[Rivet] CMS_2015_I1384119 plugin (minbias eta)

[Janssen Xavier]

Hi Holger,

I found the problem with the normalisation. Apparently we had 2 versions of the analysis floating around and I missed the updated one with correct normalisation (too many emails during hollidays / well known fact that one should never take hollidays / sic). I am attaching the correct code. Everything is in the counting of events passing cuts on #particles to veto elastic events I believe (#particles>2). You can test it but would be better if Hannes confirm as well.
At least now I get the correct plot while indeed I had the same problem as you with the version which is in rivet 2.4.0.

Thanks for reporting.
Cheers, Xavier.




On 22 Oct 2015, at 19:14, Holger Schulz <holger.schulz at durham.ac.uk<mailto:holger.schulz at durham.ac.uk>> wrote:

On 22/10/15 18:10, Janssen Xavier wrote:
Hi Holger,


On 22 Oct 2015, at 18:25, Holger Schulz <holger.schulz at durham.ac.uk<mailto:holger.schulz at durham.ac.uk>> wrote:

On 22/10/15 17:15, Janssen Xavier wrote:

Hi Holger

Oh Sorry, the label should be 1/N dNch/deta ofc as in the paper.


Hi Xavier,

thanks for the clarification, I'll fix this for the next release.

I just cross read the paper, there is one sentence that concerns me a bit:
The data points and uncertainties are symmetrized in ±η.

Yes, we have always being doing it since the 900 GeV paper I believe, which means we average the positive and negative bins and plot this average on both the + and - corresponding bins. So the normalisation is taken correctly. This means we double the statistic of the measurements as there is no reason to have different physics on both side ...

See I was running a minbias MC through the analysis and it is pretty
much a factor 2 off (attachment).

I can try to run it myself as I am just done with integrating 2.4.0 in our CMS software. I let you know.

Great!
 10k events are sufficient. I am struggling to get Pythia8 to run :)

Cheers,
Holger


Do you remember who made the validation plots that came with the tarball in rivet contrib?
I would like to run a setup present in those plots to rule out any rivet problems.

I think it was Hannes, so he is already in cc.

Thanks,
Holger


Cheers, Xavier.


Cheers, Xavier.




On 22 Oct 2015, at 18:10, Holger Schulz <iamholger at googlemail.com><mailto:iamholger at googlemail.com> wrote:

Hi Xavier, Hannes,

I am having trouble understand the 13TeV pseudorapidity
distribution in the CMS_2015_I1384119 analysis.

In the rivet version, the ylabel says 1/N dN/deta which would
mean that the integral of the curve should be =1 --- however
it is more like 24.

I had a look into the paper to see if there is some normalisation
factor but found that there on Figure 3 the ylabel is not mentioning
a normalisation, i.e. it says dNch/d eta.

Could you please clarify what the normalisation of this histogram
is?

Thanks,
Holger




_______________________________________________
Rivet mailing list
Rivet at projects.hepforge.org<mailto:Rivet at projects.hepforge.org>
https://www.hepforge.org/lists/listinfo/rivet


<eta.pdf>




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.hepforge.org/lists-archive/rivet/attachments/20151022/686f0297/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: CMS_2015_I1384119.cc
Type: application/octet-stream
Size: 2219 bytes
Desc: CMS_2015_I1384119.cc
URL: <https://www.hepforge.org/lists-archive/rivet/attachments/20151022/686f0297/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: d01-x01-y01.pdf
Type: application/pdf
Size: 18153 bytes
Desc: d01-x01-y01.pdf
URL: <https://www.hepforge.org/lists-archive/rivet/attachments/20151022/686f0297/attachment.pdf>