buggy definition of boost from LorentzVector

[Alexander Grohsjean]

Dear all,

is it possible that there is a bug in the definition of a boost Vector in tools/SampleAnalyser/MABoost.h?
void SetBoostVector(const MALorentzVector& q)
{ setBoostVector(q.X(),q.Y(),q.Z()); }

Shouldn't the vector be divided by the energy of the LorentzVector?

Moreover, I realized that assuming p,q,r a LorentzVectors
(p-q-r).Px() != p.Px() - q.Px() - r.Px()

Finally, I would like to ask if there are event shape variables, like aplanarity, already available in Madanalysis?

Cheers and thanks, Alexander.

[Benjamin Fuks (fuks)]

Hi Alexander,

Thanks for having pointed out this bug. This is now fixed. Can you please check the v1.5beta version of the code,
   https://code.launchpad.net/~ma5dev/madanalysis5/v1.5beta
and let us know whether everything is now fine. Note that the bug had no impact so far as those methods have never been used before.

Cheers,

Benjamin

[Alexander Grohsjean (alexander-grohsjean)]

Dear Benjamin, dear all,

MABoost includes now

     try
     {
       if (q.T()==0) throw EXCEPTION_WARNING("Energy equal to zero.
Impossible to compute the boost.","",0);
       setBoostVector(q.X(),q.Y(),q.Z());
     }
     catch(const std::exception& e)
     {
       MANAGE_EXCEPTION(e);
       MABoost();

but but the division by q.T() is still missing. There is only a check
that q.T() is not 0.
The other bug with (a-b-c).Px() != a.Px()-b.Px()-c.Px(), a,b,c, being
LorentzVectors, is fixed. Thanks a lot!

Regarding my question about event shape observables like aplanarity: Are
they available in madanalysis?

Cheers, Alexander.

Am 19.09.2016 um 18:45 schrieb Benjamin Fuks:
> Hi Alexander,
>
> Thanks for having pointed out this bug. This is now fixed. Can you please check the v1.5beta version of the code,
> https://code.launchpad.net/~ma5dev/madanalysis5/v1.5beta
> and let us know whether everything is now fine. Note that the bug had no impact so far as those methods have never been used before.
>
> Cheers,
>
> Benjamin
>
> ** Changed in: madanalysis5
> Importance: Undecided => Medium
>
> ** Changed in: madanalysis5
> Importance: Medium => High
>
> ** Changed in: madanalysis5
> Status: New => Fix Committed
>
> ** Changed in: madanalysis5
> Milestone: None => v1.5
>
> ** Changed in: madanalysis5
> Assignee: (unassigned) => Eric Conte (eric-conte)
>

[Eric Conte (eric-conte)]

Dear Alexander,

Sorry for this mistake. The bug is fixed now.

Concerning your last question, sphericity and aplanarity are not implemented into MadAnalysis up to now... But it is a good idea for the next release.

Cheers,
Eric

[Benjamin Fuks (fuks)]

Hi Alexander,

We have a private (validated) implementation if the expert mode of the code, if interested. Please contact me directly (outside launchpad).

Cheers,

Benjamin

On 20 Sep 2016, at 22:32 , Eric Conte <email address hidden> wrote:

> Dear Alexander,
>
> Sorry for this mistake. The bug is fixed now.
>
> Concerning your last question, sphericity and aplanarity are not
> implemented into MadAnalysis up to now... But it is a good idea for the
> next release.
>
> Cheers,
> Eric
>
> --
> You received this bug notification because you are a member of
> MadAnalysisTeam, which is subscribed to MadAnalysis 5.
> https://bugs.launchpad.net/bugs/1623116
>
> Title:
> buggy definition of boost from LorentzVector
>
> Status in MadAnalysis 5:
> Fix Committed
>
> Bug description:
> Dear all,
>
> is it possible that there is a bug in the definition of a boost Vector in tools/SampleAnalyser/MABoost.h?
> void SetBoostVector(const MALorentzVector& q)
> { setBoostVector(q.X(),q.Y(),q.Z()); }
>
> Shouldn't the vector be divided by the energy of the LorentzVector?
>
> Moreover, I realized that assuming p,q,r a LorentzVectors
> (p-q-r).Px() != p.Px() - q.Px() - r.Px()
>
> Finally, I would like to ask if there are event shape variables, like
> aplanarity, already available in Madanalysis?
>
> Cheers and thanks, Alexander.
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/madanalysis5/+bug/1623116/+subscriptions