LAMMPS WWW Site - LAMMPS Documentation - LAMMPS Commands

dump command

dump image command

dump movie command

dump molfile command

Syntax:

dump ID group-ID style N file args 

Examples:

dump myDump all atom 100 dump.atom
dump myDump all atom/mpiio 100 dump.atom.mpiio
dump 2 subgroup atom 50 dump.run.bin
dump 2 subgroup atom 50 dump.run.mpiio.bin
dump 4a all custom 100 dump.myforce.* id type x y vx fx
dump 4b flow custom 100 dump.%.myforce id type c_myF[3] v_ke
dump 2 inner cfg 10 dump.snap.*.cfg mass type xs ys zs vx vy vz
dump snap all cfg 100 dump.config.*.cfg mass type xs ys zs id type c_Stress[2]
dump 1 all xtc 1000 file.xtc 

Description:

Dump a snapshot of atom quantities to one or more files every N timesteps in one of several styles. The image and movie styles are the exception: the image style renders a JPG, PNG, or PPM image file of the atom configuration every N timesteps while the movie style combines and compresses them into a movie file; both are discussed in detail on the dump image doc page. The timesteps on which dump output is written can also be controlled by a variable. See the dump_modify every command.

Only information for atoms in the specified group is dumped. The dump_modify thresh and region commands can also alter what atoms are included. Not all styles support all these options; see details below.

As described below, the filename determines the kind of output (text or binary or gzipped, one big file or one per timestep, one big file or multiple smaller files).

IMPORTANT NOTE: Because periodic boundary conditions are enforced only on timesteps when neighbor lists are rebuilt, the coordinates of an atom written to a dump file may be slightly outside the simulation box.

IMPORTANT NOTE: Unless the dump_modify sort option is invoked, the lines of atom information written to dump files (typically one line per atom) will be in an indeterminate order for each snapshot. This is even true when running on a single processor, if the atom_modify sort option is on, which it is by default. In this case atoms are re-ordered periodically during a simulation, due to spatial sorting. It is also true when running in parallel, because data for a single snapshot is collected from multiple processors, each of which owns a subset of the atoms.

For the atom, custom, cfg, and local styles, sorting is off by default. For the dcd, xtc, xyz, and molfile styles, sorting by atom ID is on by default. See the dump_modify doc page for details.

As explained below, the atom/mpiio, custom/mpiio, and xyz/mpiio styles are identical in command syntax and in the format of the dump files they create, to the corresponding styles without "mpiio", except the single dump file they produce is written in parallel via the MPI-IO library. For the remainder of this doc page, you should thus consider the atom and atom/mpiio styles (etc) to be inter-changeable. The one exception is how the filename is specified for the MPI-IO styles, as explained below.


The style keyword determines what atom quantities are written to the file and in what format. Settings made via the dump_modify command can also alter the format of individual values and the file itself.

The atom, local, and custom styles create files in a simple text format that is self-explanatory when viewing a dump file. Many of the LAMMPS post-processing tools, including Pizza.py, work with this format, as does the rerun command.

For post-processing purposes the atom, local, and custom text files are self-describing in the following sense.

The dimensions of the simulation box are included in each snapshot. For an orthogonal simulation box this information is is formatted as:

ITEM: BOX BOUNDS xx yy zz
xlo xhi
ylo yhi
zlo zhi 

where xlo,xhi are the maximum extents of the simulation box in the x-dimension, and similarly for y and z. The "xx yy zz" represent 6 characters that encode the style of boundary for each of the 6 simulation box boundaries (xlo,xhi and ylo,yhi and zlo,zhi). Each of the 6 characters is either p = periodic, f = fixed, s = shrink wrap, or m = shrink wrapped with a minimum value. See the boundary command for details.

For triclinic simulation boxes (non-orthogonal), an orthogonal bounding box which encloses the triclinic simulation box is output, along with the 3 tilt factors (xy, xz, yz) of the triclinic box, formatted as follows:

ITEM: BOX BOUNDS xy xz yz xx yy zz 
xlo_bound xhi_bound xy
ylo_bound yhi_bound xz
zlo_bound zhi_bound yz 

The presence of the text "xy xz yz" in the ITEM line indicates that the 3 tilt factors will be included on each of the 3 following lines. This bounding box is convenient for many visualization programs. The meaning of the 6 character flags for "xx yy zz" is the same as above.

Note that the first two numbers on each line are now xlo_bound instead of xlo, etc, since they repesent a bounding box. See this section of the doc pages for a geometric description of triclinic boxes, as defined by LAMMPS, simple formulas for how the 6 bounding box extents (xlo_bound,xhi_bound,etc) are calculated from the triclinic parameters, and how to transform those parameters to and from other commonly used triclinic representations.

The "ITEM: ATOMS" line in each snapshot lists column descriptors for the per-atom lines that follow. For example, the descriptors would be "id type xs ys zs" for the default atom style, and would be the atom attributes you specify in the dump command for the custom style.

For style atom, atom coordinates are written to the file, along with the atom ID and atom type. By default, atom coords are written in a scaled format (from 0 to 1). I.e. an x value of 0.25 means the atom is at a location 1/4 of the distance from xlo to xhi of the box boundaries. The format can be changed to unscaled coords via the dump_modify settings. Image flags can also be added for each atom via dump_modify.

Style custom allows you to specify a list of atom attributes to be written to the dump file for each atom. Possible attributes are listed above and will appear in the order specified. You cannot specify a quantity that is not defined for a particular simulation - such as q for atom style bond, since that atom style doesn't assign charges. Dumps occur at the very end of a timestep, so atom attributes will include effects due to fixes that are applied during the timestep. An explanation of the possible dump custom attributes is given below.

For style local, local output generated by computes and fixes is used to generate lines of output that is written to the dump file. This local data is typically calculated by each processor based on the atoms it owns, but there may be zero or more entities per atom, e.g. a list of bond distances. An explanation of the possible dump local attributes is given below. Note that by using input from the compute property/local command with dump local, it is possible to generate information on bonds, angles, etc that can be cut and pasted directly into a data file read by the read_data command.

Style cfg has the same command syntax as style custom and writes extended CFG format files, as used by the AtomEye visualization package. Since the extended CFG format uses a single snapshot of the system per file, a wildcard "*" must be included in the filename, as discussed below. The list of atom attributes for style cfg must begin with either "mass type xs ys zs" or "mass type xsu ysu zsu" since these quantities are needed to write the CFG files in the appropriate format (though the "mass" and "type" fields do not appear explicitly in the file). Any remaining attributes will be stored as "auxiliary properties" in the CFG files. Note that you will typically want to use the dump_modify element command with CFG-formatted files, to associate element names with atom types, so that AtomEye can render atoms appropriately. When unwrapped coordinates xsu, ysu, and zsu are requested, the nominal AtomEye periodic cell dimensions are expanded by a large factor UNWRAPEXPAND = 10.0, which ensures atoms that are displayed correctly for up to UNWRAPEXPAND/2 periodic boundary crossings in any direction. Beyond this, AtomEye will rewrap the unwrapped coordinates. The expansion causes the atoms to be drawn farther away from the viewer, but it is easy to zoom the atoms closer, and the interatomic distances are unaffected.

The dcd style writes DCD files, a standard atomic trajectory format used by the CHARMM, NAMD, and XPlor molecular dynamics packages. DCD files are binary and thus may not be portable to different machines. The number of atoms per snapshot cannot change with the dcd style. The unwrap option of the dump_modify command allows DCD coordinates to be written "unwrapped" by the image flags for each atom. Unwrapped means that if the atom has passed through a periodic boundary one or more times, the value is printed for what the coordinate would be if it had not been wrapped back into the periodic box. Note that these coordinates may thus be far outside the box size stored with the snapshot.

The xtc style writes XTC files, a compressed trajectory format used by the GROMACS molecular dynamics package, and described here. The precision used in XTC files can be adjusted via the dump_modify command. The default value of 1000 means that coordinates are stored to 1/1000 nanometer accuracy. XTC files are portable binary files written in the NFS XDR data format, so that any machine which supports XDR should be able to read them. The number of atoms per snapshot cannot change with the xtc style. The unwrap option of the dump_modify command allows XTC coordinates to be written "unwrapped" by the image flags for each atom. Unwrapped means that if the atom has passed thru a periodic boundary one or more times, the value is printed for what the coordinate would be if it had not been wrapped back into the periodic box. Note that these coordinates may thus be far outside the box size stored with the snapshot.

The xyz style writes XYZ files, which is a simple text-based coordinate format that many codes can read. Specifically it has a line with the number of atoms, then a comment line that is usually ignored followed by one line per atom with the atom type and the x-, y-, and z-coordinate of that atom. You can use the dump_modify element option to change the output from using the (numerical) atom type to an element name (or some other label). This will help many visualization programs to guess bonds and colors.

Note that atom, custom, dcd, xtc, and xyz style dump files can be read directly by VMD, a popular molecular viewing program. See Section tools of the manual and the tools/lmp2vmd/README.txt file for more information about support in VMD for reading and visualizing LAMMPS dump files.


Dumps are performed on timesteps that are a multiple of N (including timestep 0) and on the last timestep of a minimization if the minimization converges. Note that this means a dump will not be performed on the initial timestep after the dump command is invoked, if the current timestep is not a multiple of N. This behavior can be changed via the dump_modify first command, which can also be useful if the dump command is invoked after a minimization ended on an arbitrary timestep. N can be changed between runs by using the dump_modify every command (not allowed for dcd style). The dump_modify every command also allows a variable to be used to determine the sequence of timesteps on which dump files are written. In this mode a dump on the first timestep of a run will also not be written unless the dump_modify first command is used.

The specified filename determines how the dump file(s) is written. The default is to write one large text file, which is opened when the dump command is invoked and closed when an undump command is used or when LAMMPS exits. For the dcd and xtc styles, this is a single large binary file.

Dump filenames can contain two wildcard characters. If a "*" character appears in the filename, then one file per snapshot is written and the "*" character is replaced with the timestep value. For example, tmp.dump.* becomes tmp.dump.0, tmp.dump.10000, tmp.dump.20000, etc. This option is not available for the dcd and xtc styles. Note that the dump_modify pad command can be used to insure all timestep numbers are the same length (e.g. 00010), which can make it easier to read a series of dump files in order with some post-processing tools.

If a "%" character appears in the filename, then each of P processors writes a portion of the dump file, and the "%" character is replaced with the processor ID from 0 to P-1. For example, tmp.dump.% becomes tmp.dump.0, tmp.dump.1, ... tmp.dump.P-1, etc. This creates smaller files and can be a fast mode of output on parallel machines that support parallel I/O for output. This option is not available for the dcd, xtc, and xyz styles.

By default, P = the number of processors meaning one file per processor, but P can be set to a smaller value via the nfile or fileper keywords of the dump_modify command. These options can be the most efficient way of writing out dump files when running on large numbers of processors.

Note that using the "*" and "%" characters together can produce a large number of small dump files!

For the atom/mpiio, custom/mpiio, and xyz/mpiio styles, a single dump file is written in parallel via the MPI-IO library, which is part of the MPI standard for versions 2.0 and above. Using MPI-IO requires two steps. First, build LAMMPS with its MPIIO package installed, e.g.

make yes-mpiio    # installs the MPIIO package
make g++          # build LAMMPS for your platform 

Second, use a dump filename which contains ".mpiio". Note that it does not have to end in ".mpiio", just contain those characters. Unlike MPI-IO restart files, which must be both written and read using MPI-IO, the dump files produced by these MPI-IO styles are identical in format to the files produced by their non-MPI-IO style counterparts. This means you can write a dump file using MPI-IO and use the read_dump command or perform other post-processing, just as if the dump file was not written using MPI-IO.

Note that MPI-IO dump files are one large file which all processors write to. You thus cannot use the "%" wildcard character described above in the filename since that specifies generation of multiple files. You can use the ".bin" suffix described below in an MPI-IO dump file; again this file will be written in parallel and have the same binary format as if it were written without MPI-IO.

If the filename ends with ".bin", the dump file (or files, if "*" or "%" is also used) is written in binary format. A binary dump file will be about the same size as a text version, but will typically write out much faster. Of course, when post-processing, you will need to convert it back to text format (see the binary2txt tool) or write your own code to read the binary file. The format of the binary file can be understood by looking at the tools/binary2txt.cpp file. This option is only available for the atom and custom styles.

If the filename ends with ".gz", the dump file (or files, if "*" or "%" is also used) is written in gzipped format. A gzipped dump file will be about 3x smaller than the text version, but will also take longer to write. This option is not available for the dcd and xtc styles.


This section explains the local attributes that can be specified as part of the local style.

The index attribute can be used to generate an index number from 1 to N for each line written into the dump file, where N is the total number of local datums from all processors, or lines of output that will appear in the snapshot. Note that because data from different processors depend on what atoms they currently own, and atoms migrate between processor, there is no guarantee that the same index will be used for the same info (e.g. a particular bond) in successive snapshots.

The c_ID and c_ID[N] attributes allow local vectors or arrays calculated by a compute to be output. The ID in the attribute should be replaced by the actual ID of the compute that has been defined previously in the input script. See the compute command for details. There are computes for calculating local information such as indices, types, and energies for bonds and angles.

Note that computes which calculate global or per-atom quantities, as opposed to local quantities, cannot be output in a dump local command. Instead, global quantities can be output by the thermo_style custom command, and per-atom quantities can be output by the dump custom command.

If c_ID is used as a attribute, then the local vector calculated by the compute is printed. If c_ID[N] is used, then N must be in the range from 1-M, which will print the Nth column of the M-length local array calculated by the compute.

The f_ID and f_ID[N] attributes allow local vectors or arrays calculated by a fix to be output. The ID in the attribute should be replaced by the actual ID of the fix that has been defined previously in the input script.

If f_ID is used as a attribute, then the local vector calculated by the fix is printed. If f_ID[N] is used, then N must be in the range from 1-M, which will print the Nth column of the M-length local array calculated by the fix.

Here is an example of how to dump bond info for a system, including the distance and energy of each bond:

compute 1 all property/local batom1 batom2 btype 
compute 2 all bond/local dist eng
dump 1 all local 1000 tmp.dump index c_1[1] c_1[2] c_1[3] c_2[1] c_2[2] 

This section explains the atom attributes that can be specified as part of the custom and cfg styles.

The id, mol, proc, type, element, mass, vx, vy, vz, fx, fy, fz, q attributes are self-explanatory.

Id is the atom ID. Mol is the molecule ID, included in the data file for molecular systems. Proc is the ID of the processor (0 to Nprocs-1) that currently owns the atom. Type is the atom type. Element is typically the chemical name of an element, which you must assign to each type via the dump_modify element command. More generally, it can be any string you wish to associated with an atom type. Mass is the atom mass. Vx, vy, vz, fx, fy, fz, and q are components of atom velocity and force and atomic charge.

There are several options for outputting atom coordinates. The x, y, z attributes write atom coordinates "unscaled", in the appropriate distance units (Angstroms, sigma, etc). Use xs, ys, zs if you want the coordinates "scaled" to the box size, so that each value is 0.0 to 1.0. If the simulation box is triclinic (tilted), then all atom coords will still be between 0.0 and 1.0. Use xu, yu, zu if you want the coordinates "unwrapped" by the image flags for each atom. Unwrapped means that if the atom has passed thru a periodic boundary one or more times, the value is printed for what the coordinate would be if it had not been wrapped back into the periodic box. Note that using xu, yu, zu means that the coordinate values may be far outside the box bounds printed with the snapshot. Using xsu, ysu, zsu is similar to using xu, yu, zu, except that the unwrapped coordinates are scaled by the box size. Atoms that have passed through a periodic boundary will have the corresponding cooordinate increased or decreased by 1.0.

The image flags can be printed directly using the ix, iy, iz attributes. For periodic dimensions, they specify which image of the simulation box the atom is considered to be in. An image of 0 means it is inside the box as defined. A value of 2 means add 2 box lengths to get the true value. A value of -1 means subtract 1 box length to get the true value. LAMMPS updates these flags as atoms cross periodic boundaries during the simulation.

The mux, muy, muz attributes are specific to dipolar systems defined with an atom style of dipole. They give the orientation of the atom's point dipole moment. The mu attribute gives the magnitude of the atom's dipole moment.

The radius and diameter attributes are specific to spherical particles that have a finite size, such as those defined with an atom style of sphere.

The omegax, omegay, and omegaz attributes are specific to finite-size spherical particles that have an angular velocity. Only certain atom styles, such as sphere define this quantity.

The angmomx, angmomy, and angmomz attributes are specific to finite-size aspherical particles that have an angular momentum. Only the ellipsoid atom style defines this quantity.

The tqx, tqy, tqz attributes are for finite-size particles that can sustain a rotational torque due to interactions with other particles.

The c_ID and c_ID[N] attributes allow per-atom vectors or arrays calculated by a compute to be output. The ID in the attribute should be replaced by the actual ID of the compute that has been defined previously in the input script. See the compute command for details. There are computes for calculating the per-atom energy, stress, centro-symmetry parameter, and coordination number of individual atoms.

Note that computes which calculate global or local quantities, as opposed to per-atom quantities, cannot be output in a dump custom command. Instead, global quantities can be output by the thermo_style custom command, and local quantities can be output by the dump local command.

If c_ID is used as a attribute, then the per-atom vector calculated by the compute is printed. If c_ID[N] is used, then N must be in the range from 1-M, which will print the Nth column of the M-length per-atom array calculated by the compute.

The f_ID and f_ID[N] attributes allow vector or array per-atom quantities calculated by a fix to be output. The ID in the attribute should be replaced by the actual ID of the fix that has been defined previously in the input script. The fix ave/atom command is one that calculates per-atom quantities. Since it can time-average per-atom quantities produced by any compute, fix, or atom-style variable, this allows those time-averaged results to be written to a dump file.

If f_ID is used as a attribute, then the per-atom vector calculated by the fix is printed. If f_ID[N] is used, then N must be in the range from 1-M, which will print the Nth column of the M-length per-atom array calculated by the fix.

The v_name attribute allows per-atom vectors calculated by a variable to be output. The name in the attribute should be replaced by the actual name of the variable that has been defined previously in the input script. Only an atom-style variable can be referenced, since it is the only style that generates per-atom values. Variables of style atom can reference individual atom attributes, per-atom atom attributes, thermodynamic keywords, or invoke other computes, fixes, or variables when they are evaluated, so this is a very general means of creating quantities to output to a dump file.

See Section_modify of the manual for information on how to add new compute and fix styles to LAMMPS to calculate per-atom quantities which could then be output into dump files.


Restrictions:

To write gzipped dump files, you must compile LAMMPS with the -DLAMMPS_GZIP option - see the Making LAMMPS section of the documentation.

The atom/mpiio, custom/mpiio, and xyz/mpiio styles are part of the MPIIO package. They are only enabled if LAMMPS was built with that package. See the Making LAMMPS section for more info.

The xtc style is part of the XTC package. It is only enabled if LAMMPS was built with that package. See the Making LAMMPS section for more info. This is because some machines may not support the low-level XDR data format that XTC files are written with, which will result in a compile-time error when a low-level include file is not found. Putting this style in a package makes it easy to exclude from a LAMMPS build for those machines. However, the XTC package also includes two compatibility header files and associated functions, which should be a suitable substitute on machines that do not have the appropriate native header files. This option can be invoked at build time by adding -DLAMMPS_XDR to the CCFLAGS variable in the appropriate low-level Makefile, e.g. src/MAKE/Makefile.foo. This compatibility mode has been tested successfully on Cray XT3/XT4/XT5 and IBM BlueGene/L machines and should also work on IBM BG/P, and Windows XP/Vista/7 machines.

Related commands:

dump image, dump_modify, undump

Default:

The defaults for the image style are listed on the dump image doc page.