Which equivalent-full-cycle convention should a grid-battery example expose?

[mohammadrezwankhan]

The Oxford energy-trading example in Battery Cycle-Life Analyzer aligns each
capacity check with the cell's measured current profile. The capacity files
contain elapsed profile time, so they cannot simply be treated as laboratory
cycle counts.

The current adapter integrates positive discharge current and divides by the
stated nominal capacity of 16 Ah:

EFC(t) = integral(max(I(t), 0) dt) / (3600 * Q_nominal)

It reports duplicate/reversed timestamp repairs and limits any terminal-time
adjustment. The example also reports the small chronological holdout and
bounded extrapolation rather than presenting the fit as broad validation.

For comparing mixed grid-service profiles, which convention would you expect
an analysis API to expose?

1. Discharge throughput divided by nominal capacity (the current default).
2. Half of absolute charge-plus-discharge throughput divided by nominal capacity.
3. Discharge throughput divided by measured initial capacity.

What metadata would you require to compare results across cells, and should
alternative conventions be explicit options or separate derived columns?

Implementation and documented assumptions:
https://mohammadrezwankhan.github.io/battery-cycle-life-analyzer/oxford-energy-trading.html

Merged implementation:
#7

The Oxford data remain separately licensed and are downloaded from the original
archive only when the example runs; they are not bundled in this repository.