[gmx-users] Fwd: density error bars

[XAvier Periole]

You may have wanted to have sent the message to the gmx list!

XAvier.

Begin forwarded message:

> From: nicoletta liguori <fisicotta at gmail.com>
> Date: July 26, 2011 5:28:29 PM MDT (CA)
> To: x.periole at rug.nl
> Subject: density error bars
>
> Hi,
> I'm using Gromacs and its tools to sample some kind of membranes and  
> characterize their features.
> Actually I don't know what to do about error bars.
> For example when I calculate the density mass distribution along the  
> axis parallel to the normal to the
> membrane surface I average on all trajectories using g_density_d on  
> my set contained
> in the .xtc file.
> Considering that, for what I understood, it makes the average on  
> time of the density, but without
> adding any kind of error (there are no possibilities with this tool)  
> I tried to check if there are other tools
> available to calculate the error bars but seem not to exist!!
> For example g_analyze reads directly the already averaged set of  
> data (in this case) cause the only
> possible input is a .xvg file!
> So I was thinking to (not yet know how, is my first computational  
> trial) to calculate the error with a simple
> standard deviation (I was thinking to use matlab ), but to do this,  
> I suppose that
> I have to consider all the points indipendent, and to do this
> I'd have to look at the autocorrelation function of density itself!
> but how to do this??
> there are again no tools available
> how do you do when u deal with errors?
> Thanks for attention and for the help
> Nicoletta

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20110726/423d3b92/attachment.html>