the projections input problem in the win file

GW, Bethe-Salpeter …

Moderators: maryam.azizi, bruneval

Locked
njuxyh
Posts: 5
Joined: Mon Apr 16, 2012 3:25 pm

the projections input problem in the win file

Post by njuxyh » Fri Jan 18, 2013 4:09 am

i am trying to calculate the band structure of graphene with the gw+wannierfunction in the abinit, after the parallel run of the dft and GW calculation, i ran a serial run of wannnier function as following the tutial.
in the *.win file, i used pz orbitals on every atom and sp2 orbitals on 'every other' atom, which is the same with example 10 in the examples directory
of the Wannier90 distribution. but the input in the begin atoms_frac block can not wirtten like in the example, the error is :
> > Begin Projections
> > C1:sp2;pz
> > C2:pz
> > End Projections
> >
> > Wannier90 is running in LIBRARY MODE
>
> > param_get_projection: Atom site not recognised c1


i put the problem in the wannier forum, a reply is
The error is caused when the labels used in the projections (C1, C2) do
> not match the labels given to the atoms. In standalone mode these are set
> in the win file, in library mode they are passed by the calling program.



and there are two possible solutions:
the first one:
You need to get the calling program to use the labels
> C1 and C2.


the second one :

>other options include - giving
> co-ordindates of the projections directly (using f= or c=) - or put a Pz
> orbital on each C atom and a 's' orbital midway between bonded atoms.

now, my question is how do i realize the first solution?

and if i used the second ways, how do i written in the *win file:

if: my system is graphene ,(a3=a3, z direction),

begin atoms_frac
C a1 a2 a3
C a4 a5 a3
begin atoms_frac
then " *put a Pz orbital on each C atom and a 's' orbital midway between
bonded atoms"*
**

begin projections
f=a1,a2,a3 :pz
f=a4,a5,a3: pz
f=(a1+a4)/2,(a2+a5)/2,a3:l=0
end projections

it is right?

you are very appreciated if you are willing to help me. Thank you very much!

Locked