PSF

[ralphbird]

Save PSF to EA file and then record it in the DL3 file.

[TarekHC]

I already said similar things in the joint-crab repo, but I will repeat it here.

I think for the that paper we don't want to add more information. That data will be publicly released, so we would need to validate the whole 3D analysis chain...

Of course don't get me wrong, internally we need to implement the full-enclosure IRFs for scientific reasons... So this definitely needs to be done.

[ralphbird]

Agree - this is a goal for down the line, we want this in place before we release VEGAS 2.6.0 though since it make sense to only run the EAs once (or at least we want the VEGAS bit in place - the V2DL3 bit will be the easy one!).

[tony32lin]

I have some preliminary implementation in a VEGAS branch for full enclosure IRF. I follow the same implementation as that of ED and included the following new histogram:

1. pfEffArea_MC_no_theta_cut
   Effective Area (MC Energy) without theta^2 cut
2. pfEnergyBias_2D_DL3
   DL3 style energy bias E_{rec}/E_{real} vs MC Energy
3. pfEnergyBias_2D_DL3_no_theta_cut
   DL3 style energy bias E_{rec}/E_{real} vs MC Energy (no theta^2 cut)
4. pfLogAngularBias_2D_DL3_no_theta_cut
   Log10(Angular difference (rec - real) ) vs MC Energy

I think the angular difference histogram already serve as what we want here.
See below for an example of such histogram at 0deg zenith angle using Moderate cut.