Message boards :
Questions :
How the Allocation Score is calculated?
Message board moderation
Author | Message |
---|---|
Send message Joined: 26 Mar 20 Posts: 5 |
If I check how projects are chosen for a specific computer, it says that the allocation score is how much work is owed to the project. Is that score specific to my account or is it an average score through all Science United accounts? And I would like to know how the actual calculation is done to get the allocation score. |
Send message Joined: 26 Mar 20 Posts: 5 |
It seems the Allocation Score is on the project, not on any account or computer. https://scienceunited.org/doc/implementation.pdf has a good amount of information on this and on overall SU implementation. |
Send message Joined: 26 Mar 20 Posts: 5 |
It is described in more details here: https://scienceunited.org/doc/su_overview.pdf. But according to this, allocation balance is subtracted when calculating the project score. And only positive values will be there for allocation balance according to the formula (as I understand). So how come we see positive and negative values as Allocation Score? Is Allocation Score not the same as Allocation Balance but something derived from Allocation Balance? |
Send message Joined: 26 Mar 20 Posts: 5 |
Found all the answers at https://github.com/davidpanderson/science_united. Negative values can be there as Allocation Balance. If more hosts were removed from a project, Average Estimated Credits will go down below 0 at some point. This causes the Allocation Balance to be negative. (https://github.com/davidpanderson/science_united/blob/bf888cc1246702190f54cba3f0a4e5689e1c617c/su_schedule.inc) And Allocation Score we see on Science United dashboard is the negative value of Allocation Balance. So while the Allocation Balance is subtracted when calculating the project score, Allocation Score should get added. (https://github.com/davidpanderson/science_united/blob/bf888cc1246702190f54cba3f0a4e5689e1c617c/su_schedule.inc) I'm just a newbie to these things. So please note that others can add to this or correct me if I am wrong. |
© 2024 UC Berkeley