How to calculate GW correction in a generic k-point

[amolina]

Hi everybody,

I have followed the tutorial http://www.yambo-code.org/doc/inputs/ypp_kk.php in order to calculate GW correction in a generic k-point.

I always find the same problem when I run the yambo setup after "yambo -i -V kpt" and it is:

<---> [WARNING]The grid is 0D but I have more than 1 kpt.
<---> [WARNING]Trying to expand the k-grid
[WARNING][RL indx] 2 equivalent points in the rlu grid found

If I ignore such warnings, in the execution I get again the same warnings plus a segmentation fault. If I check the r_setup I see this annoying thing in the
[02.04] K-grid lattice
======================

[WARNING]The grid is 0D but I have more than 1 kpt.

[WARNING]Trying to expand the k-grid

Compatible Grid is 1D
B1 [rlu]= 0.000000 0.000000 1.000000
Grid dimensions : 1
K lattice UC volume [au]: 0.272388


Has anyone had this problem before?

Thanks in advance.


Alejandro.

[claudio]

Dear Alejandro

did you set the variable IkXLim equal to the number of k-point
in the original grid? you can get this number from PWSCF/ABINIT or from the report
file r_setup
....
[RD./SAVE//ns.db1]------------------------------------------
Bands : 10
K-points : 3
G-vectors [RL space]: 10417
Components [wavefunctions]: 1308
......

also check the k-point units, I advice you to use rlu (reduced lattice units) that
in PWSCF are colled crystal while in ABINIT are the default.

Then if you still get the error message please put in the forum your input files,
including groud-state, yambo and ypp inputs.

best regards
Claudio Attaccalite

[amolina]

Dear Claudio,

yep, I check the variable IkXLim and matchs with the number of points. I have also attached the setup file that contains the warning flags.

Cheers,

Alejandro.

[claudio]

Dear Alejandro

I do not see any mistake in your input files, please put also the ABINIT input files,
in such a way we can repeat the calculation and see why it doesn't work.

Claudio

[amolina]

Dear Claudio,

at the end it works! It was a mistake in the ABINIT input file . Sorry!

Cheers,


Alejandro.

[claudio]

Ciao Alejandro

with the new version of Yambo you can also interpolate the QP band structure with the comand:

ypp -s b

Cla

[amolina]

Thanks Claudio.

Just to complete the information for beginners (like me):
ypp -s b -V qp

To refer the file db.QP where the GW eigenvalues are:

GfnQPdb= "E < ./SAVE/db.QP" # [EXTQP G] Database

[claudio]

Ciao Alejandro

the line is right:

GfnQPdb= "E < ./SAVE/db.QP" # [EXTQP G] Database

but please check the name of you quasi-particle database in the SAVE folder

if you compile Yambo with netcdf the name is ndb.QP otherwise db.QP

cla