proceeding compute msd from restart-file

Hi all,

is there any chance to continue the computation of with respect to the initial coordinates?
I’ve just run a couple of tests and found that using the outdated compute msd/molecule and running from a restart-file, the msd is computed with respect to the current coordinates in the restart-file. Here’s the essential part of my script:

read_restart diff.800.restart


thermo_style custom step temp vol etotal pe ke epair ebond eangle edihed press density
thermo 10

timestep 1.0

label diff
restart 100 diff.*.restart
compute 2 all msd/molecule
fix 1 all nve
fix 2 all langevin 300.0 300.0 100.0 12345
fix 3 all ave/time 1 100 100 c_2 file msd.dat mode vector
run 3000000
unfix 4
unfix 3
unfix 2
unfix 1

The 2nd run i restarted from step 800, i used the same ID for my computes and still, the compute is re-initialized…see the appended png (msd vs step)
any hints? Maybe i should use compute cc1 all chunk/atom molecule and compute msd all msd/chunk cc1 ?

thx in advance,


no, the compute msd/chunk (current command) or compute msd/molecule (outdated)
does not allow you to do continue across a restart. See the doc page for compute msd and compute msd/chunk
and there are NOTEs about it. It could be implemented, but no one has done it.
The reasons it’s not easy is that the data that would need to be stored in the restart file
is not per-atom data, but per-molecule (chunk) data.