Regarding pair distance using pair/local

Dear users,

I am trying to calculate the local pressure in a small subregion using the volume approach(Lion and Allen, J. Phy. Condens. Matter, 2012) which requires me to evaluate fij, rij as well the fraction of line connecting rij which intersects the subregion.
As such, I use property/local to identify the pair atoms ids and pair/local to compute fij, rij which are written using dump local. However, when I calculate the distance between the pair atoms using their atomic coordinates x,y,z corresponding to their ids, the distance is not equal to rij for some pairs.
I am attaching the geometry file, input script and dump files for 1000th timestep. For the 1000th time step, pair no 20 in dump.file.1000 ->(1 842 9.15932 -0.00163738 (id1, id2, rij,fij)) .
But the coordinates of atomid 1 (-0.315859, -0.315859, 0.433176) and atomid 842 (24.8266, 7.96162, -0.270761) corresponding to pair 20 gives a distance of 26.4793 which donot match with 9.15932. Infact, out of 77068 pairs, 24412 pairs show such behavior.
Kindly suggest what is the reason for this problem?

Regards,
Manash Borthakur
IIT Guwahati

dump.full.1000 (28.3 KB)

dump.pair.1000 (2.11 MB)

geometry.data (56 KB)

input.in (1.29 KB)

Dear users,

  I am trying to calculate the local pressure in a small subregion using
the volume approach(Lion and Allen, J. Phy. Condens. Matter, 2012) which
requires me to evaluate fij, rij as well the fraction of line connecting
rij which intersects the subregion.
  As such, I use property/local to identify the pair atoms ids and
pair/local to compute fij, rij which are written using dump local. However,
when I calculate the distance between the pair atoms using their atomic
coordinates x,y,z corresponding to their ids, the distance is not equal to
rij for some pairs.
  I am attaching the geometry file, input script and dump files for 1000th
timestep. For the 1000th time step, pair no 20 in dump.file.1000 ->(1 842
9.15932 -0.00163738 (id1, id2, rij,fij)) .
  But the coordinates of atomid 1 (-0.315859, -0.315859, 0.433176) and
atomid 842 (24.8266, 7.96162, -0.270761) corresponding to pair 20 gives a
distance of 26.4793 which donot match with 9.15932. Infact, out of 77068
pairs, 24412 pairs show such behavior.

  Kindly suggest what is the reason for this problem?

there is no problem. you ​have periodic boundary conditions and did not
account for it when computing distances. LAMMPS will always report the
distance to the closest periodic image.

axel.