[gmx-users] .cpt and .trr files

[Mark Abraham]

On 11/03/11, Anna Marabotti  <anna.marabotti at isa.cnr.it> wrote:
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Dear 
> all,
> 
> a very quick 
> question: I 
> would like to know if a checkpoint file produced with a simulation with Gromacs 
> contains really ALL information needed to restart a simulation; or better, if I 
> keep only the .cpt file and the .xtc file of a simulation, which 
> information I lost (apart from the precision of the result)? At present I'm 
> keeping he .trr files (which contains all information) only to be sure I can 
> restart a simulation, but I need DESPERATELY free disk 
> space...
> 
> 
> 
> 




There is no longer any need to keep a .trr unless you can identify a reason (e.g. keeping velocities, or you need full precision for some analytical purpose). The .cpt has a full-precision trajectory frame + state variables in it. You should be able to see that the .cpt file size is around 2/N of the size of a positions-only N-frame .trr file, because it has full-precision arrays of positions and velocities, plus a little bit more.

You can easily satisfy yourself that the .cpt restarts are exact - get a small water box and do both a split run and two part runs from the same initial conditions on a single processor using mdrun -reprod.

Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20110311/f7d71554/attachment.html>