bad normalization of dos_values
The DOS values stored in dos_values
appear to be multiplied by the cell volume before normalization.
E.g. for this FHI-aims calculation:
first non-zero value of dos_values
: 1.2101768523571676e-16
cell_volume
: 6.924703842336816e-29
first non-zero value from file: 0.00000028 eV
To find the original value from file, one can calculate:
value_from_file = dos_values
/ cell_volume
* electron charge
The units of dos_values
are then [states/J/m^3].
Therefore, after normalization the unit of dos_values_normalized
is [states/J/atom], which is different to what is stated in the Metainfo.
From random observations (not referenced here), this behavior seems to be the same regardless of the parser. So I assume it happens somewhere between parsing and normalization.