[Rivet] Profile histograms in Rivet 2

[Andy Buckley]

Hmm, we need to look into this. I think most (all?) of the cases where
invalid stats calculations are attempted (mean with 0 entries, variance
etc. with < 2) now return a 0. I don't hugely like that, but it was a
quick fix for some problems and exceptions are more awkward to handle.
To be revisited in YODA design reviews...

For what it's worth, the code that does the conversion is in
YODA/ROOTCnv.h. Spot any errors in there? It would be nice to get a fix
into the next YODA version.

Thanks,
Andy


On 07/10/14 17:39, David Bjergaard wrote:
> Hi James,
> 
> (Shooting from the hip).  It looks like it wouldn't be too hard (reading
> the Profile1D.h interface) to roll your own YODA->ROOT converter.  
> 
> Do you know where the code that handles Rivet->Root conversion?  If this
> is athena software then I suppose this issue gets punted downstream?
> Andy can better comment on what happens if ProfileBin1D::mean() is
> called with no entries (in my experience it should throw an exception).
> In that case it may be that this behavior has changed causing the athena
> code to fail.
> 
>     Dave
> 
> James Robinson <james.robinson at cern.ch> writes:
> 
>> Dear David and Andy,
>>
>> Sorry about this, but I'm having another problem with my analysis'
>> Rivet routine. One of the histograms is a Profile1DPtr and, although
>> the .yoda output gives sensible numbers, the .root file has a TProfile
>> for which all bins are 0.
>>
>> Obviously this isn't a problem in Rivet per se, but I guess it might
>> be a problem with the athena Rivet_i interface. The only thing I can
>> think of that might be weird about this histogram is that some bins
>> have no entries (so a naive call to ProfileBin1D::mean() might fail).
>>
>> Any ideas about this one?
>>
>> Thanks,
>> James


-- 
Dr Andy Buckley, Royal Society University Research Fellow
Particle Physics Expt Group, University of Glasgow / PH Dept, CERN