My Community

You can use p4vasp to plot the bandstructure from the vasprun.xml

if you are looking for examples for the most typical problems you might want to calculate,
please have a look at the examples and excercises given at the VASP workshop
held in spring 2003:
you can download the
-- tutorials of the given lectures (.pdf) and the
-- inputs of all excercises (.tgz)
from our web-site:
http://cms.mpi.univie.ac.at/vasp-workshop

concerning the visualisation of VASP results (structures, DOS, bandstructures), it is recommended to use p4vasp, a visualisation
tool developed by O. Dubay. Please have a look at the p4vasp site.
You can reach it via the institute's homepage or directly,
http://cms.mpi.univie.ac.at/odubay/p4vasp.html

I'd like to second the request for more information about how to calculate band structure. The workshop examples are good on many other topics, but for band structure, all of the important details are missing. I assume they were explained to participants in the workshop, but they are missing from the slides. Which files need to be copied from the self-consistent run? Where are the bands? in EIGENVAL? The example files are in different directories; which files are important, and which ones get overwritten? Sorry if these are dumb questions from a newbie.

I get a reasonable band structure from a 2-atom Si unit cell, but not from an 8-atom supercell.

By the way, if anybody's interested, I made a quick EIGENVAL parser in Perl to convert the bands to a format which you can plot with gnuplot or any spreadsheet. (p4vasp still won't compile on Macs. This was easier.)
<span class='smallblacktext'>[ Edited Sun Nov 20 2005, 12:33PM ]</span>

GNUPLOT also works, it has some clever features, plots columns of data. It is free to download from http://www.gnuplot.info/

+it works on mac ...

[quote="wistey"]I'd like to second the request for more information about how to calculate band structure. The workshop examples are good on many other topics, but for band structure, all of the important details are missing. I assume they were explained to participants in the workshop, but they are missing from the slides. Which files need to be copied from the self-consistent run? Where are the bands? in EIGENVAL? The example files are in different directories; which files are important, and which ones get overwritten? Sorry if these are dumb questions from a newbie.

I get a reasonable band structure from a 2-atom Si unit cell, but not from an 8-atom supercell.

By the way, if anybody's interested, I made a quick EIGENVAL parser in Perl to convert the bands to a format which you can plot with gnuplot or any spreadsheet. (p4vasp still won't compile on Macs. This was easier.)
[/quote]

for a bandstructure run:
1) the ionic and electronic structures have to be converged.
2) continue with CONTCAR and CHGCAR from the last step
(set ICHARG=11) and set up a KPOINTS file like shown in the
VASP-workshop handson examples (subdirectories
2_3_fccSi_band or 3_3_Ni100clean_BAND)
3) for plotting, either use the p4vasp tool, or some utility which
prepares and plots the information of the EIGENVAL file
4) concerning the band structures of supercells: please mind
the (un)folding of the Brillouin zone!

I am also facing the same type of problem. p4vasp is not working for bandstructure plotting. How can I covert the EIGENVAL file so that I can view it through GNUPLOT

[quote author=distance(kpointN,kpoint1)
), which can then easily be plot by grace or gnuplot.

(in case you have a combination of high symmetrylines in 1 calculation eg: X->Gamma-> L, then you will also need your KPOINTS file to get the neccesry information to create your k-point coordinate)

Danny
(PS: the above will work perfectly for a simple calculation along 1 line (of symmetry), If you have more some tweaking is needed to get your k-point coordinates correct...also I don't know wat happens to the EIGENVAL file when spin is introduced, but I gues the above can easily be adjusted to that)
<span class='smallblacktext'>[ Edited Sat Oct 21 2006, 03:40PM ]</span>

just to complete: in the EIGENVAL file, you find

# of atoms (twice), NBLOCK*KBLOCK, #spins
V(cell), | lattice vectors |[m], POTIM*10^-15
Temperature
'CAR'
the header of your files
#of electrons, #of k-points, #of bands
and for all k-points:

k-point coordinates, weight if the k-point
band index, eigenvalue (for spin-polarized systems, there are 2 columns of eigenvalues, for spin up and down respectively.


please make use of the special input format of the KPOINTS file to generate k-points along the high-symmetry lines of the BZ in order to plot a band structure, as shown for Si on page 12
of the handsonII.pdf VASP-workshop tutorial handout
(http://cms.mpi.univie.ac.at/vasp-workshop)

[quote="wistey"]I'd like to second the request for more information about how to calculate band structure. The workshop examples are good on many other topics, but for band structure, all of the important details are missing. I assume they were explained to participants in the workshop, but they are missing from the slides. Which files need to be copied from the self-consistent run? Where are the bands? in EIGENVAL? The example files are in different directories; which files are important, and which ones get overwritten? Sorry if these are dumb questions from a newbie.

I get a reasonable band structure from a 2-atom Si unit cell, but not from an 8-atom supercell.

By the way, if anybody's interested, I made a quick EIGENVAL parser in Perl to convert the bands to a format which you can plot with gnuplot or any spreadsheet. (p4vasp still won't compile on Macs. This was easier.)
[/quote]

[quote="wistey"]I'd like to second the request for more information about how to calculate band structure. The workshop examples are good on many other topics, but for band structure, all of the important details are missing. I assume they were explained to participants in the workshop, but they are missing from the slides. Which files need to be copied from the self-consistent run? Where are the bands? in EIGENVAL? The example files are in different directories; which files are important, and which ones get overwritten? Sorry if these are dumb questions from a newbie.

I get a reasonable band structure from a 2-atom Si unit cell, but not from an 8-atom supercell.

By the way, if anybody's interested, I made a quick EIGENVAL parser in Perl to convert the bands to a format which you can plot with gnuplot or any spreadsheet. (p4vasp still won't compile on Macs. This was easier.)
[/quote]

I am also trying to covert the EIGENVAL file so that I can view it through gnuplot. If it is possible could you please share your prgogram that you made to convert the mentioned file?

Hi,
I would like to find out if information on the kpoints in the OUTCAR provides the same details on the bandstructure as the EIGENVAL.

Thank you.

you need the self-consistent charge density (CHGCAR, ICHARG=11 in INCAR) and of course INCAR,POSCAR, POTCAR and an appropriate KPOINTS file to get the band structure.
the relevant data to plot a band structure are EIGENVAL (k-points, eigenvalues) and PROCAR (the 'character' of the bands, based on local charges).

please read the online manual as well for further information (eg. chapter examples).

[quote="admin"] just to complete: in the EIGENVAL file, you find

# of atoms (twice), NBLOCK*KBLOCK, #spins
V(cell), | lattice vectors |[m], POTIM*10^-15
Temperature
'CAR'
the header of your files
#of electrons, #of k-points, #of bands
and for all k-points:

k-point coordinates, weight if the k-point
band index, eigenvalue (for spin-polarized systems, there are 2 columns of eigenvalues, for spin up and down respectively.


please make use of the special input format of the KPOINTS file to generate k-points along the high-symmetry lines of the BZ in order to plot a band structure, as shown for Si on page 12
of the handsonII.pdf VASP-workshop tutorial handout
(http://cms.mpi.univie.ac.at/vasp-workshop)

[/quote]

Does the CAR line mean that the coordinates supplied are in cartesian coordinates, or are the coordinates for every k-point in reciprical coordinates? (or however you spell it