StartDirection and EndDirection switched, at least in main

[jdkio]

The starting direction of the reco tracks seems to be bifurcated. Here is reco-true for the xz angle. These are LAr-starting muons that end in the TMS, and we're measuring reco-true angle at the front plane of the detector. Using file /exp/dune/data/users/kleykamp/dune-tms/2024-09-24_test_23rd_version.tmsreco.root and branch kleykamp_issue_158

This is RHC so we expect muons and antimuons in approximately equal numbers. So this looks like what you'd expect for the end direction. If we use EndDirection (actually EndDirectioN in these files) instead, then it looks much better:

I know that kalman tracks start from the back, so I think the direction is being calculated from the back nodes instead of the front. I haven't checked if this is fixed in branch 3D-Kalman-fitter but that's my next stop.

That said, using this as a charge discriminator looks decent

[AsaNehm]

Has this issue been solved with the merge?