mc_truth/interaction + true_pos in prompt_hit + backtrack

[YifanC]

• Fill mc_truth/interactions with vertices when there is no mc_hdr available. (closed PR mc_truth/interactions #126 )
• Added true drift position and true energy reconstructed hits. (closed PR add true position and E to hits dataset #127 )
• Adapted to use configuration instead hardcoded numbers for hit reconstruction (closed PR add true position and E to hits dataset #127 )
• Copy the mc_packets_assn from larndsim output to mc_truth/packet_fraction and propagate it to mc_truth/calib_prompt_hit_backtrack
• Adapted calib_hit_merger.py to consume the updated mc_truth/calib_prompt_hit_backtrack instead.
• Use index for referencing between the packet and segment instead of taking segment_id as indices for references. Together with this PR for larndsim, we can get rid of the hotfix for segment updates.

Note:

• event_ids are now available in mc_truth/packet_fraction and mc_truth/calib_prompt_hit_backtrack. event_id is not yet available in mc_truth/calib_final_hit_backtrack (to be added). mc_truth/calib_final_hit_backtrack data type is still defined manually, which can be updated too.
• In the backtracking datasets, segment_ids is used instead segment_id given this is what's been called in larnd-sim. We can change the name in larndsim if it's preferred. In truth dataset, we use segment_id.
• The data and ref all look reasonable to me. From validating the output ofmc_truth/calib_final_hit_backtrack, I noticed some differences using this feature branch vs. using develop (commit f58217e). The differences are inherited from mc_truth/calib_prompt_hit_backtrack. Given in this feature branch it's a direct copy from the larndsim ouput and the develop version reshape and remask a few times, I am convinced the problem is in the develop branch, but I cannot explain what was wrong.

[YifanC]

Minor comments in Git Issue #129

[YifanC]

Some screenshots of some output: https://indico.fnal.gov/event/64988/contributions/293232/attachments/178588/243452/2024_06_04_ndlar_flow_readiness.pdf

[mjkramer]

Validation plots look identical to MR5-beta1

[YifanC]

Validation plots look identical to MR5-beta1

Cool! @mjkramer Thanks for checking! I think we should add some backtracking part in the sim flow validation plots. (The backtracking info are different in beta1 and this version. The beta1 or current develop version doesn't seem to be correct.)

[diaza]

Could I ask what "true energy reconstructed hits" means? Is it the true charge deposited on the pixels, or the true dE deposited in the argon?

[YifanC]

@diaza E_true_recomb_elife means using true drift time and true dEdx for recombination to correct the energy of the hits. Note that I manually put the true lifetime in the config.