Abs. and Exciton of 5.2.0 output

[sdwang]

Dear developers,
I used 5.2.0(Revision 22561 Hash (prev commit) 785b4fd6f) to obtain the abs and sorted the exciton, and I found the output of >ypp -e s have negative energies, and the absorption is also wired where the low energy absorption peak is very high. I noticed in 5.2.0 the option

Code: Select all

Lkind= "BAR"                     # [BSE,X] bar(default)/full

while for 5.1.1 it is

Code: Select all

Lkind= "Lbar"                    # [BSE] Lbar (default) / full

Why it has been changed in the 5.2.0?
I never met this wired behavior in the previous version. I do not know if the problem is here.
(attached are the alpha and exciton sorted files, and I just loaded part of the E_sorted file due to the limited size)
Thanks!

Shudong

[Daniele Varsano]

Dear Shudong,
not clear what went wrong here! Can you also attach the report/input file?


Best,
Daniele

[sdwang]

Dear Daniele,
Attached are the input and report files.
Thanks!

Shudong

[Daniele Varsano]

Dear Shudong,

I can't see errors in your input.

You can try the following:
1. Use the haydock solver, you have a very large matrix and maybe something went wrong in the direct diagonalization?
2. Even if it should work, the "slab z" is usually ysed together with the W-rim method, can you try to use a different cutoff e.g. the "box"?

Daniele