[Primexd] Accidental background subtraction

[Igal Jaegle]

Matt,

The random bkg subtraction is only changing dramatically the fcal inner and outer rings, the exact same ring locations where mcsmear is overlay event when random bkg in included in the simulation. What I do exactly is below:

vector<const DBeamPhoton *> locBeamPhotons;
  loop->Get(locBeamPhotons);

vector<const DEventRFBunch*> locEventRFBunches;
  loop->Get(locEventRFBunches);

double locRFTime = locEventRFBunches.empty() ? 0.0 : locEventRFBunches[0]->dTime;

and in the double fcal loop
for (unsigned int k = 0; k < (int) locBeamPhotons.size(); k ++) {

          const DBeamPhoton *ebeam = locBeamPhotons[k];
          double eb = ebeam->lorentzMomentum().E();
          double tb = ebeam->time();
          double zb = ebeam->position().Z();

double locDeltaTRF = tb - (locRFTime + (zb - m_targetZ)/29.9792458);
 double weight = 0;
          if (fabs(locDeltaTRF) <= 2.004) {
            weight = 1;
          } else if ( ( -(2.004 + 3.0 * 4.008) <= locDeltaTRF && locDeltaTRF <= -(2.004 + 4.008) ) ||
                      ( (2.004 + 4.008) <= locDeltaTRF && locDeltaTRF <= (2.004 + 3.0 * 4.008) ) ) {
   weight = -0.25;
          } else {
            continue;
          }

Beam Photon factory is selecting the detector with the best timing or one can request it is only using the photons.

I just scale by "weight" all histo.

tks ig.
________________________________
From: Shepherd, Matthew <mashephe at indiana.edu>
Sent: Thursday, April 16, 2020 7:00 AM
To: Igal Jaegle <ijaegle at jlab.org>
Cc: primexd at jlab.org <primexd at jlab.org>; Drew Smith <andrew.p.smith at duke.edu>
Subject: [EXTERNAL] Re: Accidental background subtraction


Igal,

On Apr 16, 2020, at 6:46 AM, Igal Jaegle <ijaegle at jlab.org<mailto:ijaegle at jlab.org>> wrote:

I am just removing the random coincidence below and in the peak of the prompt coincidence with the usual technique.

I think this only removes a background if the beam photon enters into your analysis.  If the beam photon does not enter your analysis, then it doesn't matter whether it is accidental or prompt.  How do you use the beam photon in your analysis?

I think the trigger bit is indeed very important.  If a substantial number of your triggers are FCAL + CCAL then these can almost never, by construction, provide the pi0 you desire for calibration and will only add background.  And I believe they will tend to contaminate inner layers most.

Thanks... I'm just trying to see if there is some lesson from this that can be propagated back to the standard calibration method.

Matt

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/primexd/attachments/20200416/3d0d0bed/attachment-0002.html>