Problem
Observed prices, forecasts, deterministic gross margins, and predicted realized margins must remain distinguishable when they are exposed through one economic-planning workflow.
Status: Blocked
Blocked by:
Implementation must not start until all prerequisite contracts and model cards are published.
Proposed behavior
- Keep observed prices as the default.
- Require explicit opt-in for forecast values.
- Return deterministic gross margin separately from predicted realized gross margin.
- Carry source type, horizon, interval, model version, training cutoff, dataset version, and freshness through every scenario.
Acceptance criteria
- Every price identifies whether it is observed, user supplied, quoted, imputed, or forecast.
- Forecast scenarios include horizon and lower/upper intervals.
- Observed and forecast values are never silently combined.
- Scenario ranking occurs only when required data is fresh, supported, and comparable.
- Ranking is suppressed when data is stale, unsupported, or uncertainty exceeds a documented threshold.
- Deterministic and predicted margins use different fields and labels.
- Python, JSON, CSV, CLI, and batch outputs preserve all provenance and uncertainty metadata.
- Tests cover opt-in behavior, stale data, unsupported series, wide intervals, and mixed observed/forecast scenarios.
Problem
Observed prices, forecasts, deterministic gross margins, and predicted realized margins must remain distinguishable when they are exposed through one economic-planning workflow.
Status: Blocked
Blocked by:
Implementation must not start until all prerequisite contracts and model cards are published.
Proposed behavior
Acceptance criteria