[Clas12_verystrange] [EXTERNAL] Electron energy correction

[Geraint Clash]

Hello all,

I have re-done the energy correction I shared with you previously but using
pass 1 data from this path:
/volatile/clas12/users/devita/2pi/rga-pass1v0/skim3/. I have used the
following function to produce the attached plots:

TLorentzVector  Correct_Electron(TLorentzVector x){

  Double_t E_new, Px_el, Py_el, Pz_el;
  TLorentzVector el_new;

  E_new =
x.E()-0.03689+0.1412*x.E()-0.04316*pow(x.E(),2)+0.007046*pow(x.E(),3)-0.0004055*pow(x.E(),4);

  Px_el = E_new*(x.Px()/x.Rho());
  Py_el = E_new*(x.Py()/x.Rho());
  Pz_el = E_new*(x.Pz()/x.Rho());

  el_new.SetXYZM(Px_el, Py_el, Pz_el, 0.000511);

  return el_new;
}

then calling the function like this  Correct_Electron(measured electron
TLorentzVector) and that goes in place of the measured electron
TLorentzVector.

The attached file is a short presentation containing the following: The
separate missing mass of everything and the proton plots fitted to show the
mean and sigma, for both the original and corrected cases. Then the missing
mass plots unfitted on the same axis where red is the original. Finally,
the delta energy plots where the corrected is on the left.

Thanks,
Geraint Clash.

On Wed, 15 Apr 2020 at 15:10, nicholas <nicholas at jlab.org> wrote:

> Hi Will,
>
> We have no indication that the angles are wrong as these are established
> from the clusters. We understand how the energy is off in the FT and I
> think at this stage an initial correction that Geraint is working on is
> sufficient to start looking at things in much more detail. Raffaella came
> up with a different energy reconstruction, which accounts for some of this
> difference, but it is not sufficient to correct for it in its entirety. By
> no means the correction that Geraint will send soon is the final one and he
> will work on more studies using higher stats. I think the way the
> correction is determined/applied is not very crucial especially since in
> your plot there is still a significant slope and it doesnt seem to be a
> scale factor independent of E.
>
> Cheers,
> -N
>
>
> On 15 Apr 2020, at 15:00, William Phelps <wphelps at jlab.org> wrote:
>
> Hello Geraint,
>
> I have been thinking about this electron correction (briefly). It seems to
> me that it would make sense to have the correction as a ratio
> P_missing/P_reconstructed as a function of P_reconstructed. So all we would
> need to do is scale the electron momentum (and recalculate energy). You
> mentioned that you are going to take another look at your correction during
> the meeting, is this what you are thinking about doing?
>
> I attached a 2d histogram that might make more sense than what I said in
> the paragraph above.
>
> A secondary question to the group: If the reconstructed momentum is
> incorrect does this mean that the angle is also incorrect?
>
> Thanks,
> -Will
>
> <Plot_04-15-2020_09.53.13_AM.png>
>
> On Apr 3, 2020, at 9:58 AM, Geraint Clash <geraintclash at gmail.com> wrote:
>
> Hello all,
>
> I carried out a study on the difference between the energy of the electron
> when you treat it as missing and the measured electron using skim3 data
> from this path: /volatile/clas12/users/clas12/rich/dst/recon/005038/. For
> this, the exclusive reaction ep->e’p pi+pi- was selected, with three-sigma
> cuts around missing mass squared of all products and the missing mass of
> the proton.
> The plot of the difference in energy ("missing" electron - detected
> electron) against the electron measured energy is the attached file Del_E.
> Y-projections were fitted with Gaussians and the means were fitted with a
> fourth-order polynomial establishing the correction function (see attached).
>
> The energy correction which should be added to the measured electron
> energy, E, is: -0.0004x^4+0.0071x^3-0.0432x^2+0.1356x-0.0257.
>
> Thanks,
> Geraint Clash.
> <Del_E_mean_fit.png><Del_E.png>
> _______________________________________________
> Clas12_verystrange mailing list
> Clas12_verystrange at jlab.org
> https://mailman.jlab.org/mailman/listinfo/clas12_verystrange
>
>
> _______________________________________________
> Clas12_verystrange mailing list
> Clas12_verystrange at jlab.org
> https://mailman.jlab.org/mailman/listinfo/clas12_verystrange
>
>
> _______________________________________________
> Clas12_verystrange mailing list
> Clas12_verystrange at jlab.org
> https://mailman.jlab.org/mailman/listinfo/clas12_verystrange
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/clas12_verystrange/attachments/20200415/914e1182/attachment-0002.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: E_corrections.pptx
Type: application/octet-stream
Size: 801138 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/clas12_verystrange/attachments/20200415/914e1182/attachment-0002.obj>