Dear All,
I am using LAMMPS (14 Feb 2014) and have some questions :
1.I am using read_restart command and I want to know whether compute commands used for a simulation are stored in the restart file?It seems that manual doesn’t tell us.
2.Some fixes store an internal “state” which is written to the restart file.This allows the fix to continue on with its calculations in a restarted simulation.Can computes do that like some fixes?
Thank you very much for your answer
Dear All,
I am using LAMMPS (14 Feb 2014) and have some questions :1.I am using read_restart command and I want to know whether compute
commands used for a simulation are stored in the restart file?It seems that
manual doesn't tell us.
in general, if the manual doesn't say something explicitly, then you
can assume that it is not available. it is impossible to document
everything that cannot be done.
2.Some fixes store an internal "state" which is written to the restart
file.This allows the fix to continue on with its calculations in a restarted
simulation.Can computes do that like some fixes?
a compute computes an immediate property, why should it store data in a restart?
if a more complex computation require access to previous data, it is
not launched directly, but used in combination with a fix.
axel.
Axel is correct that there is nothing for a compute to store
in the restart file, since no compute has any “state”, b/c
it is just calculating an instantaneous property.
However, a few computes invoke a fix (behind the scenes)
to store something, like compute msd which needs to have
access to time 0 positions. Those fixes are stored in the
restart file. Which means that if you re-specify the same
compute in your restart script, it will re-specify the same
hidden fix and find that it was in the restart file, and hence
restore the needed info. This means you can continue
to use compute msd in the restart script, as if you had continued
calculating the MSD from the time 0 positions in the original run.
Steve