[gpaw-users] pseudo and all-electron density

[Ask Hjorth Larsen]

Hi

2012/7/5 Brandon Cook <brandon.g.cook at vanderbilt.edu>:
> usesymm=None doesn't seem to change the result. I get these numbers with or
> without it
>
> nelectrons 32.0
> ae-den 112.0
> pseudo-den 34.8535926774
> sum wfs    33.3655212891
>
> -Brandon

nelectrons is the correct number of valence electrons in the system (4
per atom and 8 atoms).

ae-den integrates correctly to the total number of electrons (14 per
atom for Si).

The pseudodensity is not guaranteed to contain any specific number of
electrons, but typically isn't too far from the number of valence
electrons unless you go lower into the periodic table.

The small discrepancy between the two last numbers should equal the
integral of the fixed pseudo-core density (called "ñ-core" in the PAW
formalism).

Regards
Ask