[Rivet] normalisation to cross-sections

Jonathan Butterworth jmb at hep.ucl.ac.uk
Fri Oct 23 09:30:31 BST 2009


now you worried me.

I was part of the long thread and I (thought I had) understood that 
event shape variables etc would *not* be normalised to cross section. 
They would have a Norm=xx.x which would mean they don't get scaled 
further. And they don't need the cross section info to get given Norm=1.0

Or did I miss something?

Hendrik Hoeth wrote:
> Thus spake Hendrik Hoeth (hendrik.hoeth at cern.ch):
> 
>> In fact, that's why I hadn't commented on this discussion so far --
>> too much painful brain activity needed.
> 
> Okay, I've started reading the looooong mails in the normalisation
> thread, and I'm beginning to get an idea of why you want to normalise
> each and every distribution using the cross-section. Still I'm not sure
> if that's the right way to go. What Rivet writes out after such a
> "normalisation" of an event shape variable is not directly comparable
> with the data anymore. And I fear that we break Rivet for many users.
> How do we explain to someone that he suddenly needs to provide
> cross-section information for observables that are clearly normalised to
> unity in the data?
> 
> More to be written in the other thread ...
> 
>      Hendrik
> 

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Prof. Jonathan Butterworth,              http://www.hep.ucl.ac.uk/~jmb/
Physics and Astronomy Department                  Tel: +44 20 7679 3444
ATLAS, CERN                                       Tel: +41 22 76  72340
University College London                 Gower St, London WC1E 6BT, UK
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


More information about the Rivet mailing list