ToA lnL first functionas

[tschiffer]

Started to inclued the possiblity to use the ToA information for the lnL

[Vindaar]

Later we can refactor out this type of logic from both the regular and ToA versions of this proc into a morphImpl that takes only the reference data, line energies and returns the bins and result. That way we only have the two procedures to do the "data extraction" from the relevant sources and the actual interpolation logic is shared. But I can do that at a later time.

[Vindaar]

Similarly to the above procedure, the actual logic can also be refactored out into a common procedure for regular / ToA cases. Just needs an argument that hands the internal morphing procedure to call. Again, I can do that later.

[Vindaar]

Looking good for your first lines of Nim code! 🥳

Added comments mainly for explanatory purposes.

The next steps would involve:

• modifying the LikelihoodContext object to add fields for whether we use ToA cut / ToA LnL / neither and their values here: https://github.com/Vindaar/TimepixAnalysis/blob/master/Analysis/ingrid/ingrid_types.nim#L887-L907
• assign these values in the body of initLikelihoodContext here: https://github.com/Vindaar/TimepixAnalysis/blob/master/Analysis/ingrid/private/likelihood_utils.nim#L883-L963

Once that is done, we can implement the ToA cut (as you proposed) first. Certainly simpler.

This would involve:

• add the equivalent logic for the ToA cut in this block of code: https://github.com/Vindaar/TimepixAnalysis/blob/master/Analysis/ingrid/likelihood.nim#L1309-L1338
  (You might want to add another isVetoedByToACut proc or similar that you call there. But for a start, you can just add the logic inline there)
• read the ToA data to have it for the cut in this block a bit above: https://github.com/Vindaar/TimepixAnalysis/blob/master/Analysis/ingrid/likelihood.nim#L1260-L1267
  Make sure to only attempt to read the data if we work with a timepix3 detector https://github.com/Vindaar/TimepixAnalysis/blob/master/Analysis/ingrid/ingrid_types.nim#L893 and the ToA cut is active! In the other case, just declare an empty seq[float]

If this is done and you have added the ToA cut there, we'd need to do 2 other things (but we'll do that once the implementation is there and works):

1. Add the information about how many events were vetoed by the ToA counter here: https://github.com/Vindaar/TimepixAnalysis/blob/master/Analysis/ingrid/likelihood.nim#L127-L140 (needs a counter in the many logic of course). Note that the settings which were used are automatically serialized in that writeInfos proc in the h5f.toH5(...) call.
2. In downstream code handle the resulting efficiency penalty due to the ToA cut. But for that we first need that number of course (question is whether to estimate it based on a reference dataset or use a fixed number for the beginning). This is so that e.g. in a limit calculation we know the full efficiency.