AW: Minimization doesn't stop when converged

From: Norman Geist (norman.geist_at_uni-greifswald.de)
Date: Fri Oct 07 2011 - 04:54:47 CDT

Hi,

thanks for your interest. Well I just wondered because the criteria for a
successful minimization is that the total energy of the system doesn't
fluctuate anymore (mostly). Namd could see that during runtime as it prints
out the energies. I know that feature from the Amber package, where the
minimization stops then and one can save computing time, with not running a
minimization where nothing will change for hours or days or even weeks
(depending on system size). Especially when running a new system, one can
make weak expectations for the time needed for the minimization, often too
high, sometimes too short also. If the minimization could decide on its one
when it is finished, that would be nice, wouldn't it?

Well that was just a question for my interest.

thanks

Norman Geist.

-----Ursprüngliche Nachricht-----
Von: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] Im Auftrag
von Boyang Wang
Gesendet: Freitag, 7. Oktober 2011 10:56
An: Norman Geist
Cc: namd-l_at_ks.uiuc.edu
Betreff: Re: namd-l: Minimization doesn't stop when converged

Hi Norman,

during minimization, there is no momentum or temperature, so namd is always
searching.

Best,
Boyang

On Fri, Oct 7, 2011 at 4:48 PM, Norman Geist <norman.geist_at_uni-greifswald.de
> wrote:

> Hi,****
>
> I would just like to ask if there’s a special reason why namd doesn’t have
> the feature to break the minimization when the total energy gets constant=

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:57:48 CST