speed up last_instance in tcm by precomputing keys and vectorizing

[gipert]

• chore: add debug logging for tcm progress in generate_tcm_cols
• perf: speed up last_instance in tcm by precomputing keys and vectorizing

[codecov]

Codecov Report

❌ Patch coverage is 93.75000% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 52.56%. Comparing base (e26fd48) to head (c2f1eb6).
⚠️ Report is 3 commits behind head on main.

Files with missing lines	Patch %	Lines
src/pygama/evt/tcm.py	93.75%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #653      +/-   ##
==========================================
+ Coverage   52.51%   52.56%   +0.04%     
==========================================
  Files          60       60              
  Lines        8270     8274       +4     
==========================================
+ Hits         4343     4349       +6     
+ Misses       3927     3925       -2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Copilot]

Pull request overview

This PR improves the performance and observability of TCM generation in pygama.evt, focusing on speeding up the “last instance” computation during coincidence clustering and adding debug-level progress logging.

Changes:

• Precompute and cache table_keys mapping helpers to support faster per-iteration processing.
• Vectorize “last occurrence per table_key” computation using NumPy (searchsorted + maximum.at) instead of a per-hit Python loop.
• Add debug logging summarizing TCM progress per iteration and remove some noisier debug statements.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.