[AGIPD] Normalization of SPI hitscore by pulse energy

Description

This adds the normalization of SPI hitscores (number of lit-pixels) by XGM pulse energy exposed in the particular frame. This feature improve the hitfinding quite a bit in a case if the pulse energy varies much over the train, for example in proposal 2671.

How Has This Been Tested?

Tested offline on run 409, proposal 2671

Relevant Documents (optional)

SPB_DET_AGIPD1M-1_correct_002671_r409_240924_135147.pdf

Types of changes

• New feature (non-breaking change which adds functionality)

Reviewers

@schmidtp @ahmedk

This is a good lesson what will inevitably happen the more general analysis we integrate, and still in a much controlled environment with highly talented people getting it right more often than not.

As with the original code, this is entirely orthogonal and considered out of regular functionality. As such LGTM. The notebook changes can be hotfixed. Given EXtra-redu is imported and runs for all AGIPD notebooks, I would like to have the CI passing for AGIPD though. And it would be good if @ahmedk has a look if I missed something.

LGTM, thanks for running the tests

The notebooks are hotfixed in production, and I updated the EXtra-redu dependency.

added 1 commit

• d6f4d073 - Fix missing hitfinder parameter

Compare with previous version

So this version ended up working?

Yes

mentioned in merge request !1080 (merged)

merged

mentioned in commit bc08cedb

changed milestone to %3.15.2

mentioned in merge request !1125 (merged)