SHG spectrum

[sunxl]

Dear all,
When I calculated the SHG at different frequencies after calculating the collision, for example, I calculated 30 frequency points in total, and calculated them in parallel in groups of five. During the calculation process, sometimes I encountered the following situation: at one point, the calculation just stopped, such as F14. This would then affect the subsequent calculations of F19, F24, and F29. However, the calculation would still show as completed in the end. But when I deleted these interrupted calculations and resubmitted them after the calculation was completed, the system seemed not to recognize the missing output files and defaulted to thinking that the calculations had been completed. Does this affect the calculation results? How can these missing files be completed?

o-shg-bse1-2-01D.external potential F11 936KB 2025/10/22,10:15
o-shg-bse1-2-01D.external potential F12 878KB 2025/10/22,10:15
o-shg-bse1-2-01D.external potential F13 911KB 2025/10/22,10:15
o-shg-bse1-2-01D.external potential F14 367KB 2025/10/22,8:12
o-shg-bse1-2-01D.external potential F15 910KB 2025/10/22,10:15

[Daniele Varsano]

Dear sunxl,

I'm not sure I understood correctly your problem.
My suggestion is to recalculate the missing points redirecting the outputs (-J) using a different directory of the previous one.

Best,
Daniele

[sunxl]

Dear Daniele,

Thank you for your patient response.
When using the -J option to create a new directory, how should I set it up so that it only calculates the missing points? Isn't this equivalent to starting the calculation all over again?

Best,
sunxl

[claudio]

Dear sunxl

if you change directory using -J you have to copy all ndb.Nonlinear* files in that directory,
so Yambo will find them and not recalculate.

Anyway you can do also separate runs for different energy range, number of points and then joint
the file response functions

best
Claudio