search

DUNE / ND_CAFMaker

Code for making CAFs from ND inputs (whether 2x2+MINERvA prototype or full ND simulation)
Apache License 2.0
1 stars 11 forks source link

Feature/tm sconverer #9

Closed cmmarshall closed 2 years ago

cmmarshall commented 2 years ago

TMS reco/matching merge

cjbacchus commented 2 years ago

I pushed the fixes that make this work with the newly-merged SAND branch

cjbacchus commented 2 years ago

Is someone working on the most major of the comments above?

clarenc3 commented 2 years ago

@cjbackhouse I plan to address them either today or over the weekend. I agree with all the above comments for what it's worth.

Faiza is updating the track matcher so I'll wait for her and let you know

clarenc3 commented 2 years ago

Branch is now updated with a few additions:

1) Remove TF1 reliance and instead just use arithmetic with doubles 2) Update matching code to propagate the LAr track into the first layer of the TMS and do the residual on that point, instead of propagating both tracks into the dead region between the TMS and LAr. This removes reliance on the TMS unit vector for calculating the residual; the TMS unit vector can sometimes be unreliable because a straight line is a bad parameterisation of the track in the TMS due to the bending 3) Make TMS use cm as default unit for distances; LAr already does this and TMS was using mm. Also update the track matching code to know about this.

Found some bugs during this; the LAr direction vectors appear to be corrupted. Hence the matching calculates the line in LAr by using start and end points of the LAr track rather than the saved direction vector. This should be updated when the LAr unit vectors are fixed.

chenel commented 2 years ago

I think we are in a position to merge this (the ND-LAr bug above is not related to anything in this repo and will be fixed next time we respin). @clarenc3 would you do the honors?

chenel commented 2 years ago

Turns out @clarenc3 is ✈️ today, so I hereby invoke the merge. 🪄