I am currently exploring a 2x2x2 supercell of CH3NH3PbI3 to look at polaron behavior. However, the vasprun.xml file is too large for p4vasp to plot the band structure and DOS plots (500GB+). Are there other options for looking at the band structure to see if there are any localized band gap states?? I know I can do the brute force method of looking at the EIGENVAL file, but there must be a faster way. Thanks in advance for any and all help.
Hi Eric,
Indeed 500 GB is a very large file. Yes I think you can analyze the EIGENVAL file with a python script to plot it, but as you say its going to be quite time consuming. Have you try to reduce (to a certain degree) your KPOINTS-mesh, or also to inspect just the ENERGY band region you are interested in. I mean with the wavecar file try to inspect small regions so you have smaller files.
EMIN = -5.0
EMAX = 5.0
NBANDS = (aprox1.5 *#electrons)
LORBIT = 11
I hope it helps you my idea.
Hi Eric,
if you're interested in the DOS, you can also take a look at the DOSCAR and plot that using python or similar. I think it's easier than analyzing the EIGENVAL file. Link to vaspwiki:
https://cms.mpi.univie.ac.at/wiki/index.php/DOSCAR
Best,
Kai
Kai,
i also want to look at the band structure to calculate the rashba parameter due to spin orbit coupling ...so I need the bands around the band gap ...when I open the vasprun.xml file in p4vasp I usually just save to .dat and dump the data in excel to post processes these values for small structures, but the files are too large to open in p4vasp now for my supercell structures ...I know there is a way to calculate certain bands only using IBAND but I’m not sure if that would help me here
Dear Eric,
another option is to use pyProcar, https://github.com/romerogroup/pyprocar
this package has a 'filter' function, with that you can remove unneeded data, or group several rows/columns/blocks of data into a single entry. decreasing the size of the PROCAR file up to ~200 MB or less.
I found a different solution that works very well ...that is, if you have access to Matlab. So, there is a set of functions called VASPLAB (found here: https://www.mathworks.com/matlabcentral/fileexchange/36836-vasplab), and a specific function called import_eigenval. If you have your EIGENVAL file in the working Matlab directory along with the import_eigenval.m file, then all you have to type on the command line is the following:
[energy, kpoints] = import_eigenval('EIGENVAL')
plot(energy, 'k')
This will give you a nice band structure and then you can modify as you see fit. In the newer Matlab version (2018b and later) you can run your mouse pointer over the plot and it will display the given points on the plot (which is exactly what I need to calculate the Rashba paramters!!) . Let me know if you have any issues with my method here and I'll see if I can help. Thanks to everyone who answered my initial question.
- Badges
- Science topic
- Similar topics
- Psychology