Dear All,
I would like to use fix store/state (initialized with ave/atom) in my own fix
which needs info about neighbors of atoms.
It seems to me that fix store/state does not hold the info about ghost atoms at the moment.
At least when i access its fields for atoms with "index>nlocal" it gives me zeroes.
I'm not sure about the behavior of ave/atom.
Could you possibly guide me on which parts of fixes
should be closely looked at and modified to keep the info about ghost atoms ?
Perhaps this would be desireble by other users.
p/s/ i would rather not re-invent the wheel by creating another custom fix(es) which
function as the mentioned two fixes with addition of taking care of info at ghost atoms.
Regards,
Denis.
Dear All,
I would like to use fix store/state (initialized with ave/atom) in my own fix
which needs info about neighbors of atoms.
It seems to me that fix store/state does not hold the info about ghost atoms at the moment.
At least when i access its fields for atoms with "index>nlocal" it gives me zeroes.
I'm not sure about the behavior of ave/atom.
Could you possibly guide me on which parts of fixes
should be closely looked at and modified to keep the info about ghost atoms ?
Perhaps this would be desireble by other users.
there is a big principal problem with this request:
which atoms are ghost atoms depends on many parameters and changes
during an MD. so storing their information directly is pointless. you
would have to reconstruct the information of the ghost atoms for their
corresponding local atoms in the other domains.
axel.
Axel is correct. I don’t think “storing info about
ghost atoms” would have whatever effect you
are hoping it would. You probably need to
explain the motivation for what you are trying to do.
Steve