Native summary pipeline lacks legacy feature columns (shifted capacity, RIC, cumulated CE)

[jepegit]

Context

Surfaced during the native-headers flip (Stage 5a, flipping native_schema=True as default). The native cellpycore summary pipeline produces a different column set than the legacy summarizer — not just renamed columns:

Absent in native summary (present in legacy):

• shifted_charge_capacity / shifted_discharge_capacity
• cumulated_ric / cumulated_ric_sei / cumulated_ric_disconnect (RIC)
• cumulated_coulombic_efficiency
• summary date_time

Renamed semantically (not in cellpycore.legacy.mapping):

• legacy cumulated_charge_capacity → native test_cumulated_charge_capacity (a "test-level cumulated" concept)

These are real electrochemistry features (shifted capacity, RIC) the native summarizer does not compute. The D6 header shim cannot paper over them — the columns don't exist.

Decision (Stage 5a)

Stage 5a proceeds with the mechanical naming migration + golden regen; the ~10 tests that assert these columns are marked xfail/skip referencing this issue. This tracks the follow-up.

Work (later)

• Decide per column: compute in the native cellpycore summarizer (feature parity), reconcile the cumulated_* ↔ test_cumulated_* naming in the mapping, or deprecate/remove in cellpy 2 and document in the migration guide.
• Remove the corresponding xfail/skip markers once resolved.

🤖 Generated with Claude Code