Transferring restart file to new machine

LAMMPS users,

I have read on the doc page that restart files are not always compatible across machines and it might be necessary to convert to a date file.

In my script, I used the command:

write_data datafile

and it writes out a datafile with the positions and velocities, which is what I want.

After the x y z parameters though, there is a 0 0 0 or 0 -1 0 etc. which I am not sure why it is there and if there is a way to get rid of this.

Anyways, when I start my new script with the read_data command of datafile,

I get :

ERROR: Expected integer parameter in input script or data file (…/update.cpp:411)

It is not so easy to edit the datafile as it is very large.

Here is an example of the datafile that was written:

LAMMPS data file via write_data, version 5 Sep 2014, timestep = 14800

576000 atoms
2 atom types

-3.1855691821745182e+00 1.9451962473817562e+02 xlo xhi
-7.5335619696836886e-01 1.7047191491096402e+02 ylo yhi
-1.1708772362496802e+00 2.1677829945825169e+02 zlo zhi
-9.8856956361423130e+01 0.0000000000000000e+00 0.0000000000000000e+00 xy xz yz

Masses

1 15.9994
2 28.0855

Atoms # full

1 1 1 -1.2000000000000000e+00 9.3990181188726396e-01 7.5291880663208333e-01 3.3107060379865683e-01 0 0 0
16 1 2 2.3999999999999999e+00 6.6216552747162680e+00 -5.2675250213444980e-01 8.9385452636332907e-01 0 0 0
28801 401 1 -1.2000000000000000e+00 1.0981248531742009e+00 8.2872183636341168e-01 1.1263069143283678e+01 0 0 0

These are image flags. There is no information on that in the write_data page of the manual, but you can check the read_data page. Not sure about the error though!

Arthur

Hi Benjamin,

Those are the image flags (more on them on this page http://lammps.sandia.gov/doc/read_data.html) - they store which periodic replica each atom is in, and are used to calculate the unwrapped coordinates. You only need them if you’re doing something that requires unwrapped coordinates, but they’re harmless otherwise.

We need to see your input script to answer the rest.

Thanks,

Niall

The input script that wrote the datafile was run on a CPU and is:

units metal
atom_style full
kspace_style pppm 1.0e-6
box tilt large
read_restart …/1000eV_12/radiate_v1.restart
pair_style hybrid/overlay coul/long 15 table linear 15000

pair_coeff * * coul/long
pair_coeff 1 1 table …/…/…/…/…/…/…/bks_zbl.potential O-O
pair_coeff 1 2 table …/…/…/…/…/…/…/bks_zbl.potential O-Si
write_data datafile

Now I am trying to start the simulation on a GPU.
The script looks like:

package gpu 1 neigh no
suffix off
units metal
atom_style full
kspace_style pppm 1.0e-6
box tilt large
read_data datafile

pair_style hybrid/overlay coul/long 15 table linear 15000

pair_coeff * * coul/long
pair_coeff 1 1 table …/…/…/…/…/…/…/bks_zbl.potential O-O
pair_coeff 1 2 table …/…/…/…/…/…/…/bks_zbl.potential O-Si

neighbor 2.0 bin
neigh_modify delay 10 check yes

I get this output when it runs:

LAMMPS (10 Aug 2015)
Reading data file …
triclinic box = (-3.18557 -0.753356 -1.17088) to (194.52 170.472 216.778) with tilt (-98.857 0 0)
WARNING: Triclinic box skew is large (…/domain.cpp:181)
2 by 2 by 4 MPI processor grid
reading atoms …
576000 atoms
reading velocities …
576000 velocities
Finding 1-2 1-3 1-4 neighbors …
Special bond factors lj: 0 0 0
Special bond factors coul: 0 0 0
0 = max # of 1-2 neighbors
0 = max # of 1-3 neighbors
0 = max # of 1-4 neighbors
1 = max # of special neighbors
ERROR: Expected integer parameter in input script or data file (…/update.cpp:411)

The input script that wrote the datafile was run on a CPU and is:

units metal
atom_style full
kspace_style pppm 1.0e-6
box tilt large
read_restart ../1000eV_12/radiate_v1.restart
pair_style hybrid/overlay coul/long 15 table linear 15000

pair_coeff * * coul/long
pair_coeff 1 1 table ../../../../../../../bks_zbl.potential O-O
pair_coeff 1 2 table ../../../../../../../bks_zbl.potential O-Si
write_data datafile

Now I am trying to start the simulation on a GPU.
The script looks like:

package gpu 1 neigh no
suffix off
units metal
atom_style full
kspace_style pppm 1.0e-6
box tilt large
read_data datafile

pair_style hybrid/overlay coul/long 15 table linear 15000

pair_coeff * * coul/long
pair_coeff 1 1 table ../../../../../../../bks_zbl.potential O-O
pair_coeff 1 2 table ../../../../../../../bks_zbl.potential O-Si

neighbor 2.0 bin
neigh_modify delay 10 check yes

I get this output when it runs:

LAMMPS (10 Aug 2015)
Reading data file ...
  triclinic box = (-3.18557 -0.753356 -1.17088) to (194.52 170.472 216.778)
with tilt (-98.857 0 0)
WARNING: Triclinic box skew is large (../domain.cpp:181)
  2 by 2 by 4 MPI processor grid
  reading atoms ...
  576000 atoms
  reading velocities ...
  576000 velocities
Finding 1-2 1-3 1-4 neighbors ...
Special bond factors lj: 0 0 0
Special bond factors coul: 0 0 0
  0 = max # of 1-2 neighbors
  0 = max # of 1-3 neighbors
  0 = max # of 1-4 neighbors
  1 = max # of special neighbors
ERROR: Expected integer parameter in input script or data file
(../update.cpp:411)

did you provide the *complete* input??
the error message you quote would only happen, if you use the
"reset_timestep" command with a non-integer argument.

axel.

that’s it, thanks

in the future, please always remember to do these two things first:

1) run your input with -echo screen, so you get to see *exactly* where
in the input the error happens and you don't ask questions about a
totally unrelated topic. as a person that is subscribed to
lammps-users as long as you are, you should have seen that kind of
advice been given many times...

2) if you cannot solve your problem this way, always provide
*complete* input decks, so that people have a realistic chance to
provide meaningful help without having to assume that you are not
telling us everything (to put it politely). if you don't know how to
solve a problem with a LAMMPS input, how could you know what part of
your input is relevant or not?

thanks in advance,
    axel.