Page 1 of 1

py4vasp phonon band visualization

Posted: Sat Sep 21, 2024 6:43 am
by michael_walkup

Hello,

I am plotting phonon bands of phthalocyanine (H20PBc) and am using py4vasp to pull the data from the vaspout.h5 file. I'm unfamiliar with Jupyter, so I initially used Spyder and a pandas database to plot the phonon bands. I then found I couldn't look at phonon modes of specific atoms, so I plotted the same data in Jupyter using the regular py4vasp commands (specifically, mycalc.phonon_band.to_plotly()). When selecting individual atoms, the bands in the plot turn from simple lines into wide, translucent bands. With the large number of bands, this makes the details in the plot difficult to discern. What is the significance of this visualization? I have attached images of the same data in Spyder, in Jupyter, and one more in Jupyter selecting a carbon atom.

-Michael


Re: py4vasp phonon band visualization

Posted: Mon Sep 23, 2024 6:01 pm
by martin.schlipf

The width of the line corresponds to the contribution of specific atoms to a phonon mode. If you have a system with different atom types you would typically see that the heavier atoms contribute more to the lower frequency modes. If the modes are very close so that all these lines overlap you can play with the width argument to change the strength of the broadening effect.

As an alternative for complex systems, it is often helpful to look at the phonon DOS to investigate the atom contributions. By loosing the qpoint dependence you have a more compact representation of the information so it is possible to distinguish the atomwise contributions.