GW band structure behaviour not same as PBE

[Taranga]

Dear Developers,
I am writing to express my concern regarding the GW band structure calculation I recently performed using yambo. While the calculation completed without errors, the resulting band structure appears to deviate significantly from the expected behavior, particularly when compared to PBE calculations. I have attached the relevant log files, report files, and a comparative diagram of the GW and PBE band structures for your reference. Please let me know if any additional information is required.

Sincerely,
Taranga

[Daniele Varsano]

Dear Taranga,

this seems a problem of the interpolation and not of the GW calculation itself.
Can you share the ypp.in file you used to interpolate the band structure?

Additionally, I strongly recommend to you use the "slab z" geometry for the coulomb cutoff instead of the box.

Best,
Daniele

[Taranga]

Dear Daniele,
Thank you for the reply. I have attached the ypp_bands input file.

[Daniele Varsano]

Dear Taranga,

I had a look at your input file and I cannot find anything wrong, I do not know what is causing problem in the interpolation.
I suggest you to try to plot the QP bands structure using the yambopy utility. Here you can find a tutorial:

https://wiki.yambo-code.eu/wiki/index.p ... structures

Best,
Daniele

[Taranga]

Hi,
While trying to plot the GW band structure using yambopy I have faced some errors. The errors are:

Traceback (most recent call last):
File "/Users/apple/Downloads/databases_yambopy/bands_gw.py", line 27, in <module>
ydb.plot_scissor_ax(ax,n_top_vb)
File "/Users/apple/Downloads/databases_yambopy/yambopy/.venv/lib/python3.12/site-packages/yambopy/dbs/qpdb.py", line 200, in plot_scissor_ax
shift,cslope,vslope,cintercept,vintercept=self.get_scissor(valence,verbose=verbose)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/apple/Downloads/databases_yambopy/yambopy/.venv/lib/python3.12/site-packages/yambopy/dbs/qpdb.py", line 166, in get_scissor
vslope, vintercept, r_value, p_value, std_err = stats.linregress(e0,eqp)
^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/apple/Downloads/databases_yambopy/yambopy/.venv/lib/python3.12/site-packages/scipy/stats/_stats_py.py", line 10891, in linregress
raise ValueError("Inputs must not be empty.")
ValueError: Inputs must not be empty.

I am not able to figure out what is happening. I have attached the bands.py file.

One more question. How do I choose the 'n_top_vb' variable?

[palful]

Dear Taranga,

It seems that the function computing the scissor operator, which needs the energies of Kohn-Sham and quasiparticle states, is not finding the correct inputs. I cannot understand why without looking at your ns.db1 and ndb.QP files.

Anyway, if your interest is simply to plot the band structure and not to calculate a scissor operator, you can simply remove section 1 from that script.

As for the variable n_top_vb, this is the index of the highest occupied band in your system (starting from 0 because of python counting).

Cheers,
Fulvio