[Olympus] [Olympus-lumi] Follow-Up on Today's 12 Degree Discussion

[Michael Kohl]

Hi,
I would like to add that whenever new plots are shared to be discussed 
then the codes which were used to generate the plots need to be shared, 
too. The best way is to push them into the respective git repo (e.g. 
olympus_lumigem).

Without sharing the codes, meaningful discussions are hardly possible.
When presenting plots in the meeting(s), the slides should contain all 
necessary info to understand them, in order to avoid the storm of 
questions during the presentation, like it happened today.

Best regards
   Michael



On Mon, 18 Aug 2014, Brian S. Henderson wrote:

> Hello all,
>
> I have put together a few plots to examine the issues we discussed during 
> Kirill and Denis's presentation today, specifically the MC/data comparisons 
> on Slide 3.  The four plots attached shown comparisons of the reconstructed 
> theta with any four tracking for the combinations electrons/positrons in the 
> left/right arms, for two runs of each species.  The version of LumiFit I 
> used today should be essentially identical to that used by Kirill and Denis 
> last week. On each plot: red is the data, blue is the MC according to the SC 
> lumi, cyan is the MC with a scaling factor, and green is the MC 
> /without//selecting on the lumi trigger conditions/ (neither requiring a ToF 
> hit nor a hit in both SiPMs).
>
> Regarding what I did to generate these plots: the generator was run under 
> identical conditions to those specified in the talk this morning with the 
> exception of the soft fraction, which I left on the default value of 0.5. 
> The standard lumi hit finding and digitization were run as they exist in 
> olympus_release, and then both were tracked using the version of LumiFit 
> currently in olympus_lumigem, using the fitlumis4.xml and fitlumisMC4.xml 
> (the only difference between these recipes is the application of offsets, 
> which will be phased-out in the near future).  I then ran my 12 degree 
> analysis program "GEMLumiNoPro.xml" (which doesn't require proton tracks) to 
> make the histograms, and re-ran GEMLumiNoPro for the MC tracks with the 
> trigger selection conditions removed.
>
> During today's discussion, there were two primary questions regarding the 
> theta plot on the right side of Slide 3: the overall normalization of the 
> data relative to MC and the relative excess of lower theta (between 10-11 
> degrees) of the MC relative to the data. As can be seen in the plot 
> "elefttheta.png", this effect seems to be caused by not properly accounting 
> for the trigger in the processing of MC data.  When the trigger is selected 
> on in the analysis, the theta distributions agree reasonably between data 
> and MC in both the left and right arms.
>
> The overall scale, however, remains a question.  As noted, Denis and Kirill 
> saw the MC distribution to be about 1.5 or 2 times larger than the data 
> distribution, but this doesn't seem to be due to the trigger issue since 
> leaving out the trigger doesn't increase the "peak" of the MC distribution. 
> Is it possible the MC data set included more events than expected or 
> multiple runs?
>
> Based on this, I have a few questions for Kirill and Denis:
>
> * At what stage of your analysis and in what way is the trigger
>   selection done for MC events?
> * When you run the generator for a single run, do you change the
>   numbers of events, and if so how?
> * Is your analysis code currently available in the repos so that
>   people can take a look at it?
>
>
> I have full sets of kinematic distributions for these runs using the latest 
> tracking, so let me know if there is anything you would like to see.
>
> Thanks,
> Brian
>

+---------------------------------------------------------------------
| Dr. Michael Kohl, Associate Professor and Staff Research Scientist
| Physics Department, Hampton University, Hampton, VA 23668
| Jefferson Lab, C117, 12000 Jefferson Avenue, Newport News, VA 23606
| Phone: +1-757-727-5153 (HU), +1-757-269-7343 (Jlab)
| Fax:   +1-757-728-6910 (HU), +1-757-269-7363 (Jlab)
| Email: kohlm at jlab.org, Cell: +1-757-256-5122 (USA)
+---------------------------------------------------------------------