Missing eqp_co_q.dat file in BGW documentation?

Eleanor Davison

unread,

Apr 7, 2025, 7:57:55 AMApr 7

to BerkeleyGW Help

Dear BGW developers,

I am attempting to run a finite-Q calculation for monolayer MoS_2 to reproduce its exciton bandstructure, but I'm getting this error in the absorption calculations:

............

Reading quasiparticle energy corrections from eqp_co_q.dat

ERROR: cannot open file 'eqp_co_q.dat' for reading: does not exist.

forrtl: error (76): Abort trap signal

............

which suggests that I need a file called eqp_co_q.dat, which I guess would be the quasiparticle energy corrections for the bands in WFNq_co (since eqp_co.dat is the corrections for the bands in WFN_co)?

I can't see this file mentioned in the BGW 4.0 manual, but I understand why the code might need it, as it does not have QP energy corrections for any k-points beyond the WFN_co grid.

Do I need to calculate this eqp_co_q.dat file for finite-Q calculations? Or should I try to get round this error in another way?

Thanks so much for your help!

Eleanor

Zhenglu Li

unread,

Apr 7, 2025, 1:01:07 PMApr 7

to Eleanor Davison, BerkeleyGW Help

Dear Eleanor,

You can use the utility inteqp.x in BerkeleyGW to generate (interpolate) the quasiparticle energies of the shift-q grid.

Best,

Zhenglu

From: Eleanor Davison <eleanor....@gmail.com>
Sent: Monday, April 7, 2025 4:57 AM
To: BerkeleyGW Help <he...@berkeleygw.org>
Subject: Missing eqp_co_q.dat file in BGW documentation?

--
You received this message because you are subscribed to the Google Groups "BerkeleyGW Help" group.
To unsubscribe from this group and stop receiving emails from it, send an email to help+uns...@berkeleygw.org.
To view this discussion visit https://groups.google.com/a/berkeleygw.org/d/msgid/help/3f6a4a5f-610c-4ea9-a2e1-9de9962eb462n%40berkeleygw.org.

Bill Ash

unread,

Apr 8, 2025, 6:27:56 AMApr 8

to BerkeleyGW Help, zhen...@usc.edu, Eleanor Davison

Dear Zheng,

I am experiencing the same problem. But when I use inteqp.x, it needs the input file named eqp_co.dat and I don't know where that file is generated from.

Looking forward to your reply, thank you!

Best,

Bill

Zhenglu Li

unread,

Apr 8, 2025, 11:53:02 AMApr 8

to Bill Ash, BerkeleyGW Help, Eleanor Davison

Dear Bill,

You need to get it from the GW calculations (epsilon and sigma steps). sigma.x will generate eqp0.dat and eqp1.dat, and you can link (typically) eqp1.dat as your eqp_co.dat for inteqp.

Best,

Zhenglu

From: Bill Ash <ashbi...@gmail.com>
Sent: Tuesday, April 8, 2025 3:27 AM
To: BerkeleyGW Help <he...@berkeleygw.org>
Cc: Zhenglu Li <zhen...@usc.edu>; Eleanor Davison <eleanor....@gmail.com>
Subject: Re: Missing eqp_co_q.dat file in BGW documentation?

Bill Ash

unread,

Apr 9, 2025, 5:48:04 AMApr 9

to BerkeleyGW Help, zhen...@usc.edu, Eleanor Davison, Bill Ash

Dear Zheng,

Thank you very much for your help. When I submit the calculation, the following error is reported:

ERROR: Missing bands in file eqp_co.dat

Here is my inteqp.inp:

number_val_bands_coarse 6
number_val_bands_fine 6
number_cond_bands_coarse 2
number_cond_bands_fine 2
use_velocity

epsilon.inp:

epsilon_cutoff 8.0
number_bands 178
cell_slab_truncation
frequency_dependence 0
begin qpoints
0.001000000 0.000000000 0.000000000 1.0 1
0.000000000 0.333333333 0.000000000 1.0 0
0.000000000 0.666666667 0.000000000 1.0 0
0.333333333 0.000000000 0.000000000 1.0 0
0.333333333 0.333333333 0.000000000 1.0 0
0.333333333 0.666666667 0.000000000 1.0 0
0.666666667 0.000000000 0.000000000 1.0 0
0.666666667 0.333333333 0.000000000 1.0 0
0.666666667 0.666666667 0.000000000 1.0 0
end

sigma.inp:

screened_coulomb_cutoff 8.0
number_bands 178
band_index_min 33
band_index_max 44
screening_semiconductor
cell_slab_truncation
begin kpoints
0.000000000 0.000000000 0.000000000 1.0
0.000000000 0.333333333 0.000000000 1.0
0.000000000 0.666666667 0.000000000 1.0
0.333333333 0.000000000 0.000000000 1.0
0.333333333 0.333333333 0.000000000 1.0
0.333333333 0.666666667 0.000000000 1.0
0.666666667 0.000000000 0.000000000 1.0
0.666666667 0.333333333 0.000000000 1.0
0.666666667 0.666666667 0.000000000 1.0
end

I would like to use the flag of exciton_Q_shift 0 0.333333333 0.333333333 0.000000000 to run a finite Q exciton calculation for WSe2. I'm using a 3*3*1 k-points grid, and generate wavefunctions on the full k-point grid. May I ask what I am wrong about?

Looking forward to your reply, thank you!

Best,

Bill

Eleanor Davison

unread,

Apr 10, 2025, 11:56:36 AMApr 10

to BerkeleyGW Help, Bill Ash, zhen...@usc.edu, Eleanor Davison

Dear Zheng,

Thank you so much for your help! Using inteqp.x worked, and my finite-Q absorption calculation now runs fine.

Hi Bill,

I had a similar error "ERROR: Missing bands in file eqp_co_q.dat". I think it happened because the inteqp.x code only interpolates QP corrections for conduction bands for eqp_co.dat, and valence bands for eqp_co_q.dat (when you use "use_velocity" in the inteqp.inp file). But I think absorption.x requires a list of QP corrections for both conduction and valence bands in eqp_co.dat and eqp_co_q.dat files, otherwise it thinks there are missing bands.

Maybe this is the cause of your error, if your eqp_co.dat only lists the 2 conduction band corrections, and your eqp_co_q.dat only has the 6 valence band corrections?

If this is the case, I managed to fix it by using the "use_momentum" flag in my inteqp.inp file (even though I am using "use_velocity" in the absorption.x calculation). It makes sure the inteqp.x code calculates QP energies for both valence and conduction bands, rather than just the conduction or valence bands.