Posted by Jellby on June 26, 2007 at 14:11:38:
In Reply to: Re: Resuming an optimization posted by Jellby on June 07, 2007 at 18:22:40:
: That's unfortunate, indeed. The method you propose will still miss the updated Hessian and trust radius information...
OK, it seems I was wrong. The optimization continues nicely and the new iterations are added to the $Project.structure file. The only problem I see is that The Structure.xx.output files, if created, don't follow the same numbering, but start again from 1 and, moreover, they are appended and not overwritten. So, if I run successive 10-step optimizations, I would have only Structure.1.output to Structure.9.output, but each of them containing the information from several steps, for instance Structure.1.output will contain steps 1,10,19,28... (as far as I can tell).
Does this merit a bug report?