[Halld-offline] How many hypotheses per one track?

[semenov at jlab.org]

Hello Paul,

I'm trying to check how many PID hypotheses I have per one track. For each
event in the run 10913, I compared DTrackTimeBased.size() [which is the
number of charged track in the event] with DChargedTrackHypothesis.size()
[which is the number of hypotheses in the event, I hope]. The code is very
small and simple (just to be sure that the result is not affected by any
selection cuts):

 jerror_t JEventProcessor_tst1::evnt(JEventLoop *locEventLoop, uint64_t
eventnumber)
{
	vector<const DTrackTimeBased*> locTrackTimeBased;
	  locEventLoop->Get(locTrackTimeBased);
	vector<const DChargedTrackHypothesis*> locTrackHypothesis;
	  locEventLoop->Get(locTrackHypothesis);

	double mlt = 1.*(locTrackHypothesis.size()-locTrackTimeBased.size());
	double mltX = 1.*locTrackTimeBased.size();
	double mltY = 1.*locTrackHypothesis.size();

	mult1->Fill(mlt);
	mult2->Fill(mltX,mltY);

	return NOERROR;
}

The resulting histograms are in the attachment, and the conclusion is that
number of "charged" hypotheses is always equal to the number of "charged"
tracks (viz., one hypothesis per track).

I do believe that you said some time ago that we should have multiple
hypotheses for each track... Does it mean that there is something very
strange in these classes or I'm doing something incorrectly?

Thank you,
Andrei
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mult2.pdf
Type: application/pdf
Size: 16235 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20160503/3383b8c1/attachment.pdf>