Multiple read_data.

Dear Lammps developers,

I am reading multiple data files to merge several subsystems into one large simulation box. All the atoms are of the type full, so originally they have a molecule identifier. When Lammps (I am using lammps-31Mar17) reads the data files using the keyword add and the argument append, the atom identifier is correctly added. However, the molecule identifier does not change, and so every time new atoms are read the effect is like adding new atoms to the molecules that existed previously. Would it be possible to include also the append feature for the molecule identifier in versions to come?

Thanks.

Dear Lammps developers,

I am reading multiple data files to merge several subsystems into one
large simulation box. All the atoms are of the type full, so originally
they have a molecule identifier. When Lammps (I am using lammps-31Mar17)
reads the data files using the keyword add and the argument append, the
atom identifier is correctly added. However, the molecule identifier does
not change, and so every time new atoms are read the effect is like adding
new atoms to the molecules that existed previously. Would it be possible to
include also the append feature for the molecule identifier in versions to
come?

​yes, this is a reasonable suggestion and can be added in a quite
straightforward fashion without too much hassle. I've included it in a
pending pull request, which should be included into the next LAMMPS patch
release.
https://github.com/lammps/lammps/pull/815/commits/bba4bd148985896dde1096e4a7d1c0ad726fde8f

if desired, you can download and check out this modified version of LAMMPS
here: https://github.com/akohlmey/lammps/tree/collected-small-fixes​

axel.