[Rivet] Validation meeting

[James Monk]

On 27 Nov 2008, at 18:53, Andy Buckley wrote:

> By the way, we typically run O(1M) events for these plots: is that
> reasonable within Athena?

Should be possible on our batch system, but I think you'll be out of  
luck using the RTT - like I said it has to be done in 24 hours and  
shares the cpu with other jobs.

BTW, the idea of the ckin(4) cut is there is no "merge point" - in  
principle each plot contributes something to each bin (in proportion  
to the cross section in that bin from that plot), like you are  
generating the events flat(ter) in PT but with a weight.  Can't do it  
without the PT spectrum (or Yoda) though, so I'll do as you suggest if  
it's necessary.