[Pw_forum] local and global minimums in relaxation

Mike Marchywka marchywka at hotmail.com
Thu Dec 27 17:44:44 CET 2012





> Date: Mon, 24 Dec 2012 23:20:34 -0500
> From: correlation at tut.by
> To: pw_forum at pwscf.org
> Subject: Re: [Pw_forum] local and global minimums in relaxation
>
> I see..
> Thank you!

Well as I understand the situation, normally you have a molecule that you want to
optimize and hope your original structure is close enough that the "forces" the minimizer
sees get it to the right answer. Most molecules are not in some global minimum configuration
of atoms right? You could take atoms infinitely far apart I guess and see what you get.

Personally though I did have a related question and that is what is neglected by trying to
find a static ground state and ignoring any fluctuations? This is probably not yet
relevant compared to errors in getting static result but curious if anyone wanted to comment.




>
> Sergey
>
> On 12/21/2012 3:47 PM, Paolo Giannozzi wrote:
> > On Dec 21, 2012, at 20:01 , Sergey wrote:
> >
> >> when I do relaxation (using bfgs algorithm), I end up with the
> >> system in
> >> one of the minimums. It can be local one or it can be global one. And
> >> there is probably several other possible local minimums in the system.
> >> How can I publish such a result? As I see in the papers no one usually
> >> describe this issue. Usually it is just - 'we performed structural
> >> optimization (relaxation) and here is the result.'
> >>
> >> How to deal with this matter?
> > if you find a better way to deal with this matter, your future career is
> > guaranteed. It is a well-known problem. That's way smart people
> > have invented simulating annealing, genetic algorithms,
> > metadynamics, ...
> >
> > P.
> > ---
> > Paolo Giannozzi, Dept of Chemistry&Physics&Environment,
> > Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
> > Phone +39-0432-558216, fax +39-0432-558222
> >
> >
> >
> >
> > _______________________________________________
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://pwscf.org/mailman/listinfo/pw_forum
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
 		 	   		  



More information about the users mailing list