 |
| danzigrules 2006-08-03 22:00:30 | can I delete that file from my muon folder? |
| Stephen Brooks 2006-08-04 11:51:14 | results.dat holds the program's record of previous results it has generated, but if you are downloading samplefiles it will also get input data from there, so if you want you can delete results.dat, and your simulations will be based around the current network best and less so your own previous results. |
| Pascal 2007-03-23 09:59:45 | Another question: Is it possbile to start the client from its own without any sample files? It has been told several times that it may be better to try own simulations without any given data. Is that possible? Thanks.. |
| [TA]JonB 2007-03-23 11:45:35 | If you delete the results.dat file, then edit the config.txt to never download any sample files, then your client will run all on its own and come up with widely variant attempts, uninfluenced by the other clients from around the world. That increases the possibility that your client may find a completely new and efficient set of variables for the simulation. It also increases the possibility that it will produce nothing but already tried and inefficient settings. Regardless, it will still produce Mpts and increase your stats  . |
| Pascal 2007-03-23 13:52:48 | Thanks, JonB. I just forgot to use the config_nonet.bat script. |
| K`Tetch 2007-03-23 22:41:54 | [ta]jonb With the average lattice having at least 200 variables, that gives somewhere in the region of 10^600 (aka 1000^200) different simulations. Thus, the odds of comming up with an 'already tried and inefficient setting' are extremely small. Not sure how many variables the current lattices have, but at 220, its more like 10^660 - thats sixty more zeros after the one. I'm one of the people that run without any samplefiles, and whilst i don't have th worlds best computer, it does help by examining a broader range. |
