(Date: 9.2.2015 – 14.2.2015)
A whole week without any talks and stuff!
I’m programming, in particular I’m looking at \(\chi^2\) distributions of my vertex- and mass-fitted D mesons. Because of doubly found tracks (and other effects), particles can be found twice, thrice, …, during track reconstruction. There’s, sometimes, no definite way to find out which is the real track representation — both are real, for that matter. So you consider, for the time being, both. That means, in a pure signal event data set, you can reconstruct a mother particle (a D meson, for my case) more than once, since there is more than one good daughter tracks. But, at some point in the reconstruction / analysis chain, you have to make a decision, which one of the particle hypotheses you trust more. And how you can distinguish them. And that’s what I’m trying to find out.
Also, during the last week, I look into something I call a count flow. The number of reconstruction ingredients of every reconstruction stage. Say, you start of with a bunch of signal particles to reconstruct. How many of them do leave hit points in the tracking detectors? How many of them such good hit points that they can be assembled into a particle? Et cetera. Here I see something I got hints of before very clearly: Positive and negative particles are not reconstructed equally. There’s a charge discrimination. That’s strange. Something I have to look into.
The third analysis subject of this week is background suppression. I have two data sets of background-like events — events, which are not my signal channel but try to mimic the QCD stuff happening at the targeted energy all the time. The enemy to fight against, if you will. I discuss some possible cuts with a colleague, write a script to test them quickly. During a meeting I show them, but most of the cuts I chose can just be applied in exclusive events1 — but I’d like to have, first, some cuts also working when only one of the signal candidates is reconstructable. Because statistics.
During Wednesday and Friday I stay at home and write. A lot of writing these days. Also on Friday: A quick online meeting concerning a pattern recognition day during the next collaboration meeting in ~1 month. I was one of the people suggesting such a meeting, but I’m not too sure to make in March. It’s just so close to my due date. :/
Events, where you reconstruct the whole decay. ↩