[Rivet] inclusive kt jet cross section analysis

[James Monk]

Someone not using RivetGun to run Rivet will still be able to use any  
analyses that are not normalised without doing anything.  It will not  
run if they try and use a normalised analysis unless they have set the  
cross section using the setCrossSection method of analysis handler.   
In RivetGun, this gets set after event generation is finished but  
before the analysis is finalised so that the cross section benefits  
from increased statistics.

I think that calculation of the cross section is something we can  
expect the generator to do and so there is no need for Rivet to  
replicate it.  It is probably a good idea to be able to embed the  
cross section information in the HepMC data (is this possible  
already?) so that a program that reads events from a file can report  
that cross section to Rivet.  If the cross section information is not  
in the HepMC file then there is no way around the user having to tell  
their program what it is if they are running analyses normalised to  
the cross section.


On 27 Nov 2007, at 11:43, Jon Butterworth wrote:

> Hi Lars,
>
> That would be good I think yes.
>
> Regarding the cross section, one has to provide the info somehow if  
> you
> want the normalisation, so your main program should be able to use the
> setCrossSection() function to do this. James can probably say more...
>
> Cheers,
> Jon
>
> Lars Sonnenschein wrote:
>> Hello Jon
>> looking into the HepEx0701051 analysis code I realized that you only
>> produce the histograms for kT separation parameter D=0.7.
>>
>> Mike is also interested in the D=0.5 and D=1.0 plots, just one y  
>> bin as
>> in the publication, which requires to run the jet algo three times.
>>
>> Should I do the appropriate changes in your code?
>>
>> 	Lars
>>
>> P.S.: I see also the normalization via the CrossSection() function,  
>> which
>> gets set from RivetGun via setCrossSection().
>> How is this supposed to work if Rivet is just fed with a file of  
>> HepMC
>> events (stand alone without Rivet)?
>>
>> Wouldn't it be better to write the instantaneous generator cross  
>> section
>> into each HepMC event, so that Rivet could compute the normalization
>> factor while looping over the events?
>> This of course would require a HepMC data member storing the  
>> information.
>> But it would disentangle Rivet and RivetGun.
>>
>>
>>
>> ________________________________
>> Lars Sonnenschein
>> ________________________________
>> Home Institution:
>> PH/TH 4/1-030
>> CERN
>> CH-1211 Geneve 23
>> Switzerland
>> Tel.:+41(22)767-2827
>> --------------------------------
>> ________________________________
>> FNAL:
>> D0, PK151
>> Mailstop #352
>> Fermilab, P.O.Box 500
>> Batavia, IL 60510-500
>> Tel.: +1(630)840-8740
>> ________________________________
>
> -- 
> ~ 
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Prof. Jonathan M. Butterworth,                   
> J.Butterworth at ucl.ac.uk
> Head of High Energy Physics              http://www.hep.ucl.ac.uk/ 
> ~jmb/
> Physics and Astronomy Department                  Tel: +44 20 7679  
> 3444
> University College London                 Gower St, London WC1E 6BT,  
> UK
> ~ 
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>
>
>
>
>
>
>
>
> _______________________________________________
> Rivet mailing list
> Rivet at projects.hepforge.org
> http://www.hepforge.org/lists/listinfo/rivet