Voronoi calculation for a particular group

Dear LAMMPS developers and users,

Greetings! I am using the VORONOI package to calculate the Voronoi tesselation, and it appears that all the atoms are considered. Say, I am only interested Voronoi volume of the O atoms in water, is it possible to count only O’s? Thank you for the attention.

Sincerely,

LC Liu

No that is not yet possible during the simulation (of course you can
always do it in post processing).
You have to keep in mind that the voronoi volume can quickly become
ill-defined when you start to pick and choose participating atoms. Do
you just have water molecules in your simulation?
That being said, modifying the voronoi compute to only take into
account the atoms in the group is trivial. I can add an option for
that. But not before Sunday.
Cheers,
Daniel

Nah, well, here is a quick fix to the last compute voronoi version
Add the option 'only_group' to your voronoi compute and it will only
send the atoms in the compute group to the voronoi calculation.
I.e. create a group for your oxygen atoms and do

compute 1 oxygen voronoi/atom only_group

Keep in mind that you've now effectively reduced the density of
voronoi atoms. Read the docs carefully to understand the implications!
You may have to increase the communication cutoff to get correct
tesselations near the boundaries of your simulation domain!
Best,
Daniel

compute_voronoi_atom.cpp (11.8 KB)

compute_voronoi_atom.h (1.77 KB)

Dear LAMMPS developers and users,

Greetings! I am using the VORONOI package to calculate the Voronoi
tesselation, and it appears that all the atoms are considered. Say, I am
only interested Voronoi volume of the O atoms in water, is it possible to
count only O's? Thank you for the attention.

how about defining a group that has your oxygen atoms and assigning
the compute to an atom style variable while multiplying it with
gmask()?

axel.

That will not solve the problem at all. Dude, read the manual!

That will not solve the problem at all. Dude, read the manual!

yeah, perhaps i should make it a rule to not answer questions before i
had my morning coffee...