[Frost] FW: [EXTERNAL] Update on Scale Factor

[Michael Dugger]

Chan Kim,

You should contact the FROST group directly instead of routing through
Eugene.

It does not seem like you took ANY advice regarding the carbon scaling.
You will save a bunch of time if you listen to the advice given to you by
people that spent years looking at FROST data. Otherwise, you will end up
inventing the wheel over and over.

You should first look to see if the carbon from the run period without ice
contamination has any peaking structure underneath your signal region. If
no peaking structure, you can probably get away with simply using a
polynomial background that would include the bound nucleon contribution to
your signal. If you are lucky, you can get away with analyzing the data
without doing a carbon subtraction.

If you do need to use carbon subtraction, you should at least look at the
CLAS-Note on the construction of scale factors. That CLAS-Note can be
found at

https://misportal.jlab.org/ul/Physics/Hall-B/clas/viewFile.cfm/2015-001.pdf?documentId=744

If you have to use carbon subtraction for the data that has ice
contamination, you are going to have a whole bunch of problems that have
nothing to do with the type of fitting routine used. The problems are
bigger than tweaks associated with splines and least square minimization.

If you are not careful, you will end up spinning your wheels and getting
nowhere.

Take care,
Michael

>
>
> -Eugene
>
> From: Chan Kim <kimchanwook at gwmail.gwu.edu>
> Sent: Sunday, November 10, 2019 02:22
> To: Igor Strakovsky <igor at gwu.edu>; Eugene Pasyuk <pasyuk at jlab.org>;
> Stuart Fegan <sfegan at jlab.org>
> Subject: [EXTERNAL] Update on Scale Factor
>
> Hello All,
>
> For computing scale factor, I used splines interpolation (as Hao Jiang
> from USC) and below plot is the result (x-axis is Missing Mass Square).
>
> First, I used splines polynomials to fit carbon at specific x intervals.
> The, I used that splines polynomials + gaussian to fit butanol.
> While fitting the butanol using the carbon splines polynomials, it gives
> out the how much the splines polynomials need to be scaled to make a good
> fit on butanol, which I can use as the scale factor.
>
> The plot is a result of only one of a energy + angle bin, and it yet
> doesn't work on all energy/angle bins. I now will use least square
> minimization to see if this method works on all bins..
> Please let me know what you think.
>
> Thank you,
> Chan
>
> [background_subtract.png]
> _______________________________________________
> Frost mailing list
> Frost at jlab.org
> https://mailman.jlab.org/mailman/listinfo/frost