Thanks a lot Dr. Elias Assmann and Mr. Kadda Amara for your valuable 
suggestions. I am running the calculations to fix the problem. 
Best Regards,,,   

    On Friday, 18 December 2015, 17:07, Elias Assmann <> 

 On 12/18/2015 08:27 AM, Adam Bakheet wrote:
> I am running a bandstructure calculations on doped Cax(PO4)y compound.
> After x spaghetti I got a distorted bandstructure plot. I checked the
> case.spaghetti_ene, and the last two columns of some bands appeared
> merged together.
> Please how can fix this problem? Your help is appreciated.

That sounds like the energy numbers got “too large” for the Fortran
format they are printed with (or rather, so large that they only just
fit into the format).  On my installation, a line from
case.spaghetti_ene looks like this:

  0.41296  0.41296  0.41296  0.00000 -48.30905

So apparently you have three-digit negative energies (or even four-digit
positive ones!).  If this is correct, some suggestions:

1. Do you really want those large energies?  Otherwise, just change the
energy range.

2. Change the format in the spaghetti source to include a space between
the fields.

3. Parse the file in a way so as to account for the fixed format, e.g.
using a standard Unix tool (see ‘man cut’):

  cut --output-delimiter=\  -c 1-40,41- case.spaghetti_ene

4. Dirty hack: Run spaghetti in Rydberg mode instead of eV.


Elias Assmann
Institute of Theoretical and Computational Physics
TU Graz                  ⟨⟩

Wien mailing list

Wien mailing list

Reply via email to