Thanks @LoryWangxx proving the edge case. When the sequential atoms objects provided to VaspInteractive are read from a trajectory file, the cell parameter may suffer from round-off error. This may cause VaspInteractive to think they are having different cell parameters (due to inherited ase-calculator behavior).
See atoms in moving.traj.zip
A resolution to this issue may be to add default round-off tolerance for cell parameter alone in VaspInteractive. User can request a slightly loose criteria for cell but position change still using default limit (1e-15)
Thanks @LoryWangxx proving the edge case. When the sequential atoms objects provided to
VaspInteractiveare read from a trajectory file, the cell parameter may suffer from round-off error. This may causeVaspInteractiveto think they are having different cell parameters (due to inherited ase-calculator behavior). See atoms in moving.traj.zipA resolution to this issue may be to add default round-off tolerance for cell parameter alone in
VaspInteractive. User can request a slightly loose criteria for cell but position change still using default limit (1e-15)