[E16-e1f] Fwd: cc cuts

[Silvia Pisano]

Dear Mauri and dear all,

in the mail below you can find a summary of the procedure I developed to
implement the cc matching cuts on e1f.

Let me know if you have any question.

Best regards,

silvia


---------- Forwarded message ----------
From: Silvia Pisano <silvia.pisano at lnf.infn.it>
Date: 8 April 2016 at 17:00
Subject: cc cuts
To: Harut Avagyan <avakian at jlab.org>, Nathan Harrison <nathanh at jlab.org>,
Aram Movsisyan <maram at mail.desy.de>


Dear all,

after some weeks I managed to implement and test the Osipenko's cuts on phi
(from Nathan), theta and time. For theta and time I took definitions (and
inspiration!) from what Aram did (thanks a lot for your help with it,
Aram!), and developed an automatic fit procedure to define the cuts.

I wrote a set of function that I will put on github, so that if someone has
to implement those cuts can use it (Nathan, if you will be asked to do it
for the analysis review just let me know and I will help you with it).

For the theta fit I used a polynomial background plus a gaussian signal,
while for time I used a polynomial background plus N gaussians in order to
account for multiple reflection peaks.

This is an example for theta fits:

http://www.lnf.infn.it/~pisanos/dh/cc_theta_cuts_Sect3.pdf

(other sectors and in the same page, changing the corresponding number)

and for time:

http://www.lnf.infn.it/~pisanos/dh/cc_time_cuts_Sect3.pdf

The number of peaks is determined through the root class TSpectrum, and it
seems to work in a pretty good way (it failures just in few cases, if any).

I followed the procedure delined by Wes, cutting around three sigma from
the mean for theta and applying just a lower cut on time, so to account for
the multiple relfection. However, the combined action of the three cuts -
theta, time and phi - seems to be quite effective in selecting just the
gaussian signal, as you can see from, e.g.,:

http://www.lnf.infn.it/~pisanos/dh/cc_time_Sect_4_check.pdf

for time and in

http://www.lnf.infn.it/~pisanos/dh/cc_theta_Sect_4_check.pdf

for theta, where the empty histogram shows the variable distribution before
the cuts, while the green, solid filled histogram represents the
distribution after the cuts.

The global effects on nphe is

http://www.lnf.infn.it/~pisanos/dh/cc_nphe_check.pdf

I am now going to redo the higher-level analysis to see what happens to the
observables, and I will keep you updated on it.

Thanks again to all of you for your help,

silvia


-- 
 "Life was just chemistry,
  and chemistry was just physics." - Jonah Lehrer
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/e16-e1f/attachments/20161018/c8f5404a/attachment.html>