arbitrarily iterate Trajectory with given frame indices: parallel benefit about cpptraj HOT 5 CLOSED

Currently pytraj has parallel implementation but this is only good for single calculation. For multiple actions, pytraj can't not beat cpptraj due to python overhead (example shown below, again).

python has builtin multiprocessing so I don't bother user to install MPI.

ref restart.nc
autoimage
rms reference @CA
distance :1 :2

from cpptraj.

@hainm This is currently an area of active development. It's not just splitting up the trajectory, it's making sure the data is correctly used afterwards as well (e.g. a covariance matrix needs to be properly summed up, data set time series need to be put in the correct order, etc). I'm currently working on this in-house and will keep you up to date on my progress.

from cpptraj.

cool. I am waiting for it. Hopefully you release it soon. ha ha.

from cpptraj.

@drroe

I just double-check cpptraj's man and found 'crdframes', which is what pytraj currently need.

It's likely that cpptraj has everything I need to make pytraj and cpptraj better. So my job now is only to ask you does cpptraj have this, where is it?. lol.

from cpptraj.

'crdframes' solves my problem. Just added a PR for pytraj. I am closing this.

from cpptraj.