quaternions

[Pierre de Buyl]

Hi all,

I have started to use quaternions in molecular simulations some time ago.

I believe that it is a worthwhile addition. LAMMPS and ESPResSo use quaternions
as well and could store them in H5MD.

I intend to file a H5MD enhancement proposal but would like feedback. The
storage scheme would be simple: instead of the cartesian coordinates, the last
index of the dataset would run over the quaternion components w, x, y, z. This
order is used internally by LAMMPS, by the Python library rowan. My own code
uses x, y, z, w so I'll change it :-(

The name of the data would be "quaternion".

Regards,

Pierre

-- 
-----------------------------------------------------------
Pierre de Buyl
KU Leuven - Institute for Theoretical Physics
T +32 16 3 27355
W http://pdebuyl.be/
-----------------------------------------------------------