[Rivet] numerical difference between data and mc aida

[Andy Buckley]

Hi Albert & Sunil,

Could you send us a minimal example pair of AIDA files (just one
distribution would be best) that demonstrate this problem? (I guess we
can cook one up from what you've sent, but complete files are easier and
more reliable.)

Probably we just need to add a little more numerical tolerance to the
bin comparison in the make-plots script -- if you can send these today
we should get the fix into the imminent 1.8.3 release.

Thanks for the report!

Andy


On 03/04/13 11:39, Albert Knutsson wrote:
> Dear Rivet experts,
> 
> we have a numerical problem with an aida file.
> 
> The data aida file have points looking like this:
> 
>   <dataPoint>
>     <measurement value="5.000000e-02" errorPlus="1.666667e-02"
> errorMinus="1.666667e-02"/>
>     <measurement value="1.186100e+00" errorPlus="1.535834e-01"
> errorMinus="1.535834e-01"/>
>   </dataPoint>
>   <dataPoint>
>     <measurement value="8.333333e-02" errorPlus="1.666667e-02"
> errorMinus="1.666667e-02"/>
>     <measurement value="1.763658e+00" errorPlus="2.291718e-01"
> errorMinus="2.291718e-01"/>
>   </dataPoint>
> 
> 
> The corresponding points in the MC aida produced by rivet, looks like this:
> 
>    <dataPoint>
>     <measurement value="5.000001e-02" errorPlus="1.666666e-02"
> errorMinus="1.666666e-02"/>
>     <measurement value="1.281114e+00" errorPlus="2.092222e-02"
> errorMinus="2.092222e-02"/>
>   </dataPoint>
>   <dataPoint>
>     <measurement value="8.333334e-02" errorPlus="1.666667e-02"
> errorMinus="1.666667e-02"/>
>     <measurement value="1.924731e+00" errorPlus="2.560152e-02"
> errorMinus="2.560152e-02"/>
>   </dataPoint>
> 
> 
> The consequence is that rivet-mkhtml complains about
> "+++ Error in Histogram.divide(): Binning of histograms differs"
> 
> The ratio plots are still there, but there is a discontinuity in the
> distribution at the second point above.
> 
> Do you have a solution to this?
> 
> Thanks in advance!
> 
> Cheers,
> Sunil and Albert
> _______________________________________________
> Rivet mailing list
> Rivet at projects.hepforge.org
> http://www.hepforge.org/lists/listinfo/rivet
> 


-- 
Dr Andy Buckley, Royal Society University Research Fellow
Particle Physics Expt Group, University of Edinburgh

The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.