| 3:15:48 AM PDT - Thu, Aug 28th 2014 |
|
Unless you're unlucky, you should be able to, yes. Have a look in the manual for the START/RESTART top-level directives.
In short, it should be enough if you have your .movecs file left over, although the .db file could contain something useful wrt restarting as well (I'm no NWChem expert). If a direct restart fails, you can try to just read in the orbitals from the .movecs file with the VECTORS directive in the DFT block.
Recommend you copy your files to a new directory and (re)start there, so that if the restart fails for some reason (like syntax error), you don't have your old/present restart files overwritten.
If you are unlucky, the calculation crashed while writing the .movecs file. Unless I'm wrong, NWChem doesn't first write a temporary version of the .movecs file before starting to overwrite the old one, so a crash in the middle of a file-write is a bit of a disaster. To prevent this for long calculations, it might be a good idea to, with some regularity, manually make a backup of the .movecs during the run.
Have fun!
|