[Rivet] Born and dressed level in Rivet

Tim Martin tim.martin at cern.ch
Thu Oct 3 16:40:17 BST 2013
Previous message: [Rivet] Born and dressed level in Rivet
Next message: [Rivet] Born and dressed level in Rivet
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Thanks,

A flag in the W/ZFinder constructor which is passed on to
ClusteredPhotons sounds like a good solution to me. Giving the option
to use this slightly more involved but more technically correct
version excluding photons from hadronisation and beyond.

We could then reassess on a case-by-case basis if we want to modify
some of our current Rivet analyses which corrected to a only FSR
photons definition.

Oleg has shown the difference between all photons and FSR only can be
up to ~3% in some variables for more extreme phases spaces, so high
pTz or with around 100 GeV of hadronic activity in the vicinity of the
Z.

Thanks, Tim.

On 3 October 2013 16:28, Ulla Blumenschein <ublumenschein at googlemail.com> wrote:
> Hi Andy,
>
> Many thanks for the feedback.
> Yes, what you detail sounds as good as we can get.
> Excluding the photons from any hadron decay chain, would be desirable.
>
> We would have to talk to SM subconveners and conveners of course to
> discuss the approach.
> As far as I know Oleg is running on standard evgen files or even
> ntuples, so I think he has all the flexibility to mimick both
> definitions, I hope.
>
>
> Cheers,
> Ulla
>
>
>
> On Thu, Oct 3, 2013 at 5:19 PM, Andy Buckley <andy.buckley at cern.ch> wrote:
>> On 03/10/13 17:00, Ulla Blumenschein wrote:
>>> Dear Andy,
>>>
>>> I  would be very happy if that works out.
>>> I cc to Oleg  and Tim, who did a comparison of various dressing
>>> definitions recently.
>>> Maybe Oleg would be interested to reproduce his plots with the
>>> proposed new Rivet implementation, to check what the residual effect
>>> is of the missing " from W/Z" requirement.
>>
>> I can certainly tell you how to hack an existing Rivet version to do
>> that. It's a bit low-level, though, so not obvious to me how it would be
>> done inside a particular analysis. But contact me/us and we'll do what
>> we can to help.
>>
>>> I have a technical question: Would only photons dircetly from
>>> quarkonia decays be excluded?  Where would photons from leptons from
>>> semileptonic HF decays end up?
>>
>> Our fromDecay(Particle&) function walks back up the decay chain looking
>> for a parent hadron or tau with status=2, meaning "physical hadron/tau
>> which has already been decayed". (Note the double specification: PYTHIA6
>> predates the standard status codes so we overspecify!) So photons from
>> charmonia and from hadron decay leptons would count as being from
>> decays: I assume that is what you want? Special case treatments for
>> these would take more coding, but if it'll be needed then we should know
>> now: at the very least this would indicate that the steering flag needs
>> to have more states than a bool would allow!
>>
>> Of course this would mean that the ClusteredPhotons projection would
>> need to be configured differently for charmonia analyses than for W/Z
>> ones, but that is hardly surprising. And we can leave the flag as
>> "include all" by default on ClusteredPhotons, but turn it to "non-decay
>> only" by default for the W/ZFinders. Sound good?
>>
>> Andy
>>
>>
>>
>>> On Thu, Oct 3, 2013 at 3:16 PM, Andy Buckley <andy.buckley at cern.ch> wrote:
>>>> Hi Ulla, all,
>>>>
>>>> I've discussed this a bit with Daniel Froidevaux and think that it would
>>>> make sense for us to at least add a flag to the ClusteredPhotons tool in
>>>> Rivet to exclude photons which come from hadron decays. Taus are a bit
>>>> of a borderline case, but that does not affect (current) EW measurements.
>>>>
>>>> This is not *exactly* the same definition, but I think that "only
>>>> photons from W/Z" is hard to implement in a generator-unspecific way.
>>>>
>>>> Maybe Frank Siegert wants to comment more on this proposal to use the
>>>> fromDecay() function in ClusteredPhotons, since he did most of the
>>>> detailed work to put these W/ZFinders and related classes together.
>>>>
>>>> Re. the Born comparisons, I believe that the ATLAS W/Z analyses written
>>>> so far *don't* compare to the Born-level results, since there *isn't* a
>>>> generator-portable way to write them. I remember removing a
>>>> generator-specific code along those lines in a previously submitted
>>>> analysis for that reason: it's fine for ATLAS-internal versions of
>>>> analyses but won't go into the public Rivet library.
>>>>
>>>> Hope that helps,
>>>> Andy
>>>>
>>>> PS. For those who care, if we agree that that is a reasonable route, I
>>>> propose that the exclude-decay-photons functionality would go into Rivet
>>>> 2.0.1 or 2.0.2. The exact version depends on the extent to which we want
>>>> to separate different types of change -- 2.0.1 might just be adding some
>>>> "queued" analysis codes, or it could be a combination of that and some
>>>> definition changes. The numbering does not really reflect anything about
>>>> the timescales! But note that we'll only be making this change in the
>>>> Rivet 2 series -- 1.x is in "maintenance mode".
>>>>
>>>>
>>>> On 03/10/13 14:51, Roman Lysak wrote:
>>>>>
>>>>>  Hi Ulla,
>>>>>
>>>>> I think the Rivet authors are better to replay to your questions so I'm
>>>>> cc-ing this to them.
>>>>> Could anyone address Ulla's questions?
>>>>>
>>>>> Cheers,
>>>>>   Roman
>>>>>
>>>>> On 10/03/2013 12:25 PM, Ulla Blumenschein wrote:
>>>>>> Hi Roman,
>>>>>>
>>>>>> we have currently a discussion in SM about how to synchronize  the
>>>>>> Atlas internal definition of dressed leptons with the Rivet
>>>>>> implementation.
>>>>>> Currently in SM, we sum all photons emitted from the W/Z decay
>>>>>> products whereas Rivet sums all photons.
>>>>>> With the extreme phase spaces which we access with 8TeV there are
>>>>>> differences up to percentage level between these definitions due to
>>>>>> pi0 decays from hadronic activity close to leptons.
>>>>>>
>>>>>> Is it possible to  implemet an alternate Z finder which can
>>>>>> discriminate photon sources? I understand that this might be against
>>>>>> the philosophy of Rivet (MC independence). Or are there other
>>>>>> problems?
>>>>>>
>>>>>> Also the born level Z finder is not completely clear to me. I assumed
>>>>>> so far that the born level is defined by different status codes in
>>>>>> different MC generator, so how does Rivet handle this issue?
>>>>>>
>>>>>>
>>>>>> Many thanks in advance,
>>>>>> Ulla
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Sun, Sep 15, 2013 at 7:46 PM, Roman Lysak <lysak at fzu.cz> wrote:
>>>>>>>   Hi Ulla,
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On 09/12/2013 06:51 PM, Ulla Blumenschein wrote:
>>>>>>>> Hi Roman,
>>>>>>>>
>>>>>>>> Could you explain how we define Born and dressed in a generator
>>>>>>>> independent way in Rivet?
>>>>>>>
>>>>>>> it depends on the object type.
>>>>>>> For example, for Z boson (similarly for W boson), you can specify
>>>>>>> whether to
>>>>>>> add photons (next to last one parameter in the constructor below) in
>>>>>>> a given
>>>>>>> radius (2nd next to last one parameter) around the leptons from Z
>>>>>>> decay to
>>>>>>> the Z boson 4-momentum.
>>>>>>>
>>>>>>>        //fiducial phase space + born level
>>>>>>>        ZFinder zfinder_mu(-2.4, 2.4, 20, MUON, 66.0*GeV, 116.0*GeV, 0.1,
>>>>>>> false, false);
>>>>>>>        addProjection(zfinder_mu, "ZFinder_mu");
>>>>>>>
>>>>>>>        //for combined cross-sections (combined phase space + dressed
>>>>>>> level)
>>>>>>>        ZFinder zfinder_comb_mu(-2.5, 2.5, 20, MUON, 66.0*GeV,
>>>>>>> 116.0*GeV, 0.1,
>>>>>>> true, false);
>>>>>>>
>>>>>>>
>>>>>>> Sorry for late response.
>>>>>>>
>>>>>>> Cheers,
>>>>>>>    Roman
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>> Cheers, Ulla
>>>>>>>>
>>>>>>>> /_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
>>>>>>>>       Ulla Blumenschein
>>>>>>>>       II Physik, Uni Goettingen
>>>>>>>>       Friedrich-Hund-Platz 1, D01.110
>>>>>>>>       phone: 0049-551-397645
>>>>>>>> /_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> 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 Glasgow / PH Dept, CERN
>>>
>>>
>>>
>>
>>
>> --
>> Dr Andy Buckley, Royal Society University Research Fellow
>> Particle Physics Expt Group, University of Glasgow / PH Dept, CERN
>
>
>
> --
> /_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
>     Ulla Blumenschein
>     II Physik, Uni Goettingen
>     Friedrich-Hund-Platz 1, D01.110
>     phone: 0049-551-397645
> /_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/



-- 
-+-+-+-+ Tim Martin +-+ ATLAS +-+-+-+-
Room 448A, University of Warwick, CV47AL
Tel: (+44) 0-24761-51769
Web: http://Tim-Martin.co.uk
Previous message: [Rivet] Born and dressed level in Rivet
Next message: [Rivet] Born and dressed level in Rivet
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Rivet mailing list