[Wien] Georg K. H. Madsen and Pavel Novak method

[Gavin Abo]

username at computername:~/wien/wiendata/NiO/NiO_+.50$ grepline :3dd001 
'NiO.scf' 2 | head -2 | tail -1
  NiO.scf::3DD001: 3D*                  0.109619395 Ry [1]

username at computername:~/wien/wiendata/NiO/NiO_+.50$ grepline :3dd001 
'NiO.scf' 2 | tail -1
  NiO.scf::3DD001: 3D*                  0.142149349 Ry

username at computername:~/wien/wiendata/NiO/NiO_+.50$ grepline :3dd001 
'NiO.scf' 2
  NiO.scf::3DD001: 3D*                  0.109619395 Ry
  NiO.scf::3DD001: 3D*                  0.142149349 Ry

I could be wrong, but check your :log file, you will likely see in it:

(x) lcore -up
(x) lcore -dn

It may be that the 0.109619395 Ry (or similar value that you got) comes 
from the "x lcore -up" while the 0.142149349 Ry comes from the "x lcore 
-dn".  Since the equation in section V of [2] has epsilon_3d_up, it may 
be that the values are needed from the "-up" instead of the "-dn".

[1] 
http://wien2k-algerien1970.blogspot.com/2016/08/how-to-calculate-u-parameter-in-nickel.html
[2] http://susi.theochem.tuwien.ac.at/reg_user/textbooks/Constraint_U.pdf

On 3/5/2018 4:00 PM, karima Physique wrote:
> Dear wien2k users:
>
> According to Georg K. H. Madsen and Pavel Novak method for the 
> estimation of Ueff, they use the penultimate eigenvalue of d orbital d 
> (head -2 | tail -1), I would be grateful if you can explain to me why 
> they did not take the last value (tail -1)?
> Thank you in advance

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20180305/387bfa59/attachment.html>