safety

Crime safety scoring and ward-level data ingestion. This app owns the data pipeline for crime statistics; it uses models from the housing app (Ward, WardSafetySnapshot, Neighbourhood) rather than defining its own.

Models

No models defined — uses housing.Ward, housing.WardSafetySnapshot, and housing.Neighbourhood.

Helper functions:

Function	Description
score_to_safety_level(score)	Maps 0–10 score → categorical level (Very Safe, Safe, …)
score_to_map_color(score)	Maps 0–10 score → hex color for map markers

Management Commands

Command	Description
fetch_crime_safety	Ingest crime_safety.json from R2 / local fallback
verify_crime_data	Validate crime data integrity and consistency

Data Flow

flowchart LR R2[R2: crime_safety.json] -->|fetch_crime_safety| WSS[WardSafetySnapshot] WSS -->|aggregation| NH[Neighbourhood.safety_score] NH --> PORTAL[neighbourhood_portal view] WSS --> REPORTS[reports app]

Key Design Decisions

• Shared models with housing — crime data lands in WardSafetySnapshot (owned by housing) to keep the single source of truth in one app.

• Score → level/color helpers are defined here rather than on the model to keep the housing app free of safety-specific display logic.

• verify_crime_data is a validation command (not a data fetch) — it checks for missing wards, score outliers, and period gaps.