From chemistry-request- at -server.ccl.net Wed Sep 12 05:10:12 2001
Received: from fecit-gate.fecit.co.uk (firewall-user- at -[193.119.134.250])
	by server.ccl.net (8.11.0/8.11.0) with SMTP id f8C9AB327677
	for <chemistry /at\ccl.net>; Wed, 12 Sep 2001 05:10:11 -0400
Received: by fecit-gate.fecit.co.uk; id KAA04052; Wed, 12 Sep 2001 10:21:46 +0100
Received: from suisei(192.168.101.50) by fecit-gate.fecit.co.uk via smap (V5.0)
	id xma004048; Wed, 12 Sep 01 10:21:11 +0100
Received: from fecit.co.uk (NT11.fecit.co.uk [192.168.101.116]) by suisei.fecit.co.uk (8.6.9/8.6.9) with ESMTP id JAA03752; Wed, 12 Sep 2001 09:50:01 +0100
Message-ID: <3B9F264F.A6CC627F #at# fecit.co.uk>
Date: Wed, 12 Sep 2001 10:09:35 +0100
From: Herbert Fruchtl <fruechtl*- at -*fecit.co.uk>
Organization: Fujitsu European Centre for Information Technology
X-Mailer: Mozilla 4.77 [en] (WinNT; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Antonio Morreale <antonio.morreale&$at$&uah.es>
CC: chemistry /at\ccl.net
Subject: Re: CCL:Two close energy states
References: <3B9E3844.2020102 $#at#$ uah.es>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

It usually doesn't mean that there are two minima, but that your
optimization algorithm, in its choice of step size and direction, has
run into an endless loop. If the difference is too big to just call it
converged, you should discard your approximate Hessian and restart from
one of these points. If you can afford it, calculate a complete Hessian
first (possibly on a lower level of theory). It might help to change the
optimization algorithm, most programs offer more than one.

Antonio Morreale wrote:
> When doing some optimization, the energy values flutuates between two
> close values alternatively. As far as I know, it seems like there are
> two close energy states and the optimization algorithm is not able to
> choose for one. Is there any way to solve this problem?
>