Solid Si tutorial data processing

[AnOnion]

I am trying to reproduce the plots of the Indirect and Direct band gaps of solid silicon as done in the GW tutorial.

My build of Yambo does not have netcdf support, so I have done my own DFT calculations in QE 5.0.3 using the input files and pseudopotentials provided (scf, then nscf). I am able to reproduce the yambo reference r-* files for the most part, with the difference that I also receive some "cc" units in adition to the "iru" and "iku" units. I have attached both my r- file and a reference one.

In the script included with the tutorial (HF_gap.sh), grep is used to extract numerical data from files which are not created by following the procedure at the tutorial link above or in the Tutorial.info file. These files have names like "o-HF_7Ry.HF", which leads me to believe that these files are created with ypp. However, I cannot figure out how to actually do this, and there are not reference o-* files for this step included in the tutorial archive.

How would I use ypp to process the data and extract what I need to calculate the band gaps, as done in the tutorial?

[Daniele Varsano]

Dear JP Ninno,
yes there are some confusion we need to fix, we will do it very soon.
The tutorial was done with a slightly different version of the code, anyway we are almost ready to release a new version of the code, where a lot of thing are changed so we will update all the tutorials according with the new release.
Now, the o.* file are produced for COHSEX and GW, but not for the HF calculation. So in order to check the HF gaps, you need to look at the report files r-*
I invite you to follow the tutorial analyzing the results by inspecting the report and output files without the use of the scripts, or alternatively by writing your own script.
Ypp is not meant for this kind of analysis.
You can type ypp -H in order to see the ypp features.
Sorry for the inconvenience and we will fix it soon.
Kind regards,

Daniele

[Daniele Varsano]

Dear JP Nimmo,
if you download the last stable version of che code from svn, it should now be compatible for the scripts in the tutorial.
Best,
Daniele

[AnOnion]

Thank you very much. I have written for access to qe-forge. In the mean time, what numbers do I subtract (reading from the script) to find the band gap? (i.e., what goes in to the variables d_HF_v, d_HF_c1, d_HF_c2 in the script?)

I'm only a chemist, and am trying to do some sanity checking before I start to do "real" science!

[Daniele Varsano]

They are the HF correction to the DFT eigenvalues for the valence band minimum and conduction band maximum.
This is done for the direct and indirect gap.
Gap=E^Qp_cbm-E^Qp_vbm

E^QP=Eo+<Sigma_HF>-<Vxc>

Sigma_HF is indicated in the report as <HF>.
In the output (once you download from the svn) you should have directly the corrections.
Best,
Daniele