p2y crashes with nonlocal pseudo commutator option

[martin.gmitra]

Dear all,

I would like to generate yambo database with nonlocal contribution using: p2y -N -S -c
I have found the following:
(1) p2y works fine without the nonlocal option: p2y -N -S
(2) p2y crashes for Hydrogen psedopotential when invoking: p2y -N -S -c
The error message is
<---> Import structure data...done
<---> Import atoms data... ** On entry to CHEEV parameter number 5 had an illegal value

The Hydrogen PP is attached to the mail.

I would like to do study on compounds containing atoms of type Ge, Si, O, H. Perhaps it is
silly question but do you think that the nonlocal pseudo commutator would play important role
for such atoms?

Best regards,
Martin

[Daniele Varsano]

Dear Martin,
what version of yambo are you using? In the last revision yambo 3.4, the non local commutation should be taken into account without specifying any option. The non local commutator could have effects on the oscillation strength of the absorption spectrum, even if usually the effect is not so strong, clearly it is a system dependent issue.

Best,

Daniele

[martin.gmitra]

Dear Daniele,
Thanks for replay. I am running the following:
Tool: p2y 3.4.0 rev.2132
Tool: yambo 3.4.0 rev.2132
I think the -c option is not default in the rev.2132. Do you recommend to go with last revision?

Thanks,
Martin

[Daniele Varsano]

Dear Martin,
yes, try with the last revision. If this does not solve the problem, anyway it's easier for us to look if bug are present.
Best,
Daniele

[Daniele Varsano]

Dear Martin,
indeed there is a problem for the H atom, which does not have non local component. We will fix it very soon both in the stable and in devel version.
Try to update later and it should work.

Best,
Daniele

[martin.gmitra]

Dear Daniele,
Thanks for dealing with. Could you let me know where I can check in the svn tree what is the current revision (just downloaded)?
Thanks,
Martin

[Daniele Varsano]

Dear Martin,
it is displayed when you do the svn up:

Code: Select all

Updated to revision 47

This should work.
Best,
Daniele

[martin.gmitra]

Dear Daniele and developers,

Do you have some news concerning fix of the nonlocal correction for Hydrogen atom?

Best regards,
Martin

[Davide Sangalli]

Dear Martin,
indeed it was fixed with revision 47.
Do you still have problems ?

Davide

[martin.gmitra]

Dear Davide,

Thanks for your replay. I have compiled recent devel revision 48.
The p2y -N -S -c gives the following error:

[ERROR] STOP signal received while in :
[ERROR]Error in num_sphbes Bad l value ierr:  1

I think this is due to Hydrogen atom.
Thanks in advance for your help.
Best,
Martin