Hi,
On 04/03/2015 04:07 PM, Jason Swails wrote:
> On Thu, Apr 2, 2015 at 4:49 PM, Brian Radak <brian.radak.accts.gmail.com>
> wrote:
>
>> Is there a reason the write to disk has not been moved? Is it just legacy
>> purposes?
>
>
> I'm not sure if that's the only reason, but it's certainly a
> consideration. This has been the behavior for the better part of 2 decades
> as far as I know (possibly longer).
As far as I understand, it's a way of having mdcrd and restrt files in
sync: if you set ntwr=ntwx, then these files are in sync. You can't have
both the energy and the restart file in sync, since the restart file
contains the next step (i.e., you restart by computing energy and
forces, not by integrating the step).
Gerald.
>
>
>
>> I don't understand the specific utility in *not* having the
>> energies match. I know there is a slight difficulty with the kinetic energy
>> (which is a result of leapfrog), does that confuse things?
>>
>
> You can get the velocities at the same timesteps as the coordinates by
> averaging the half time-step before and after (and this is what is done to
> make sure that the kinetic energy reported is given for the same time step
> as the energies are in the mdout file.
>
> You lose a little accuracy in the kinetic energy by doing this, but I think
> the error is still only of the order of dt^-3 (maybe dt^-2, I'd have to
> review the leapfrog derivation).
>
> All the best,
> Jason
>
--
____________________________________________________________________________
Prof. Gerald MONARD
SRSMC, Université de Lorraine, CNRS
Boulevard des Aiguillettes B.P. 70239
F-54506 Vandoeuvre-les-Nancy, FRANCE
e-mail : Gerald.Monard.univ-lorraine.fr
tel. : +33 (0)383.684.381
fax : +33 (0)383.684.371
web : http://www.monard.info
____________________________________________________________________________
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Apr 03 2015 - 09:00:03 PDT
