[gmx-users] Regarding binning technique

[Manik Mayur]

Hi,
I have a doubt regarding binning technique to generate density profile and
streaming velocity. I followed Allen and Tildesley and use the following
formulae to calculate density and velocity profiles. (Consider monoatomic
species)

1) No. density of atoms in a bin 'i' = Total no. of atoms in bin 'i' over
all time/(bin volume*total time steps)
2) velocity in bin 'i' = sum over time steps((Total sum of velocities in bin
'i'/total no. of atoms in bin 'i'))/total time steps

The thing that I am observing is that upon plotting, the profile changes
with the no. of bins(e.g. no. density at x=a is different for #bins=m and
#bins=n). Which is valid, but is there any mathematical way of arriving at
the optimum no. of bins by some "bin convergence test" analogous to "grid
convergence test" used in CFD. That is. no. density at x=a doesn't vary much
with the #bins.

If you have faced the same problem, how do you resolve it?

Thanks and regards,

Manik Mayur
Graduate student
Microfluidics Lab
Dept. of Mechanical Engg.
IIT Kharagpur
INDIA
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090713/9ffd6c99/attachment.html>