climateprediction.net home page
Time remaining wrong

Time remaining wrong

Message boards : Number crunching : Time remaining wrong
Message board moderation

To post messages, you must log in.

AuthorMessage
Martin P.

Send message
Joined: 5 Feb 05
Posts: 28
Credit: 1,846,059
RAC: 0
Message 19127 - Posted: 10 Jan 2006, 9:22:14 UTC
Last modified: 10 Jan 2006, 9:22:42 UTC

In all my WUs I receive very weird numbers for the time remaining.

Example: model 10vs_100068301_0 shows (client hadsm3 4.12):
Time elapsed: 300:19:22
Percentage done: 50,98%
Time remaining: 178:14:54

Do the maths!

ID: 19127 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 19128 - Posted: 10 Jan 2006, 9:35:12 UTC

Wait till you get to the sulphur models. Now they ARE way off.

ID: 19128 · Report as offensive     Reply Quote
Profile old_user5994

Send message
Joined: 31 Aug 04
Posts: 239
Credit: 2,933,299
RAC: 0
Message 19130 - Posted: 10 Jan 2006, 9:44:01 UTC

The 5.2.x versions do a slightly better job of estimating. Bill Michel did suggest an even more accurate (perhaps) estimation, but the formula is more complex. Not sure if that is what stopped them from using the suggestion.

I also don\'t know if BIll has been collecting the data to see if the formula he suggested is still better or not. For me, what is worse is that the current version of BOINC View uses the older method still ... the next version should be at least showing the value that the BOINC Daemon is using ... so that will be an advance ...
ID: 19130 · Report as offensive     Reply Quote
belgix

Send message
Joined: 5 Aug 04
Posts: 85
Credit: 2,924,043
RAC: 0
Message 19132 - Posted: 10 Jan 2006, 10:35:30 UTC - in response to Message 19128.  

Wait till you get to the sulphur models. Now they ARE way off.


I agree and it should be horrible at the beginning of phase 5 ...

ID: 19132 · Report as offensive     Reply Quote
Martin P.

Send message
Joined: 5 Feb 05
Posts: 28
Credit: 1,846,059
RAC: 0
Message 19134 - Posted: 10 Jan 2006, 11:01:29 UTC - in response to Message 19130.  
Last modified: 10 Jan 2006, 11:02:56 UTC

\"The 5.2.x versions do a slightly better job of estimating. Bill Michel did suggest an even more accurate (perhaps) estimation, but the formula is more complex. Not sure if that is what stopped them from using the suggestion.

I also don\'t know if BIll has been collecting the data to see if the formula he suggested is still better or not. For me, what is worse is that the current version of BOINC View uses the older method still ... the next version should be at least showing the value that the BOINC Daemon is using ... so that will be an advance ...\"



Paul,

I do use BOINC Manager 5.2.13! The version number I posted above is the science-app\'s version.

ID: 19134 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 19136 - Posted: 10 Jan 2006, 11:34:02 UTC

The new version is better at estimating, but it takes time for it to learn. Possibly several models.
Which is OK with the other projects, but a bit of a bore with cpdn.

ID: 19136 · Report as offensive     Reply Quote
Profile Honza
Volunteer moderator

Send message
Joined: 5 Aug 04
Posts: 390
Credit: 2,475,242
RAC: 0
Message 19140 - Posted: 10 Jan 2006, 13:43:42 UTC - in response to Message 19136.  

The new version is better at estimating, but it takes time for it to learn. Possibly several models.
Which is OK with the other projects, but a bit of a bore with cpdn.

Yes, I\'m using 5.3.6 from trux and running on a Pentium D CPDN exclusive - time estimates are settled down are conrespondents with command sense/math.

<i>phpBB forum for CPDN, all are </i><a href="http://www.climateprediction.net/board">invited</a>
ID: 19140 · Report as offensive     Reply Quote

Message boards : Number crunching : Time remaining wrong

©2024 climateprediction.net