Optimal cut for PSD

[fanurs]

Previously, we put in a cut of psd > 0.5 & psd < 2.0 & psd_perp > -0.5 & psd_perp < 0.5.

It turns out that this previous cut was too narrow, throwing away neutron data points. To see why, we look at the original data:

Points above psd > 2.0 roughly correspond to the "appendix" circled above. We thought these are "bad data points". But according to a plot from S. Mouatassim et al. / Nucl. Instr. and Meth. in Phys. Res. A 359 (1995) 530-536, they are actually from neutrons that knock out helium-3 or helium-4 from carbon-12:

Geant4 simulation verifies this hypothesis too:

So perhaps a better gate would be psd > 0.5 & psd < 4.0 & psd_perp > -0.7 & psd_perp < 0.7?

[fanurs]

The latest commit has already fixed all known PSD issues. A gate of NWB_psd > 0.5 || NWB_total_L > 3500 || NWB_total_R > 3500 would be sufficient.