[gmx-users] spatial distribution function

[Christoph Freudenberger]

Hi Nguyen,

Nguyen Hoang Phuong wrote:
> you can expand the distribution function in term of spherical harmonics
> (for linear molecule) or in term of the rotation matrix (for any
> molecule). i.e., you separate the distance dependence and angles
> dependence. For distance dependent part, you can use g_rdf. For the
> angles dependent part, basically, you can do the same.
> 
> Reference: Journal of Chemical Physics 115, 7227-7234 (2001).
> 
Thanks for the reference... I must admit I didn't understand much
while browsing your paper ;-)
Did I get it right you where doing the analysis in spherical harmonics here?

Well, my molecules are not linear and I was planing to use
a rotation matrix to convert from the box frame to the molecule frame.
I'm not so sure yet wether to do the sampling of data in cartesian
or polar coordinates.
The problem is, I am rather limited in configurations to average over.
(some 10000 or so)
I figuered that I polar coordinates might be better for that, because
I could play with the binwidth's more specifically. (e.g. I don't think
I would need much resolution in the phi coordinate, but more in the theta
coordinate).
On the other hand the cartesians would come right out of
the coordinate transformations and I would not have to normalise the
volume elements... and last but not least: for visualisation with
gopenmol i would have to retransform anything back to cartesian anyway...

Maybe I should just try both options.

Could somebody help me how to normalise the volume elements in
polar coordinates?

regards
-- 
Christoph Freudenberger
Abt. Organische Chemie I AK Siehl - Uni Ulm -Tel: ++49-731-502-2785