PDOS eigenvalues near Fermi level are not sorted properly

[Alex Vozny]

OK, I see.
Yes, this is OT, and, apparently, the system is metallic, at least in this 
unrelaxed configuration.
Working with fractional occupation is a pain in CP2K, or at least it 
provides no benefits/speedup compared to other programs.
Alex.

On Friday, April 7, 2017 at 5:26:46 PM UTC-4, Matt W wrote:
>
> Hi,
>
> is your SCF fully converged? Is your system (very close to) metallic?
>
> If you are running OT, it calculates the ground state density first, then 
> at the pdos states finds LUMOs that are orthogonal to the occupied states. 
> This is saying that it finds some LUMOs lower in energy than occupied 
> states.
>
> While we are at it, is this system similar (same?) to the one you reported 
> with inconsistency between pdos and cubes?
>
> Matt
>
>
> On Friday, April 7, 2017 at 3:52:05 PM UTC+1, Alex Vozny wrote:
>>
>> Here is an example:
>>     1819         -0.096060        2.000000        0.00228579       
>>  0.06867136        0.00357453
>>     1820         -0.093844        2.000000        0.00815732       
>>  0.01590175        0.00444991
>>     1821         -0.093705        2.000000        0.00936543       
>>  0.01842093        0.00464511
>>     1822         -0.092709        2.000000        0.00117320       
>>  0.00508035        0.00225085
>>     1823         -0.092626        2.000000        0.00317422       
>>  0.00800068        0.00255608
>>     1824         -0.091870        2.000000        0.00503839       
>>  0.00999660        0.00359990
>>     1825        * -0.098011*        0.000000        0.03037013       
>>  0.02025094        0.00472899
>>     1826         -0.096724        0.000000        0.01318675       
>>  0.04356782        0.00400355
>>     1827         -0.096528        0.000000        0.01638755       
>>  0.04630115        0.00407204
>>     1828         -0.093383        0.000000        0.00480571       
>>  0.00915687        0.00292543
>>     1829         -0.092029        0.000000        
>>
>> Why does this happen sometimes?
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20170408/e5b22fab/attachment.htm>