Re: about Hydrogen Bond correlation function

From: Hannes Loeffler (
Date: Thu Nov 04 2010 - 08:11:16 CDT

On Thu, 4 Nov 2010 13:55:28 +0800
Jun Zhang <> wrote:

> Hello everyone:
> I would like to calculate the HB correlation function according to:
> c(t)=<h(0)h(t)>/<h>
> in "Nature,379:55-57". However, I have been convoluted by the meaning
> of <>: Is it an average over all atom pair at a particular time or
> over the entire trajectory for a particular pair? I prefer the second
> definition, but I am not sure. Please help me. Thank you in advance.

I guess what you are asking is whether the angle brackets denote an
ensemble average or a time average. But in the context of the paper
cited above wouldn't you want to average over all water molecules at
some stage anyway (obviously assuming that all water molecules are
equivalent)? In other words, time average or ensemble average first or
the other way round? I would think, mathematically speaking, that
should not make any difference (or am I wrong?). Eventually, think
about the situation where you have only a single HB pair.


This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:54:44 CST