[gmx-developers] Tool to estimate the error for SPME

[Florian Dommert]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/27/2010 01:10 PM, Carsten Kutzner wrote:
> Hi,
> 
> On Aug 27, 2010, at 12:14 PM, Florian Dommert wrote:
> 
> On 08/27/2010 11:38 AM, hess at sbc.su.se wrote:
>>>> Hi,
>>>>
>>>> Does this already contain the change (wrt the previous g_tune_pme
>>>> error estimate) of not using the initial coordinates, but a uniform
>>>> distribution?
>>>>
> 
> Hi,
> 
> not so far. I am thinking about the best way to sample the box
> uniformly without meeting the PME grid nodes. My current thoughts are to
> sample the box in (k+1)(k-1) points in every dimension, where k is the
> number of grid points in the respective direction. However I have to
> test if this is enough or too much.
> 
> I put the current code in gromacs to make it easier to Carsten to access
> it and adapt g_tune_pme if desired. At the beginning we wanted to
> incorporate the tool into g_tune_pme, but as g_pme_error is MPI enabled
> problems arise, when an MPI program starts a further MPI program.
> 
> As soon as possible I will implement this and contact Carsten to discuss
> how and if g_pme_error will be incorporated into g_tune_pme.
>> A convenient way could be to call g_pme_error from g_tune_pme and
>> then let g_tune_pme parse the output file of g_pme_error. This way the
>> error estimate could run in parallel on as many processors the user
>> has at hand for the tuning run.
> 
>> Carsten
> 

Yes, this is exactly the way I also was thinking about.

> 
> 
> /Flo
> 
>>>> Berk
>>>>
>>>> Hello,
>>>>
>>>> finally I have prepared a tool g_pme_error that allows to estimate the
>>>> error of SPME and tunes ewald_rtol to minimize the error. It is MPI
>>>> enabled and contains a claim for citing the corresponding reference
>>>> where the error estimate is described.
>>>>
>>>> Are you interested to put this into the master branch of gmx ?
>>>>
>>>> /Flo
>>>>
> --
> gmx-developers mailing list
> gmx-developers at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-developers
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-developers-request at gromacs.org.
>>>>>
> 
- -- 
gmx-developers mailing list
gmx-developers at gromacs.org
http://lists.gromacs.org/mailman/listinfo/gmx-developers
Please don't post (un)subscribe requests to the list. Use the
www interface or send it to gmx-developers-request at gromacs.org.

> --
> Dr. Carsten Kutzner
> Max Planck Institute for Biophysical Chemistry
> Theoretical and Computational Biophysics
> Am Fassberg 11, 37077 Goettingen, Germany
> Tel. +49-551-2012313, Fax: +49-551-2012302
> http://www.mpibpc.mpg.de/home/grubmueller/ihp/ckutzne






- -- 
Florian Dommert
Dipl.-Phys.

Institute for Computational Physics

University Stuttgart

Pfaffenwaldring 27
70569 Stuttgart

Phone: +49(0)711/685-6-3613
Fax:   +49-(0)711/685-6-3658

EMail: dommert at icp.uni-stuttgart.de
Home: http://www.icp.uni-stuttgart.de/~icp/Florian_Dommert
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEUEARECAAYFAkx3pcYACgkQLpNNBb9GiPmw4wCfWq6q+vCz1JlWR+j9y9yld1HI
N94AmMfpmNMUzRdPAnI7K1MJUXYog7g=
=BRzM
-----END PGP SIGNATURE-----