Density of States Queries

[danielk]

In the VASP manual it states that "The density of states (DOS) , is actually determined as the difference of the integrated DOS between two pins". If this is the case then I'm a bit confused as to why my DOSCAR file seems to be missing a peak, where the integrated density of states increases by 0.7 from one pin to the next, but nothing registers on the DOS curve. I could increase the energy resolution but the increase in the integrated DOS suggests that this should be unnecessary. Eg. from my DOSCAR:

1.290 0.0000E+00 0.0000E+00 0.1440E+03 0.1440E+03
1.299 0.0000E+00 0.0000E+00 0.1450E+03 0.1440E+03

The number of spin up electrons/states has increased by one.

I'm also a bit puzzled by the units. In this case the integrated number of electrons/states increases by 0.7 and the size of the peak seems to indicate that there are 5 electrons/states eg:

0.859 0.0000E+00 0.0000E+00 0.1460E+03 0.1460E+03
0.903 0.1330E+03 0.0000E+00 0.1467E+03 0.1460E+03

where 133 * 0.044 ~ 5.8.

Are these related? Am I missing something about the units, or about the method of calculation of DOS.

Thank you in advance for your comments.

[admin]

1) you are right, the DOS is only evaluated on the NEDOS energy grid points. Hence, if the integrated dos jumps up without showing a peak in the DOS, please either increase NEDOS, or decrease the energy interval for which the DOS in calculated (EMIN,EMAX), to increase teh peak resolution.
2) probably the peak width is much smaller than the energy interval width.