Granted credit less than Claimed

Message boards : RALPH@home bug list : Granted credit less than Claimed

To post messages, you must log in.


Send message
Joined: 21 Feb 06
Posts: 6
Credit: 1,419
RAC: 0
Message 2319 - Posted: 5 Oct 2006, 18:55:47 UTC

May be that I don't understoend the credit system well, but I think that with initial replication:1 and minimum quorum:1 should be granted credit equal to claimed, shouldn't it?
ID: 2319 · Report as offensive    Reply Quote
Profile feet1st

Send message
Joined: 7 Mar 06
Posts: 313
Credit: 116,623
RAC: 0
Message 2322 - Posted: 6 Oct 2006, 1:35:11 UTC

No that isn't the way it works here. They use a work-based credit system. So credit is granted based on the number of models your complete (you can see this in the graphic display), and this varies depending upon how long your run your work units (see Ralph or Rosetta preferences for WU runtime preference), and how fast your machine is.

So they grant credit based on an average of the other machines that are working on different WUs for the same protein, and same scientific method. So as results are reported in, a weighted average of credit claimed per model is maintained, and you are granted credit based upon your number of models processed, times that running average per model.

If you think of it in terms of other projects, it would be like have a rolling quorum where credit is issued immediately based on the current rolling average value of the work reported by everyone.

This is a fairly new credit system for Rosetta (and Ralph) and so the claimed credit is being displayed as a frame of reference to compare the new credit system with the old one, where credit granted was always whatever was claimed.
ID: 2322 · Report as offensive    Reply Quote

Message boards : RALPH@home bug list : Granted credit less than Claimed

©2020 University of Washington