Context: Cases API lacked demographic breakdowns and per-diagnosis
monthly seasonality data for epidemiological analysis.
Approach:
- Two new GET endpoints with typed Pydantic response models
- Demographics uses get_inpatient_data() (outpatient lacks gender/age)
- Disease-seasonality uses get_combined_data() grouped by diagnosis+month
- pandas groupby/value_counts for vectorized aggregation
Changes:
- Models: AgeBin, GenderSplit, GenderSplitData,
AgeDiagnosisMatrixItem, DemographicsResponse
- Models: DiseaseSeasonalityPoint, DiseaseSeasonalityResponse
- GET /api/cases/demographics: age distribution (0-17), gender split,
age-diagnosis matrix (5 age groups)
- GET /api/cases/disease-seasonality: top 10 diagnoses by month
(120 entries with 1月-12月 labels)
Impact: Enables frontend demographic charts and disease seasonality
heatmaps. All 42 existing API tests continue passing.
Akiba So
f58f2f612c