Re: pairlist violations

From: Thomas Brian (thomasbrianxlii_at_gmail.com)
Date: Tue Jul 10 2012 - 15:15:07 CDT

thanks for your inputs everyone.

I found the section of documentation which states that pairlist violations
do not affect the dynamics. From
http://www.ks.uiuc.edu/Research/namd/wiki/?NamdPerformanceTuning:

"...the original pairlist distance determines the patch size but is
automatically tuned during the simulation and pairlist violations reduce
performance but do not result in missed interactions"

Thus, namd must keep track and reverse a time step or something, unless I
am mis-interpreting the statement.

On Tue, Jul 10, 2012 at 12:25 PM, JC Gumbart <gumbart_at_ks.uiuc.edu> wrote:

> I'm not sure about this - it seems NAMD does some pretty smart things to
> avoid such violations:
> http://www.ks.uiuc.edu/Research/namd/2.9/ug/node39.html
>
> See the *outputPairlists* option.
>
> On Jul 10, 2012, at 5:24 AM, Axel Kohlmeyer wrote:
>
> On Tue, Jul 10, 2012 at 1:44 AM, Thomas Brian <thomasbrianxlii_at_gmail.com>
> wrote:
>
> Hi namd-l,
>
>
> If an atom enters another atoms cutoff without being in its pairlist, what
>
> error/warning, if any is displayed in the output file?
>
>
> no. such a test would *kill* performance.
> it is up to the user, to choose reasonable
> settings, so that this does not happen.
>
> Also, does namd back up and redo the last "cycle" number of timesteps? (I
>
>
> no.
>
> saw somewhere that pairlist violations do not affect the dynamics and I
>
> don't know how that would be possible unless namd backed up and redid
>
> calculations)
>
>
> well, if you are missing interactions due to too infrequent pairlist
> rebuilds
> to too small a cutoff for it, then you *do* miss those interactions. if
> this is
> infrequent, however, you may not notice. whether this is a good or bad
> thing, is a different discussion.
>
> axel.
>
> Thanks,
>
> Tom
>
>
>
>
> --
> Dr. Axel Kohlmeyer akohlmey_at_gmail.com http://goo.gl/1wk0
> International Centre for Theoretical Physics, Trieste. Italy.
>
>
>

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:22:14 CST