Wacu/kenya-population-analysis

# Kenya Population Analysis ## Project description This project processes WorldPop gridded population rasters into a county level dataset for Kenya and provides a dashboard for exploring it. The pipeline downloads age and sex structured population estimates for 2021 to 2025, validates them, aggregates them from 1km grid cells to the 47 counties, and calculates demographic indicators used in health planning. No file is downloaded or edited by hand. Running one command reproduces the full output. Outputs: - `data/processed/kenya_population_by_county.csv`, one row per county per year with population counts, dependency ratios and age structure shares - `data/processed/kenya_population_age_sex.csv`, the same data broken down by five year age band and sex, used by the age pyramid in the dashboard - `data/processed/kenya_counties.geojson`, the 47 county boundaries used in the aggregation, saved so the map and the numbers stay consistent - `logs/validation_log.txt`, a record of the checks run and the decisions taken about the data - A Streamlit dashboard ## Public health context Age structure determines what a health system has to provide, and Kenya's counties vary enough that a national average is not much use for planning. Children under five drive demand for routine immunisation, growth monitoring, nutrition programmes and paediatric care. A county with a high share of under fives needs a different service mix and different staffing from one with a low share. The population aged 65 and over drives demand for chronic disease management, particularly hypertension and diabetes, and for services that most primary facilities in Kenya were not originally set up to deliver. This group is small nationally but growing, and it is concentrated in some counties more than others. The dependency ratio compares the dependent age groups to the working age population that supports them through the tax base and household income. A high ratio means more people needing care relativ …