Energy - temperature

Is it a mistake?
For two particles i get temperature which is twice the kinetic energy. I usually expect this vice versa.
LAMMPS of 31march2011.

Step Temp PotEng KinEng TotEng
0 0 -0.47963592 0 -0.47963592
2000 0.19440569 -0.576837 0.097202844 -0.47963415
4000 11.688544 -6.3663636 5.8442721 -0.52209143
6000 0.18838473 -0.57382235 0.094192363 -0.47962999

ADDITION:

I increased number of particles to 4, two pairs far away from each other. I see that energy (total, kin, pot) is intensive value (averaged over number of particles). Is it OK?

Is it a mistake?
For two particles i get temperature which is twice the kinetic energy. I usually expect this vice versa.
LAMMPS of 31march2011.

Step Temp PotEng KinEng TotEng
0 0 -0.47963592 0 -0.47963592
2000 0.19440569 -0.576837 0.097202844 -0.47963415
4000 11.688544 -6.3663636 5.8442721 -0.52209143
6000 0.18838473 -0.57382235 0.094192363 -0.47962999

ADDITION:
I increased number of particles to 4, two pairs far away from each other. I
see that energy (total, kin, pot) is intensive value (averaged over number
of particles). Is it OK?

you are using reduced units, right?

then the thermo output for ekin, epot, is divided by the number of particles.
see the thermo_modify norm command.

axel.

As the compute_modify doc page explains, T is normalized
by N-1 particles. So if N=2, that's a big difference.
You can change this via the compute_modify command.

Steve