climateprediction.net home page
Should I complete this HadCM3?

Should I complete this HadCM3?

Message boards : Number crunching : Should I complete this HadCM3?
Message board moderation

To post messages, you must log in.

AuthorMessage
old_user141342

Send message
Joined: 20 Dec 05
Posts: 9
Credit: 127,973
RAC: 0
Message 37044 - Posted: 3 Jun 2009, 17:05:22 UTC

I'm (in theory) running this task. Because of short-deadline work for other projects, there haven't been any trickles since January. Before I continue working on the model, could someone tell me

(a) if it has already been abandoned by the project because of the missing trickles (I can't tell from the task status page),

(b) and if not, will the work still have any scientific value when it is finished (could be some months from now, it's running very slowly), or will you already have moved on to something else?

Thank you!
ID: 37044 · Report as offensive     Reply Quote
Profile Iain Inglis

Send message
Joined: 9 Jan 07
Posts: 467
Credit: 14,549,176
RAC: 317
Message 37046 - Posted: 3 Jun 2009, 17:53:38 UTC

In reply:

(a) The CPDN task scheduler is not very interactive. When tasks fail or pause it does not issue new tasks, so no task is abandoned by the project. Similarly, the task deadlines are long - but are ignored. In short, the project will accept whatever you are able to produce.

(b) The results are made available to climate researchers on the data portal site. No doubt the CPDN staff have their own research interests, which change from time to time. However, a safe working assumption is to assume that every model is worth completing.

Of course, if running a long model doesn't fit very well with the way your computer is used, you could select shorter models next time.
ID: 37046 · Report as offensive     Reply Quote
old_user141342

Send message
Joined: 20 Dec 05
Posts: 9
Credit: 127,973
RAC: 0
Message 37076 - Posted: 6 Jun 2009, 19:31:07 UTC - in response to Message 37046.  

The results are made available to climate researchers on the data portal site. No doubt the CPDN staff have their own research interests, which change from time to time. However, a safe working assumption is to assume that every model is worth completing.


Thank you for your prompt reply! I thought sometimes models were run in order to gather preliminary data for further research steps, making the preliminary models pointless when the "real" research has started. This seemingly not being the case, I'll do my best with the HadCM3.

Of course, if running a long model doesn't fit very well with the way your computer is used, you could select shorter models next time.


I've already done this. The next one will be a HadSM3 ... if they are still around.
ID: 37076 · Report as offensive     Reply Quote

Message boards : Number crunching : Should I complete this HadCM3?

©2024 cpdn.org