about the first step in LR

[deyulu]

Dear Yambo developers:
I have a naive question about running Yambo. After creating the database, we need to run "yambo"
to set up db.gops and db.kindx files under SAVE. Suppose I did first the ground state
calculation with a small number of bands. Later, I want to include more bands in the
ground states, and set up the calculation for epsilon^-1. After I regenerate the s.wf and
s.db1, do I need to rerun "yambo"? Or can I recycle the previous db.gops and db.kindx
obtained with less bands? It seems to me that those files contain the information about the
shells of G vectors, but I don't know if they also depend on the number of bands.
Thanks
Deyu Lu

***************************************************************************
Deyu Lu (Ph.D)

190 Chemistry Building
University of California, Davis
One Shields Avenue
Davis, CA 95616
Office phone: (530) 754-9663
Group Webpage: http://angstrom.ucdavis.edu/

***************************************************************************

[Daniele Varsano]

Dear Deyu Lu,
I suspect that Yambo will complain about inconsistencies in the header of the databases.
But I'm not sure, I never did it, why do you want to do this?? The Yambo inizialisation it
is not an heavy part and I suggest you to run it after having calculated your ground state
and the empty states you need for the epsilon.
Cheers,
Daniele

[andrea marini]

I have a naive question about running Yambo. After creating the database, we need to run "yambo"
to set up db.gops and db.kindx files under SAVE. Suppose I did first the ground state
calculation with a small number of bands. Later, I want to include more bands in the
ground states, and set up the calculation for epsilon^-1. After I regenerate the s.wf and
s.db1, do I need to rerun "yambo"? Or can I recycle the previous db.gops and db.kindx
obtained with less bands? It seems to me that those files contain the information about the
shells of G vectors, but I don't know if they also depend on the number of bands.

The exchange of databases between different SAVE folders is a delicate issue. By default it is not possible as the interfaces p2y and a2y assign to db1/wf/kb_pp databases a unique serial number (if you do yambo -D you will see it)

Code: Select all

>yambo -D
[RD./SAVE//ns.db1]------------------------------------------
 Bands                           : 10
 K-points                        : 3
 G-vectors             [RL space]:  359
 Components       [wavefunctions]:  359
 Symmetries       [spatial+T-rev]:  48
 Spinor components               : 1
 Spin polarizations              : 1
 Temperature                 [ev]: 0.000000
 Electrons                       : 8.000000
 WF G-vectors                    :  359
 Max atoms/species               : 2
 No. of atom species             : 1
- [b]S/N 005606[/b] ---------------------------- v.03.00.12 r.000 -

The S/N assigned to the core databases is then exported to ALL databases created. Most of these databases must respect the S/N. In the sense, if you overwrite the core databases yambo should automatically overwrite the database that do not respect the S/N. For example, db.gops and db.kindex are S/N dependent.

[deyulu]

Thank you for the answer. I see the point.

Deyu