stephenbrooks.orgForumMuon1GeneralBest 250 results.dat?
Username: Password:
Search site:
Subscribe to thread via RSS
[AMD Users] IMDb Addict
2003-06-25 11:14:32
I read something on a [DPC] forum about a better(?) results.dat (best 250 results.dat) which lets you get higher muon transfer ratings, but the post was from oct.  2002. Does anyone know anything about this issue?

Thanks,

-IMDb Addict
[AMD Users] Michal Hajicek
2003-06-25 12:10:29
http://stephan202.qik.nl
[AMD Users] IMDb Addict
2003-06-26 00:36:36
But what does it do?  What's the benefit of using such a file and don't I lose all my previous results?

-IMDb Addict
DukeBox
2003-06-26 02:02:58
You can merge them with your current file.. it wil use the entry's to calculate higher results.
MaFi
2003-06-26 04:12:28
you dont lose your previous results, because they are also in the results.txt, which is uploaded)... what you download from stephan202.qik.nl is the results.dat (after the download you have to make sure it is called results.dat or rename it!).  the resutls.dat is the database (you dont upload this!). 
markus
[OCAU] Noodles
2003-06-26 04:12:43
This results file really helps, I contribute to that top results file every 24 hours.  And if you look at the http://muon.qik.nl , I currently have the 3rd best muon transfer at 14.94ish Smile
[AMD Users] IMDb Addict
2003-06-26 08:24:54
But which one should I download?  Results250, results500 or results1000?
And one of you guys said I could merge them and another said I don't even need it?  What's the deal?
And I guess you have less particles simulated because I read somewhere that only the best muon transfers are saved and not the bad ones, so on the top list, someone with an inproved results.dat will not rise in number of particles as fast as someone without the improved results.dat, but he _will_ rise in best muon transfers?  (and that's what DPAD is all about isn't it?  Wink)

-IMDb Addict
[AMD Users] Michal Hajicek
2003-06-26 12:01:35
IMHO best250 is enough in the current project state, and if you're running windows client, try file_version with both carriage-return and linefeed (/r/n).  And: all simulation results are saved to results.txt and finally sent to the server, so the statistics shows everything you have simulated and sent, not only your best results.
[AMD Users] IMDb Addict
2003-06-28 09:30:13
Ok, thanks for the info.

-IMDb Addict
Mike Malis
2003-06-28 11:51:46
By downloading these results and using them for the future calculations wouldn't this compromise the program such that it is cornering a specific design and not allowing the full spectrum of design patterns to emerge?

Mike Malis
[AMD Users] Michal Hajicek
2003-06-28 15:40:20
It was dicussed many times here (more than one thread) with no direct advice at the end - it's expected that some of us will use some best files and others won't, and it works like this till now, so everything is OK (design refining and diversity maintaining).  But you are right, if everyone used the same bestfile, diversity could quickly go near zero (which wouldn't be optimal - with one exception-finall refining).
Michal
[AMD Users] Michal Hajicek
2003-06-28 15:53:25
Duck-Shaped croissant !?!?  what the hell is this status (no stats.  info in forum faq) Big Grin Big Grin Big Grin
Maybe another Stephens joke.

edit

it disappeared just after posting this (member sign back), strange.
Pollock[Romulus2]
2003-06-29 21:22:10
Michal, I still have the same status thing in another thread.  We must be special or...........something like that.  Smile
MaFi
2003-07-03 11:24:37
hey, i have it as well...
Smile

MaFi

[edit]now i dont have it any longer... may be if you have exactly 30 posts...?[/edit]
: contact : - - -
E-mail: sbstrudel characterstephenbrooks.orgTwitter: stephenjbrooksMastodon: strudel charactersjbstrudel charactermstdn.io RSS feed

Site has had 25162916 accesses.