reset_ids command

LAMMPS LAMMPS Mailing List Mirror
1

Dear lammps users,

I’ve wanted to delete overlapped molecules and used the command “delete_atoms”, as follows.

Intialization

units real
dimension 3
boundary p p p
atom_style full

read_data water_a.dat ~~~

replicate 3 4 4

read_data AAA.dat add append offset 2 1 1 0 0 group AAA

2

Dear lammps users,

​[...]

I can't use compress keyword of delete atoms since I use molecular atom

style.
So. I tried to use "reset_ids" command

But I got an error on both linux and Mac envrionments.
ERROR: Unknown command: reset_ids (../input.cpp:246)

I checked my lammps is the lastly updated version.

​which version is that exactly?

have you looked at this page?

http://lammps.sandia.gov/bug.html​

Could somebody let me know the way to solve this problem or another way to
reset atom IDs?

​my updated executable does have the reset_ids command:

$ ~/compile/lammps/src/lmp_mpi
LAMMPS (20 Apr 2018)
  using 1 OpenMP thread(s) per MPI task
reset_ids
ERROR: Reset_ids command before simulation box is defined
(../reset_ids.cpp:33)
Last command: reset_ids
​
​axel.​

3

Thank you so much, Axel.

I checked my version is the latest stable version Mar 16 2018 and the command was newly added at Mar 20, 2018.
I understand I can use it when I update my lammps to the version.

However, was there a similar way to renumber atoms before the version release?
How could renumber atoms with molecular atom style?

Thanks,
Mook

4

No, before you had to write a script to remap (and keep track of all bonds) ids.

Anders

5

Thanks, Anders and Alex.

Dear someone who developed “reset_ids” command.

I found the command doesn’t work for the case Tip4p water model which requires a rule of consecutive atom ID order for O and H. If you updated the command, it would be really helpful to many lammps users.

Thank you so much.
Mook.

6

Thanks, Anders and Alex.

Dear someone who developed "reset_ids" command.

I found the command doesn't work for the case Tip4p water model which
requires a rule of consecutive atom ID order for O and H. If you updated
the command, it would be really helpful to many lammps users.

​it would be helpful to *you*. i don't think there are many LAMMPS users
that need it, or else people would have asked for it already. don't assume
that what you are doing is what many LAMMPS users are doing.
in fact, i would opine, that the cleaner and more flexible way to "tailor"
systems for LAMMPS is through using external tools, e.g. VMD and the
topotools plugin or moltemplate. doing this kinds of modeling operations
are in strong conflict with the needs for an efficient parallel program.​
so in most cases, system and especially topology building and modifying
options in LAMMPS should be seen as a last resort method, not the preferred
solution.

beyond that, ​if you want somebody to improve the code or fix an issue, you
have to first provide a simple and easy to reproduce test case. it is also
highly advisable to upload those together with a sufficiently detailed
description to the LAMMPS project issue tracker on github: ​
https://github.com/lammps/lammps/issues

axel.