Synthetic NISR-style household dataset for childhood stunting risk modelling.
Generated for the AIMS KTT Hackathon (S2.T1.2).
File
Description
raw/households.csv
2,500 households with features
raw/gold_stunting_flag.csv
300 labelled households (50/50)
raw/districts.geojson
Simplified district polygons
processed/households_scored.csv
All 2,500 households with risk scores
generate_data.py