wfc_preprocess option in the PYXAID tutorial step2

[Kyungtae Lee]

Dear PYXAID users

I'm following the PYXAID tutorial of solid pentacene. As show in the tutorial, we can get H(ij) and H'(ij) plots for both real and imaginary parts at the step 2.  

But I can get only H(ij) plots, because Hprime files in the res directory are filled with zeros. 

I tried several tests and found out that I was able to get non-zero Hprime files and correct H'(ij) plots by setting "wfc_preprocess=normalize" instead of "restore" in the submit.pbs file .

However, H(ij) plots in this setting are inconsistent with the tutorial. 

I ended up running the step 2 calculation twice, one for H(ij) with "wfc_preprocess=restore" option and another for H'(ij) with "wfc_preprocess=normalize" option. 

Am I doing this correctly? Any comments would be really appreciated.

[brendanqhd]

Hi Kyungtae Lee,

Are using using Kpoints = gamma or Kpoints = automatic ? One would use the restore when doing the former. Try using Kpoints = gamma with "wfc_preprocess=restore", with the flag for the computation of Hprime set equal to 1.

Best,

Brendan

[Yoonhoo Ha]

Hi Kyungtae Lee,

I also encountered the same problem. I am not sure of this, but I got a contour plot similar to the tutorial by changing "x_re", "y_re", "z_re" parts of Hprime to "x_im", "y_im", "z_im" in py-scr6.py.

Best regards,

Yoonhoo Ha

[Kyungtae Lee]

Thank you so much, Brendan and Yoonhoo.

I was able to fix the problem by changing the py-scr6.py file according to Yoonhoo's suggestion.

Best,

Kyungtae