"ERROR: All shapes are not set"

Greetings:

I am trying to run a single point energy calculation of a system using:

atom_style dipole

and:

pair_style dipole/cut 10.0

A line of the atomic positions looks like this:

  1 1 -0.0227600 0.147890 0.128080 1.207528 -0.03756
-0.03985 0.09539

Which should be fine with the description of this atom style in
LAMMPS, as it contains the information needed:

atom index, atom type, charge, x, y, z, mux, muy, muz

I don't understand why LAMMPS wants me to define a "Shape". Am I
missing something?

Camilo

Greetings:

I am trying to run a single point energy calculation of a system using:

atom_style dipole

and:

pair_style dipole/cut 10.0

A line of the atomic positions looks like this:

  1 1 -0.0227600 0.147890 0.128080 1.207528 -0.03756
-0.03985 0.09539

Which should be fine with the description of this atom style in
LAMMPS, as it contains the information needed:

atom index, atom type, charge, x, y, z, mux, muy, muz

I don't understand why LAMMPS wants me to define a "Shape". Am I
missing something?

are you using the current version of LAMMPS?
the online documentation always refers to that.

also, please provide a functional input when reporting
problems. it is extremely difficult to check something
with just a few tiny fragments.

thanks,
     axel.

Hi Axel,

No, I am not using the most current version of LAMMPS. However, the
version I am using does feature the dipole/cut pair_style. My
version runs the example included in the distribution (v. 10 SEP 2010)
without a hitch. Also, I've been running into some trouble compiling
the latest from the svn repository, so I've been unable to verify it
with that version.

My input file is the following:

# SPCFw

units real

atom_style dipole

read_data test.data

pair_style dipole/cut 10.0

pair_coeff * * 1.0000000 1.000000

neighbor 0.3 bin
neigh_modify delay 0

fix 1 all nvt temp 5.0 5.0 100

thermo 1
thermo_style custom cpu pe evdwl ecoul epair etotal
thermo_modify format float %15.8f

timestep 0.50
run 2

The first few lines of the data file looks like this:

# water

  54 atoms

3 atom types

0.000000 8.873483 xlo xhi
0.000000 7.684662 ylo yhi
0.000000 24.150563 zlo zhi
-4.436740 0.000000 0.000000 xy xz yz

Masses

1 196.988589
2 107.8682
3 35.453

Atoms

  1 1 -0.0227600 0.147890 0.128080 1.207528 -0.03756
-0.03985 0.09539
  2 1 -0.0227600 3.105715 0.128080 1.207528 -0.03759
-0.03985 0.09539
  3 1 -0.0227600 6.063548 0.128080 1.207528 -0.03753
-0.03985 0.09539

The error message I get is this one:

LAMMPS (4 Dec 2010)
Reading data file ...
  triclinic box = (0 0 0) to (8.87348 7.68466 24.1506) with tilt (-4.43674 0 0)
  1 by 1 by 1 processor grid
  54 atoms
ERROR: All shapes are not set

Hi Axel,

No, I am not using the most current version of LAMMPS. However, the
version I am using does feature the dipole/cut pair_style. My
version runs the example included in the distribution (v. 10 SEP 2010)

nobody will try to sort out issue with an old version
that cannot be validated with the current version.
some things have changed since 2010.

without a hitch. Also, I've been running into some trouble compiling
the latest from the svn repository, so I've been unable to verify it
with that version.

try a clean checkout. the current code compiles fine.

axel.

That error is referring to the "shape" command, which
no longer exists in current LAMMPS. It is replaced
by per-atom shapes (for ellipsoids) or diameters (for spheres).

So I think you need to update to the current version. The example
file for dipole will also be changed in the current version.

Steve