<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body dir="auto">
Greetings Andrea,
<div><br>
</div>
<div>I believe you just verified what Diane Schott and I had been trying to convince other of for many years. :)</div>
<div><br>
</div>
<div>However, the correction for simulation is quite trivial since we already know the generated 4vector. </div>
<div><br>
</div>
<div>I don’t believe the kinfit package handles this problem.<br>
<br>
<div id="AppleMailSignature" dir="ltr">
<div style="direction: inherit;"><span style="background-color: rgba(255, 255, 255, 0);">BR<br>
MK <br>
----------------------------------------<br>
Michael C. Kunkel, USMC, PhD<br>
</span></div>
</div>
<div dir="ltr"><br>
On Jan 16, 2019, at 05:59, Andrea Celentano <<a href="mailto:andrea.celentano@ge.infn.it">andrea.celentano@ge.infn.it</a>> wrote:<br>
<br>
</div>
<blockquote type="cite">
<div dir="ltr"><span>Hi MK,</span><br>
<span>last meeting before holidays I presented the corrections necessary for photons, for both MC and data. For MC, the reconstructed photon energy is affected by a major shift with respect to the generated one - much larger than what happens for real data.
I remember you pointed out that this may be due to a non correct parametrization of the EC sampling fraction in the MC chain (gsim + gpp + a1c). For data, this effect is not present, since the difference between the measured photon energy and the “real” one
(as obtained from the kin. fit of gamma p -> p pi0 -> p e+ e- (gamma)) is much smaller.</span><br>
<span>Therefore, I suspect this effect is related to an error in gsim or gpp rather than a1c, or that different reconstruction routines are used in a1c for data and MC.</span><br>
<span></span><br>
<span>Can you comment more on this?</span><br>
<span></span><br>
<span>Thanks</span><br>
<span>Bests,</span><br>
<span>Andrea</span></div>
</blockquote>
</div>
</body>
</html>