I've implemented momentum recording and restarting using momentum. If Upside is run with the argument --record-momentum, then the momentum is recorded to f.root.output.mom, the same way as we did for pos. With the use of the python script I included in example/13.RestartSimulation, the end configuration and end momentum will be copied as the input position and momentum, such that we can properly restart an ended trajectory. The restart needs another argument --restart-using-momentum.
I've checked my implementation with test run on chig.pdb, and monitor the velocities here:
Before the implementation, the fluctuation near the resubmitted point (step=1000) is within the normal range of velocity fluctuation, so this should not have any huge effects on the simulation we ran previously.
I've implemented momentum recording and restarting using momentum. If Upside is run with the argument
--record-momentum
, then the momentum is recorded tof.root.output.mom
, the same way as we did forpos
. With the use of the python script I included inexample/13.RestartSimulation
, the end configuration and end momentum will be copied as the input position and momentum, such that we can properly restart an ended trajectory. The restart needs another argument--restart-using-momentum
.I've checked my implementation with test run on chig.pdb, and monitor the velocities here:
Before the implementation, the fluctuation near the resubmitted point (step=1000) is within the normal range of velocity fluctuation, so this should not have any huge effects on the simulation we ran previously.