[Pw_forum] eigenvalue ordering in bands structure calculation

[Paolo Giannozzi]

On May 14, 2008, at 18:47 , zx3 at rice.edu wrote:

> During calculating band structure of a zigzag band structure which has
> partial flat band at the fermi level, I found that at some k  
> points, the
> eigenvalues written by bands.x are not ordered from lower to  
> higher, it
> introduces problems when I plot using gnuplot.
>     I can modify band_plot.f90 to reorder this, but I wonder where  
> is the origin
> of this problem

it doesn't work as one would like, but it works as promised. In PP/ 
bands.f90:

!-----------------------------------------------------------------------
SUBROUTINE punch_band (filband, spin_component, lsigma, no_overlap)
   !-------------------------------------------------------------------- 
---
   !
   !    This routine writes the band energies on a file. The routine  
orders
   !    the eigenvalues using the overlap of the eigenvectors to give
   !    an estimate crossing and anticrossing of the bands. This  
simplified
   !    method works in many, but not in all the cases.

Notice the last sentence!

Paolo
---
Paolo Giannozzi, Dept of Physics, University of Udine
via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222