[gmx-users] g_energy averages

[David L. Bostick]

Okay thanks..

so I guess I want to ask from my pasted previous example.. though perhaps
it cannot be classified as an "error."  In the sense that the average
surface tension <#Surf*SurfTen> is taken over 250000 steps even if
nstenergy is 500, then the standard deviation of the mean of this is

2263.84/sqrt(250000)

and not

2263.84/sqrt(500) .. i.e. 500 = 250000/500

for brevity.. I just want to be sure that g_energy takes ALL of the steps
for averaging or just the subset that you specify with nstenergy.  From the
manual, I understand that ALL of them are used.  I just want validation on
this.

Thanks
David

> > If I get from the energy file
> > Energy                      Average       RMSD     Fluct.
> > #Surf*SurfTen               147.805    2270.69    2263.84
> >
> > and the run was over 250000 steps, the standard error is
> >
> > 2263.84/sqrt(250000)



-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
David Bostick					Office: 262 Venable Hall
Dept. of Physics and Astronomy			Phone:  (919)962-0165
Program in Molecular and Cellular Biophysics
UNC-Chapel Hill
CB #3255 Phillips Hall				dbostick at physics.unc.edu
Chapel Hill, NC 27599	           		http://www.unc.edu/~dbostick
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

On 17 Jul 2002, David wrote:

> On Wed, 2002-07-17 at 20:15, David L. Bostick wrote:
> >
> > .. So, if I want to find the standard error in an average value, I just
> > take the fluctuation and divide it by sqrt(nsteps)?
> >
> > i.e.
> > If I get from the energy file
> > Energy                      Average       RMSD     Fluct.
> > #Surf*SurfTen               147.805    2270.69    2263.84
> >
> > and the run was over 250000 steps, the standard error is
> >
> > 2263.84/sqrt(250000)
> No you should divide by the sqrt of the number of *independent* data
> points. For energy, there is a correlation on the order of a few
> picoseconds (depends on system). For e.g. the dipole moment of the box
> it is usually 10 ps (can be related to Debye relaxation). g_analyze may
> help you estimating errors, but it is not trivial...
>
>
> --
> Groeten, David.
> ________________________________________________________________________
> Dr. David van der Spoel, 	Biomedical center, Dept. of Biochemistry
> Husargatan 3, Box 576,  	75123 Uppsala, Sweden
> phone:	46 18 471 4205		fax: 46 18 511 755
> spoel at xray.bmc.uu.se	spoel at gromacs.org   http://zorn.bmc.uu.se/~spoel
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>