Mari,
If you’re familiar with the visualization program vmd, do you know if the pdb file produced by the lammps2pdb tool can be used to generate a video of the trajectory in vmd?
I’m trying to get a first video of a simple simulation of water running.
Directly below is a message that I sent to the vmd-mailing list.
regards,
I use LAMMPS trajectories with VMD, and I’d recommend reading the LAMMPS trajectory directly into VMD. The intermediate post-processing steps you mention are not necessary. Here’s the process:
Include a dump command in your input script that names your dump file and uses the lammpstrj extension. Something like this: ”dump 1 all custom 100 your_system.lammpstrj id type x y z”
Run your simulation.
Open your trajectory file in VMD in the normal way, and let VMD auto-detect that you’ve got a LAMMPS trajectory.
If, for some reason, you really have to have a different file format, note that you can also do a similar process with the dcd file format (might also need corresponding psf file), or make a pdb trajectory using a post-processing step like you mention.
Paul
daniel,
a few comments on using VMD or other visualization tools with LAMMPS.
first of all, you always have to keep in mind, that visualization
programs are dumb and can only use information that you
provide. if there is no indication that you are simulating water
molecules, no software can know this. in most programs, and
that includes VMD, these properties are _guessed_ through
heuristics, typically based on atom and residue names, following
the .pdb standard.
if you add LAMMPS to the mix, there will be the first problem.
LAMMPS does not produce output that has those properties
preserved. LAMMPS only knows about numerical atom types
and atom ids. if you feed those to a program like VMD, it does
not know what they stand for.
you seem to be using the .xyz file format. this is probably the
worst possible choice for any serious post processing. .xyz
(or more correctly xmol) format is a lowest common denominator
format that contains only the absolute essential information, i.e.
has the maximum loss of information, so it will be the most difficult
for a program like VMD to infer something from it. one example
is that .xyz contains no information about the simulation cell, so
you cannot use a periodic display or use any of the pbctools
plugins (outside the one to set the cell dimensions manually).
the most information about your system is typically contained
in a data file. so the best way to start a visualization project is
to read that one into VMD: topo readlammpsdata data.myfile full
now you could try to recover some information, e.g. you can
try to guess the element name from its mass:
topo guessatomdata element mass
and then copy it to the name field:
set sel [atomselect top all]
$sel set name [$sel get element]
$sel delete ; unsel set
and finally write this out to a directly supported topology file
in .psf format: animate write psf myfile.psf
this should write out all atoms, bonds and more to the psf file
in a way that VMD can read it back easily.
if you then load the .psf file first and then the .xyz or dump files
(as .lammpstrj), you should be seeing something more reasonable.
if you have an original .pdb file that matches the order of the
particles in the data file, you could try loading that after the
data file and transfer the information this way.
once you have a suitable .psf file, you can also switch to more
space efficient file formats, like .dcd or .xtc that only store the
coordinate and in a binary (compressed form). this will not only
save space, but also makes reading much faster.
if you spend a little time working your way through this and
understand why certain ways work the way the work (or don't).
it will be all pretty straightforward. at that point, i highly recommend
to write down your experience and procedure in a little tutorial
and make it available to other lammps users. for people like me,
it is all too obvious and thus often impossible to write a good
tutorial, since we tend to gloss over steps that are difficult to
understand for the less experienced, but are not even thought
about, once you understood this. only people who have just
sorted this out are in the unique position, to describe their
progress in a really helpful way where they are still able to
recall their own struggles.
cheers,
axel.