[ERROR]Fractional e/h occupations found

[wlyim]

Dear all,

My computed system is metallic by DFT and becomes semiconducting after GW correction. But even I applied the scissor operator by setting the KfnQP_E parameter in BSE calculation, the job crashed because of the fractional e/h occupations. I guess that the program still took the original occupations by pure DFT. Am I right? And how can I fix this problem?

Best regards,
William

[claudio]

Dear William

you can use the Bethe-Salpeter without the Tamm-Dancoff approximation
that should work with any occupation.
See the tutorial on SiH4

http://www.yambo-code.org/tutorials/SiH4/index.php

best regards
Claudio