[gmx-users] g_analyze - negative parameters

[Berk Hess]

Hi,

Your data might have intrinsically strange time correlations,
but the chances are high that you have not sampled enough.
I would first also try the g_analyze of Gromacs 4.0, I have improved
some subtleties in the fitting weights there, which can help in certain
unstable cases.
But tau1 seems extremely long, or does your input have even longer times?

Berk

> From: maximilian.andrews at tu-dortmund.de
> To: gmx-users at gromacs.org
> Date: Mon, 1 Feb 2010 19:49:17 +0100
> Subject: [gmx-users] g_analyze - negative parameters
> 
> Dear Gromacs-Users,
> 
> I've been trying to figure out g_analyze for a while now (having read
> the manual references and papers), but I still haven't found out what
> a negative parameter obtained during fitting means. It's clear that
> when Tau2 is longer than the total simulation time the sampling is
> bad, but I don't know how to interpret a negative parameter and how it
> affects the sampling, as can be seen from the output excerpt obtained
> from the following command (in Gromacs 3.x):
> 
> g_analyze -f data.xvg --ee data_ee.xvg
>  
> a fitted parameter is negative
> invalid fit:  e.e. 0.9166  a 0.0376076  tau1 -21618.1  tau2 383059
> Set   1:  err.est. 61.1951  a 1  tau1 1.63958e+09  tau2 0
> 
> I'm interested in the average value and the error estimate, but I
> obviously would like to know if my data is good. I guess that since
> Tau2 is set to zero, I only have the short time correlation, and not the
> long time one.
> 
> Thanks in advance for your time!
> 
> Best regards,
> 
> Maximilian
> -- 
> Dortmund University of Technology
> Department of Chemistry
> Physical Chemistry I  -  Biophysical Chemistry
> Otto-Hahn-Str. 6
> D-44227 Dortmund
> Germany
> 
> Office:   C1-06 room 176
> Phone:  +49 231 755 3916
> Fax:    +49 231 755 3901
> 
> -- 
> gmx-users mailing list    gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
 		 	   		  
_________________________________________________________________
New Windows 7: Simplify what you do everyday. Find the right PC for you.
http://windows.microsoft.com/shop
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20100202/d162a686/attachment.html>