Obtaining atomic energies of each atom

80 views
Skip to first unread message

인원여

unread,
Nov 22, 2018, 3:41:02 AM11/22/18
to aenet
Hello,

I want to get the each atomic energy of the test structures. 

In prediction using predict.x, however, the atomic energies are summed up and there is only total energy in the output

I was wondering if there is any way I can obtain the atomic energies of each atom.

Best,
In Won Yeu

Nongnuch Artrith

unread,
Nov 22, 2018, 7:34:30 AM11/22/18
to 인원여, ae...@googlegroups.com
Dear In Won Yeu,

Note that the absolute values of the atomic energies are not unique in compositions with more than one chemical species.  So you should be careful with interpreting the atomic energies directly - they are probably not meaningful.  This is why the energies are not typically written out by 'predict.x'.

However, if you really want to take a look at the atomic energies, 'predict.x' supports an experimental input file keyword called 'PRINT_ATOMIC_ENERGIES' that activates output of the atomic energies.  This parameter has not been well tested, so please double-check if the reported energies add up to the correct total energy.

Please let us know if you have any further questions.

Best,
Nong


--
You received this message because you are subscribed to the Google Groups "aenet" group.
To unsubscribe from this group and stop receiving emails from it, send an email to aenet+un...@googlegroups.com.
To post to this group, send email to ae...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Message has been deleted

인원여

unread,
Nov 24, 2018, 3:20:27 AM11/24/18
to aenet
Dear Nong,

Thank you for your very helpful reply.

The atomic energies was well written out by using the keyword and the reported energies added up to the predicted total energy.

Best,
In Won Yeu

 
--
You received this message because you are subscribed to the Google Groups "aenet" group.
To unsubscribe from this group and stop receiving emails from it, send an email to aenet+unsubscribe@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages