Multiplicity not specified

[Megha Anand]

Thanks Conrad! 
The absolute energy of Fe atom from GAPW computation in CP2K differ from 
that of G09 by ~0.1 hartree. But, I agree with you that it would make sense 
to compare the reaction energies across the two programs.



On Tuesday, September 13, 2016 at 6:00:57 AM UTC-4, Conrad wrote:
>
> Hi Mega,
>
> The atomic code is there to give you your initial wavefunction. 
> In principle, because your are then self-consistently seeking the 
> wavefunction in 
> the main quickstep code, it should be independent of the initial guess. 
> There's going 
> to be exceptions to this, of course, but my feeling is that the difference 
> is coming 
> from somewhere else. 
>
> You mention that you're comparing energies with Guassian  -
> this only makes sense if you're looking at reaction energies or similar, 
> as Matt
> mentioned in your other thread.
>
> Best wishes,
> Conrad
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20160913/089fb146/attachment.htm>