1819330563
H 0.4451659376 0.3845426363 0.3180669437
H 0.0548340624 0.8845426363 0.6819330863
H 0.5190780232 0.0552309360 0.8758769424
H 0.9809219768 0.5552309360 0.1241230576
H 0.4809219768 0.4447690640 0.3758769424
H 0.0190780232 0.9447690640 0.6241230576
H 0.4506904794 0.2283614435 0.8573409878
H 0.0493095206 0.7283614435 0.1426589832
H 0.5493095506 0.2716385565 0.3573410168
H 0.9506904494 0.7716385565 0.6426590122
H 0.3090842331 0.4297632716 0.1266381401
H 0.1909157669 0.9297632716 0.8733618449
H 0.6909157369 0.0702367284 0.6266381551
H 0.8090842631 0.5702367284 0.3733618449
H 0.4441101247 0.3878316830 0.1709994655
H 0.0558898753 0.8878316830 0.8290005495
H 0.5558898453 0.1121683170 0.6709994505
H 0.9441101547 0.6121683170 0.3290005495
H 0.5047171824 0.4119710971 0.1122976050
H 0.9952828176 0.9119710971 0.8877023720
H 0.4952828176 0.0880289029 0.6122976280
H 0.0047171824 0.5880289029 0.3877024020
H 0.7865815119 0.1367111100 0.8673986603
H 0.7134184881 0.6367111100 0.1326013697
H 0.2134184881 0.3632888900 0.3673986303
H 0.2865815119 0.8632888900 0.6326013397
H 0.8897471464 0.1470120762 0.9414068673
H 0.6102528536 0.6470120762 0.0585931067
H 0.1102528536 0.3529879238 0.4414068973
H 0.3897471464 0.8529879238 0.5585931327
N 0.7410779150 0.3420051996 0.8400956813
N 0.7589220850 0.8420051996 0.1599043337
N 0.2589220850 0.1579948004 0.3400956813
N 0.2410779150 0.6579948004 0.6599043187
N 0.2700972318 0.1449656419 0.1550537314
N 0.2299027682 0.6449656419 0.8449462836
N 0.7299027682 0.3550343581 0.6550537164
N 0.7700972318 0.8550343581 0.3449462836
K_POINTS automatic
4 4 1 0 0 0
CELL_PARAMETERS angstrom
8.7848061275 0.0000000000 0.0000000000
0.0000000000 9.1711784478 0.0000000000
0.0000000000 0.0000000000 28.5782799514
For nscf calculation, you'll need change calculation = 'nscf', uncomment "!nbnd = 600", and change k mesh to 16x16x1.
Best,
Xiaoming
Re: Lkind="full"
Posted: Wed Sep 01, 2021 11:54 am
by Davide Sangalli
Dear Xiaoming,
sorry few more questions:
Xiaoming Wang wrote: ↑Thu Jul 22, 2021 7:25 pm
Finally, I get correct exciton dispersion both with LBAR and FULL by disabling the time reversal symmetry according to Davide's suggestion #1. Thanks Davide and Daniele for all your help.
1 - removing the t-rev, did it change only the dispersion with Lkind="full" or also with Lkind="Lbar"?
2 - is the epsilon obtained with "Lbar" the same as the one obtained with "full"?
Best,
D.