[Q-e-developers] problem with PAW + stess + Hubbard U

[Davide Ceresoli]

You're right, and this is why I hesitated modifying the '>='. In this
particular pseudopotential, both two l=2 orbitals are bound. I will
change UPF occupations by hands.

Thanks,
     Davide



On 01/20/2014 02:52 PM, Kucukbenli Emine wrote:
> Ciao
> as far as I remember during generation of a PP, occupation < 0 is reserved to indicate unbound states,
> while you can have a bound state with occupation zero by selecting the reference energy=0.d0 and occupation=0.d0.
> One could imagine systems and pseudopotentials where Hubbard-relevant channel has indeed has zero occupations in the pseudo. So I would not change that if clause.
> Instead, I would put a negative occupation for any unbound state, as it would be done during generation.
> But again I might be wrong..
> cheers
> emine