[Pw_forum] k point selection

[Stefano Baroni]

On Jun 13, 2006, at 4:48 PM, Luke Thulin wrote:

> Hello everyone,
>
> I don't have a very good understanding of how to properly choose my  
> k points for sampling.  Therefore, I'm currently staying with  
> automatic generation.

I would rather use something to be done automatically, when I  
understand how it works. You may want to get a general idea from the  
historic paper by Monkhorst and Pack, Phys. Rev. B 13, 5188 (1976).

> To do band structure plots, is it necessary to choose the same  
> number of k points for both the scf and nscf calculations,

NO

> or can less be used for the scf?

sure: the number ok kpoints in a scf calculation is determined by the  
accuracy by which you want to calculate integrals over the 1st  
Brillouin Zone. The number of points in a nscf calculation depends on  
the points YOU are interested in. (some given symmetry directions,  
rather than others, etc.)

> If I try to enter a list of k points manually so that I can choose  
> my path, how do you determine the weights?

In a nscf calculation, weights are read, but not used, as it has been  
repeatedly noted in this mailing list. In a SCF, the weights would be  
all alike if the whole BZ was sampled. In order to save cpu time,  
only the portion of the BZ which is "irreducible" by symmetry (the  
"irreduciple wedge") is sampled, and the weight ha to be proportional  
to the number of points in the BZ which are euivalent by symmetry to  
the given k-point. Full details can be found in the above paper by  
Monkhorst and Pack.

> Thanks,
> Luke

You are most welcome. In the future, please do not forget to browse  
the archives of this mailing list before posting. I am pretty sure  
that the questions you raised have been answered many times, already.

Stefano Baroni

> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center -  
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

Please, if possible, don't  send me MS Word or PowerPoint attachments
Why? See:  http://www.gnu.org/philosophy/no-word-attachments.html



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20060613/6e442d70/attachment.html>