rerun file1 file2 ... keyword args ...
keyword = first or last or every or skip or start or stop or dump first args = Nfirts Nfirst = dump timestep to start on last args = Nlast Nlast = dumptimestep to stop on every args = Nevery Nevery = read snapshots matching every this many timesteps skip args = Nskip Nskip = read one out of every Nskip snapshots start args = Nstart Nstart = timestep on which pseudo run will start stop args = Nstop Nstop = timestep to which pseudo run will end dump args = same as read_dump command starting with its field arguments
rerun dump.file dump x y z vx vy vz rerun dump1.txt dump2.txt first 10000 every 1000 dump x y z rerun dump.vels dump x y z vx vy vz box yes format molfile lammpstrj rerun dump.dcd dump x y z box no format molfile dcd rerun ../run7/dump.file.gz skip 2 dump x y z box yes
Perform a psuedo simulation run where atom information is read one snapshot at a time from a dump file(s), and energies and forces are computed on the shapshot to produce thermodynamic or other output.
This can be useful in the following kinds of scenarios, after an initial simulation produced the dump file:
Conceptually, using the rerun command is like running an input script that has a loop in it (see the next and jump commands). Each iteration of the loop reads one snapshot from the dump file via the read_dump command, sets the timestep to the appropriate value, and then invokes a run command for zero timesteps to simply compute energy and forces, and any other thermodynamic output or diagnostic info you have defined. This computation also invokes any fixes you have defined that apply constraints to the system, such as fix shake or fix indent.
Note that a simulation box must already be defined before using the rerun command. This can be done by the create_box, read_data, or read_restart commands.
Also note that reading per-atom information from dump snapshots is limited to the atom coordinates, velocities and image flags as explained in the read_dump command. Other atom properties, which may be necessary to compute energies and forces, such as atom charge, or bond topology information for a molecular system, are not read from (or even contained in) dump files. Thus this auxiliary information should be defined in the usual way, e.g. in a data file read in by a read_data command, before using the rerun command.
If more than one dump file is specified, the dump files are read one after the other. It is assumed that snapshot timesteps will be in ascending order. If a snapshot is encountered that is not in ascending order, it will cause the rerun command to complete.
The first, last, every, skip keywords determine which snapshots are read from the dump file(s). Snapshots are skipped until they have a timestamp >= Nfirst. When a snapshot with a timestamp > Nlast is encountered, the rerun command finishes. Note below that the defaults for first and last are to read all snapshots. If the every keyword is set to a value > 0, then only snapshots with timestamps that are a multiple of Nevery are read (the first snapshot is always read). If Nevery = 0, then this criterion is ignored, i.e. every snapshot is read that meets the other criteria. If the skip keyword is used, then after the first snapshot is read, every Nth snapshot is read, where N = Nskip. E.g. if Nskip = 3, then only 1 out of every 3 snapshots is read, assuming the snapshot timestamp is also consistent with the other criteria.
The start and stop keywords have the same meaning that they do for the run command. They only need to be defined if (a) you are using a fix command that changes some value over time, and (b) you want the reference point for elapsed time (from start to stop) to be different than the first and last settings. See the doc page for individual fixes to see which ones can be used with the start/stop keywords. Note that if you define neither of the start/stop or first/last keywords, then LAMMPS treats the pseudo run as going from 0 to a huge value (effectively infinity). This means that any quantity that a fix scales as a fraction of elapsed time in the run, will essentially remain at its intiial value.
The dump keyword is required and must be the last keyword specified. Its arguments are passed internally to the read_dump command. The first argument following the dump keyword should be the field1 argument of the read_dump command. See the read_dump doc page for details on the various options it allows for extracting information from the dump file snapshots, and for using that information to alter the LAMMPS simulation.
In general, a LAMMPS input script that uses a rerun command can include and perform all the usual operations of an input script that uses the run command. There are a few exceptions and points to consider, as discussed here.
Fixes that perform time integration, such as fix nve or fix npt are not invoked, since no time integration is performed. Fixes that perturb or constrain the forces on atoms will be invoked, just as they would during a normal run. Examples are fix indent and fix langevin. So you should think carefully as to whether that makes sense for the manner in which you are reprocessing the dump snapshots.
If you only want the rerun script to perform analyses that do not involve pair interactions, such as use compute msd to calculated displacements over time, you do not need to define a pair style, which may also mean neighbor lists will not need to be calculated which saves time. The communicate cutoff command can also be used to insure ghost atoms are acquired from far enough away for operations like bond and angle evaluations, if no pair style is being used.
Every time a snapshot is read, the timestep for the simulation is reset, as if the >reset_timestep command were used. This command has some restrictions as to what fixes can be defined. See its doc page for details. For example, the fix deposit and fix dt/reset fixes are in this category. They also make no sense to use with a rerun command.
If time-averaging fixes like fix ave/time are used, they are invoked on timesteps that are a function of their Nevery, Nrepeat, and Nfreq settings. As an example, see the fix ave/time doc page for details. You must insure those settings are consistent with the snapshot timestamps that are read from the dump file(s). If an averaging fix is not invoked on a timestep it expects to be, LAMMPS will flag an error.
The various forms of LAMMPS output, as defined by the thermo_style, thermo, dump, and restart commands occur on specific timesteps. If successvive dump snapshots skip those timesteps, then no output will be produced. E.g. if you request thermodynamic output every 100 steps, but the dump file snapshots are every 1000 steps, then you will only see thermodynamic output every 1000 steps.
To read gzipped dump files, you must compile LAMMPS with the -DLAMMPS_GZIP option - see the Making LAMMPS section of the documentation.
The option defaults are first = 0, last = a huge value (effectively infinity), start = same as first, stop = same as last, every = 0, skip = 1;