From: Axel Kohlmeyer (
Date: Mon Sep 05 2016 - 22:01:25 CDT

On Mon, Sep 5, 2016 at 10:06 PM, sanat tiwari <> wrote:
> Hi Axel and other mailing list users,

two remarks up front:

- there is no point in copying me on an e-mail to a mailing list, that
i am very obviously subscribed to. all you are doing is confusing my
e-mail filters and forcing me to make an extra effort to sort your
e-mail in the proper folder. not something that will rather place you
on my "naughty-list" than on the "nice-list"...

- cross-posting to multiple mailing lists is generally frowned upon,
especially when the overlap is rather small, as in this case. pick the
mailing list, where you have the highest hopes to get a helpful
response and repost to the other list, if you get told that you picked
the wrong list or don't get an answer within a couple of (work) days.
with cross-posting, everybody that responds while not being subscribed
to both lists will get a "bounce" from the "wrong" list, so - again -
you are not making friends.

in this case, you are very much asking about VMD, so that e-mail
should have only been sent to vmd-l.
i am cross-posting the reply to both lists, but any followup
discussion should please only go to vmd-l, thanks.

> I am trying to visualize lammps custom data (with positions and velocities
> in 3D) in VMD. Data contains few thousand free charge particles. I save
> total time step data in one single file say test.lammpstrj. I can directly
> load it in VMD and using graphic interface and further can modify its few
> properties but only with "point" representation.
> I want to use VDK etc representations but VMD won't let me do it (anything
> other than point representation is not working). How can I make it possible
> to see particles with VDK representation?

there is no VDK representation in VMD. do you mean VDW?

VMD can only show you information that is either present or can be
inferred from the data in your trajectory file. if your file only
contains coordinates and velocities, how on earth should VMD know
anything else, e.g. atom radii (which are required for VDW)? i have
yet to see a computer program with magical abilities, to make
something out of nothing.

> Tried to follow topo tool instructions but it gives following error:
> "Warning) readlammpsheader: skipping unkown header line: 220099 : 13076 1
> 0.000401159 0.000282808 0.000462548 999.367 926.541 662.331
> ERROR) readlammpsdata: failed to parse lammps data header. abort."

> Trying to make it work since few days without any success. Will appreciate
> suggestions.

you need to study documentation better. here is the documentation for
"topo readlammpsdata"

readlammpsdata <file name> [<atom style>]

Read in atom coordinates, properties, bond, angle, dihedral and other
related topology info from a LAMMPS data file, i.e. a file suitable
for the read_data command. This can be used to check a data file for
its validity (e.g. after usingrestart2data), for manipulations from
within VMD, or to generate a .psf file to be used for visualization of
.dcd or .xtc format trajectory files in VMD. The 'atom style' is the
value given to the atom_style command in the LAMMPS input file
(default is 'full'). This subcommand creates a new molecule in VMD and
returns its molecule id or -1 in case of failure.

the file you are trying to read is *definitely* not a "data" file, so
the error you are getting is wholly deserved. the command you are
using wasn't meant to for reading trajectory dumps.

> Even with point particle representation, if I want to use GLSL rendering,
> again it just show few big nicely colored particles.

VMD as a visualization program for atoms and molecules, will assume
that any coordinate files have atoms and molecules and try to infer
bonding and atom properties from information like atom type, names or
other data. it will assume coordinates are angstrom and the better the
information you provides agrees with typical conventions for PDB
files, the better a guess it will produce. if such information is not
present or the heuristic guesses are wrong, this can be corrected via
Tcl scripting.

i suggest you spend some more time reading documentation, user's
guides, and particularly search the web. i am certain, that there have
been other people with similar problems before, and that those have
received help through either the LAMMPS or VMD mailing list. i feel
that i at least must have answered such similar questions at least a
dozen times (forgive me, that my memory is a bit hazy. the number of
e-mails that i have sent to either list is not a small one).


> Following information might be useful:
> 1. My plasma simulations use Coulomb and/or other related multiple
> interaction potentials.
> 2. Using VMD 1.9.3 and Lammps March 16 versions
> 3. Mac OSX EI Captain with Nvidia graphics
> Thanks,
> Sanat
> ----------------------------------------------------
> Sanat Kumar Tiwari
> Department of Physics and Astronomy
> University of Iowa

Dr. Axel Kohlmeyer
College of Science & Technology, Temple University, Philadelphia PA, USA
International Centre for Theoretical Physics, Trieste. Italy.