From: Philipp Schoen1 (
Date: Thu May 10 2007 - 05:16:57 CDT

Dear Users,
at the risk of boring you with a comment about water density calculation
again I would like to put the following question for discussion.
I wrote a program to calculate the density of water. My results deviate
from experiments by about <5% , i.e. approx. 0.975 g/cm3 @ 300K which is
quite ok.
For the calculations I select water by
set vol [atomselect 0 "water and same residue as $zlowerbound <= z and z <
$zupperbound" frame $i].
The "same residue" ensures that the boundaries are "weak", i.e. single
atoms of a molecule can be out of the range so that the domain size

I track the thus occuring fluctuations in the volume for all frames and
calculate the density with the average out of that.
I would like to ask the list members if this is a reasonable approach or do
you think I got nice values accidently?

Cheers and thanks in advance