About BS calculation

[sdwang]

Dear developers:
I install yambo-3.2.1-r.448 without linking NetCdf.I used p2y interfaces and I calculate bulk Si optical properties.My problem isthe RPA_nlf and RPA_lf calculatios seam OK.But the BSE has some problems.When I type >yambo -b -o b -y h ,the yambo.input is:
optics # [R OPT] Optics
bse # [R BSK] Bethe Salpeter Equation
em1s # [R Xs] Static Inverse Dielectric Matrix
bss # [R BSS] Bethe Salpeter Equation solver
BSresKmod= "xc" # [BSK] Resonant Kernel mode. ('x';'c';'d')
% BSEBands
1 | 200 | # [BSK] Bands range
%
BSENGBlk= 1 RL # [BSK] Screened interaction block size
BSENGexx= 6450 RL # [BSK] Exchange components
% QpntsRXs
1 | 19 | # [Xs] Transferred momenta
%
% BndsRnXs
1 | 200 | # [Xs] Polarization function bands
%
NGsBlkXs= 1 RL # [Xs] Response block size
% LongDrXs
1.000000 | 0.000000 | 0.000000 | # [Xs] [cc] Electric Field
%
BSSmod= "h" # [BSS] Solvers 'h/d/i/t'
% BEnRange
0.00000 | 20.00000 | eV # [BSS] Energy range
%
% BDmRange
0.10000 | 0.10000 | eV # [BSS] Damping range
%
BEnSteps= 100 # [BSS] Energy steps
% BLongDir
1.000000 | 0.000000 | 0.000000 | # [BSS] [cc] Electric Field
%

And then I type >yambo -N,
it appears:
| | |/ | | | \ / \
| | | o | _ _ | o ) |
| ~ | | \_/ | | O |
|___, | _ | | | O | |
| | | | | | | |
|____/|__|__|___|___|_____|\___/

<---> [01] Job Setup
<---> [02] Input variables setup
<---> [02.01] Unit cells
<---> [02.02] Symmetries
<---> [02.03] RL shells
<---> [02.04] K-grid lattice
<---> [02.05] Energies [ev] & Occupations
<---> [03] Transferred momenta grid
<---> [04] Game Over & Game summary
There is not the output of the BSE,it seems initio output. I try >yambo -b/>yambo -N ,it also appear the same results as above.
Where is my false? Why the RPA_lf and RPA_nlf are OK?
Wy scf and nscf input are:
&CONTROL
title="2"
calculation='scf',
restart_mode = 'from_scratch',
tprnfor=.t.,
PSEUDO_DIR='/home/sdwang/espresso-4.1.2/pseudo/',
prefix='2',
outdir = '/home/sdwang/1'
/
&SYSTEM
ibrav = 1,
celldm(1) = 10.2,
nat = 2,
ntyp = 1,
ecutwfc = 20.0 ,
/
&ELECTRONS
diago_thr_init = 1.0e-6
diago_full_acc = .true.

/
ATOMIC_SPECIES
Si 28.08550000 Si.vbc.UPF
ATOMIC_POSITIONS crystal
Si 0.00 0.00 0.00
Si 0.25 0.25 0.25
K_POINTS automatic
4 4 4 1 1 1


&CONTROL
title="2"
calculation='bands',
restart_mode = 'from_scratch',
PSEUDO_DIR='/home/sdwang/espresso-4.1.2/pseudo/',
prefix='2',
outdir = '/home/sdwang/1'
wf_collect= .true.
/
&SYSTEM
ibrav = 1,
celldm(1) = 10.2,
nat = 2,
ntyp = 1,
ecutwfc = 20.0 ,
nosym = .false.
nbnd = 200,
/
&ELECTRONS
diago_thr_init = 1.0e-6
diago_full_acc = .true.

/
ATOMIC_SPECIES
Si 28.08550000 Si.vbc.UPF
ATOMIC_POSITIONS crystal
Si 0.00 0.00 0.00
Si 0.25 0.25 0.25
K_POINTS automatic
4 4 4 1 1 1


Thanks!
S D Wang

Southeast University,Nanjing,China.

[andrea marini]

I install yambo-3.2.1-r.448 without linking NetCdf.I used p2y interfaces and I calculate bulk Si optical properties.My problem isthe RPA_nlf and RPA_lf calculatios seam OK.But the BSE has some problems.When I type >yambo -b -o b -y h ,the yambo.input

Dear sdwang, what is the exact file name you get ? If you run yambo without specifing an input file it will try to read the yambo.in and not the yambo.input. TO read the yambo.input type

Code: Select all

yambo -N -F yambo.input

[myrta gruning]

Dear Shudong Wang,

It looks like yambo cannot find any input, or that it thinks it already ran the job!

About the input: either you should have a yambo.in in the directory where you are running yambo, or if you renamed the input, then you should use the -F options to specify the input file. From what you tell it seems that it should work and everything is correct. Check that the input is saved correctly in yambo.in.
Also checked in the SAVE dir. What db files are there?
I would try again (in principle you should not need the -N option)
In case it still does not work, can you post the result of ">ls SAVE"?
Also what is the system you are running on and the compiler/mpi you are using?

Regards,
Myrta

[sdwang]

Dear Shudong Wang,

It looks like yambo cannot find any input, or that it thinks it already ran the job!

About the input: either you should have a yambo.in in the directory where you are running yambo, or if you renamed the input, then you should use the -F options to specify the input file. From what you tell it seems that it should work and everything is correct. Check that the input is saved correctly in yambo.in.
Also checked in the SAVE dir. What db files are there?
I would try again (in principle you should not need the -N option)
In case it still does not work, can you post the result of ">ls SAVE"?
Also what is the system you are running on and the compiler/mpi you are using?

Regards,
Myrta

I configure yambo with :./configure --with-iotk=/home/sdwang/espresso-3.2.3/iotk/ --with-p2y=3.2 --with-mpi=/usr/local/mpich/lib/libmpichf90.a FC=ifort (when I use FC=mpif90 ,it dose not work) ,system:x86_64-redhat-linux;
>ls SAVE it appears:
db.gops db.kindx db.ostnts s.db1 s.wf wf_00001_00001 wf_00002_00001 wf_00003_00001 wf_00004_00001
If I do not use >yambo , it does not work, but >yambo -N is OK.
SDwang

[sdwang]

I install yambo-3.2.1-r.448 without linking NetCdf.I used p2y interfaces and I calculate bulk Si optical properties.My problem isthe RPA_nlf and RPA_lf calculatios seam OK.But the BSE has some problems.When I type >yambo -b -o b -y h ,the yambo.input

Dear sdwang, what is the exact file name you get ? If you run yambo without specifing an input file it will try to read the yambo.in and not the yambo.input. TO read the yambo.input type

Code: Select all

yambo -N -F yambo.input

I type >yambo -N -F BSE, but it is always "

____ ____ ___ .___ ___. .______ ______
\ \ / / / \ | \/ | | _ \ / __ \
\ \/ / / ^ \ | \ / | | |_) | | | | |
\_ _/ / /_\ \ | |\/| | | _ < | | | |
| | / _____ \ | | | | | |_) | | `--" |
|__| /__/ \__\ |__| |__| |______/ \______/

<---> [01] Job Setup
<---> [02] Input variables setup
<---> [02.01] Unit cells
<---> [02.02] Symmetries
<---> [02.03] RL shells
<---> [02.04] K-grid lattice
<---> [02.05] Energies [ev] & Occupations
<---> [03] Transferred momenta grid
<---> [04] Game Over & Game summary

[andrea marini]

Code: Select all

yambo -N -F yambo.input

I type >yambo -N -F BSE, but it is always

To my understanding "yambo -N -F yambo.input" is quite different from "yambo -N -F BSE". The name following the -F option MUST be a filename. Is "BSE" a non-empty filename ?

[andrea marini]

If I do not use >yambo , it does not work, but >yambo -N is OK.

What does it mean "it does not work" ? Seg fault, error message, nothing ? Or maybe yambo is compiled in parallel and you are trying to launch it serially on a machine where this is not allow ? The "-N" simply switches off MPI.

[sdwang]

Well ,I need to explain my process in detail.My another procedure is:
>p2y -S -N (-ls ./SAVE : s.db1 wf_00001_00001 wf_00003_00001 s.wf wf_00002_00001 wf_00004_00001)
>yambo -i
>yambo -N -F yambo.in ,it appears:
__ __ ____ ___ ___ ____ ___
| | |/ | | | \ / \
| | | o | _ _ | o ) |
| ~ | | \_/ | | O |
|___, | _ | | | O | |
| | | | | | | |
|____/|__|__|___|___|_____|\___/

<---> [01] Job Setup
<---> [02] Input variables setup
<---> [02.01] Unit cells
<---> [02.02] Symmetries
<---> [02.03] RL shells
<---> Shells finder |####################| [100%] --(E) --(X)
<---> [02.04] K-grid lattice
<---> [02.05] Energies [ev] & Occupations
<---> [03] Transferred momenta grid
<---> [RL indx] X grid is not uniform. Gamma point only.
<---> [04] Game Over & Game summary
>yambo -b -F W,in W file,I input:
em1s # [R Xs] Static Inverse Dielectric Matrix
% QpntsRXs
1 | 13 | # [Xs] Transferred momenta
%
% BndsRnXs
1 | 200 | # [Xs] Polarization function bands
%
NGsBlkXs= 200 RL # [Xs] Response block size
% LongDrXs
1.000000 | 1.000000 | 0.000000 | # [Xs] [cc] Electric Field
%

Then I type
>yambo -N -F W, it appears:
__ __ ______ ____ _____
/\ \ /\ \\ _ \ /"\_/`\/\ _`\ /\ __`\
\ `\`\\/"/ \ \L\ \/\ \ \ \L\ \ \ \/\ \
`\ `\ /" \ \ __ \ \ \__\ \ \ _ <" \ \ \ \
`\ \ \ \ \ \/\ \ \ \_/\ \ \ \L\ \ \ \_\ \
\ \_\ \ \_\ \_\ \_\\ \_\ \____/\ \_____\
\/_/ \/_/\/_/\/_/ \/_/\/___/ \/_____/

<---> [01] Job Setup
<---> [02] Input variables setup
<---> [02.01] Unit cells
<---> [02.02] Symmetries
<---> [02.03] RL shells
<---> [02.04] K-grid lattice
<---> [02.05] Energies [ev] & Occupations
<---> [03] Transferred momenta grid
<---> [04] Game Over & Game summary

My pwscf scf and nscf input are:
&CONTROL
title="2"
calculation='scf',
restart_mode = 'from_scratch',
tprnfor=.t.,
PSEUDO_DIR='/home/sdwang/espresso-4.1.2/pseudo/',
prefix='2',
outdir = '/home/sdwang/1'
/
&SYSTEM
ibrav = 1,
celldm(1) = 10.2,
nat = 2,
ntyp = 1,
ecutwfc = 20.0 ,
/
&ELECTRONS
diago_thr_init = 1.0e-6
diago_full_acc = .true.

/
ATOMIC_SPECIES
Si 28.08550000 Si.vbc.UPF
ATOMIC_POSITIONS crystal
Si 0.00 0.00 0.00
Si 0.25 0.25 0.25
K_POINTS automatic
4 4 4 1 1 1

nscf:
&CONTROL
title="2"
calculation='bands',
restart_mode = 'from_scratch',
PSEUDO_DIR='/home/sdwang/espresso-4.1.2/pseudo/',
prefix='2',
outdir = '/home/sdwang/1'
wf_collect= .true.
/
&SYSTEM
ibrav = 1,
celldm(1) = 10.2,
nat = 2,
ntyp = 1,
ecutwfc = 20.0 ,
nosym = .false.
nbnd = 200,
/
&ELECTRONS
diago_thr_init = 1.0e-6
diago_full_acc = .true.

/
ATOMIC_SPECIES
Si 28.08550000 Si.vbc.UPF
ATOMIC_POSITIONS crystal
Si 0.00 0.00 0.00
Si 0.25 0.25 0.25
K_POINTS automatic
4 4 4 1 1 1

[sdwang]

"does not work"meas when I type >yambo,it appears:
yambo: cannot access CORE database (SAVE/*db1 and/or SAVE/*wf)

[andrea marini]

__ __ ____ ___ ___ ____ ___
| | |/ | | | \ / \
| | | o | _ _ | o ) |
| ~ | | \_/ | | O |
|___, | _ | | | O | |
| | | | | | | |
|____/|__|__|___|___|_____|\___/

<---> [01] Job Setup
<---> [02] Input variables setup
<---> [02.01] Unit cells
<---> [02.02] Symmetries
<---> [02.03] RL shells
<---> Shells finder |####################| [100%] --(E) --(X)
<---> [02.04] K-grid lattice
<---> [02.05] Energies [ev] & Occupations
<---> [03] Transferred momenta grid
<---> [RL indx] X grid is not uniform. Gamma point only.
<---> [04] Game Over & Game summary

You have some problem with the grid. Yambo cannot map it in a uniform grid. Did you remove the non symmorphic symmetries? When Yambo switeches to the Gamma-point only mode several runlevels (like BSE and Static dielectric matrix) are switched off. This is way you are always dropped to the same run.