[Rivet] [Rivet-svn] r1549 - in trunk: . src/Analyses

[Andy Buckley]

rivet-svn at projects.hepforge.org wrote:
> Author: hoeth
> Date: 2009-05-23 17:12:11 +0100 (Sat, 23 May 2009)
> New Revision: 1549
> 
> Modified:
>    trunk/ChangeLog
>    trunk/src/Analyses/ALEPH_1996_S3486095.cc
>    trunk/src/Analyses/DELPHI_1996_S3430090.cc
>    trunk/src/Analyses/DELPHI_2003_WUD_03_11.cc
>    trunk/src/Analyses/JADE_OPAL_2000_S4300807.cc
> Log:
> First step of fixing the e+e- jet analyses that depend on
> JADE and DURHAM from fastjet. What used to be called dmerge
> is now ymerge. Of course there is still a dmerge around, but
> that one has a completely different meaning.

Thanks Hendrik; that's a good FastJet 2.3 -> 2.4 gotcha to know about.

Can you check the FastJets projection... it also uses the dmerge method 
in the ySplit stuff that was based on Adam Davison and Gavin's h->bb 
code, so I suspect it no longer does what was intended. Jon, maybe you 
can take a look, too, since I think you added it in the first place...

So, no Rivet release on the hoped-for timescale, but having that 
pressure to produce releaseable code has been really useful in itself 
and we've made a lot of progress in the last few weeks. Well done.

I think the main thing to fix is now the jet shapes; I'll have another 
good look at that tomorrow. I suspect there is nothing wrong with the 
projection, but that filling histograms with quantities which already 
use bin widths is responsible for the wrong result. I'll try converting 
it to use data point sets like Emily's jet shape analysis does, and 
hopefully all we have left to do is the validation testing against 1.1.2.

Andy