[tmplummer]

…not match cal-prod file

Change Summary

Overview

Closes #3551. When a pointing's detector gain state matches no gain_config_id in the cal-prod file, L1C already sets its counts to zero, but its exposure stayed at full length. That bias pulls L2 map intensities low. Goodtimes now culls the whole pointing in that case, so its exposure is zero too and L2 leaves it out.

File changes

• imap_processing/hi/hi_goodtimes.py
  • New mark_unmatched_gain_config() filter. It reads mcp_delta_v, cem_a_delta_v, cem_b_delta_v and tof_v from the L1B DE global attributes and passes them to match_gain_config_id(). If that returns None (no match, more than one match, or a NaN or missing value), every time and spin bin is marked BAD_HV_VALUE and a warning is logged.
  • It runs in _apply_goodtimes_filters() right after mark_bad_voltage, using the cal-prod config goodtimes already loads.
• imap_processing/tests/hi/test_hi_goodtimes.py
  • New TestMarkUnmatchedGainConfig class.
  • Existing TestApplyGoodtimesFilters tests now patch out the new filter.

Testing

• New unit tests cover the match (no change), no-match, multiple-match, NaN, missing-attribute and existing-cull-code cases.
• All tests in tests/hi/ pass.

Closes: #3551

[Copilot]

🟢 Approval recommended

The implementation matches the issue requirements and includes comprehensive focused tests.

0 open findings

What changed in this PR

Ensures unmatched Hi gain configurations receive zero exposure, preventing biased L2 map intensities.

Changes:

• Cull entire pointings whose gain state lacks exactly one cal-prod match.
• Add coverage for matching, ambiguous, missing, NaN, and existing cull states.

File	Description
imap_processing/​hi/​hi_goodtimes.py	Adds and invokes gain-config culling.
imap_processing/​tests/​hi/​test_hi_goodtimes.py	Tests the new filter and orchestration.

🧠 Review effort: Balanced

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[leowerneck]

LGTM