electricsheepafrica/african-demographic-projections

---
license: cc-by-4.0
tags:
- demographics
- population
- projections
- sub-saharan-africa
- synthetic
- urbanization
- africa
- world-population-prospects
- scenario-analysis
task_categories:
- tabular-regression
pretty_name: African Demographic Projections (2020-2050)
language:
- en
size_categories:
- 1K<n<10K
---

# African Demographic Projections (2020-2050)

## Dataset Summary

Synthetic demographic projections for 20 Sub-Saharan African countries spanning 2020-2050 under three policy scenarios. Calibrated against UN World Population Prospects 2024 (medium variant) with stochastic perturbations to enable scenario analysis.

## Scenarios

| Scenario | Description |
|---|---|
| `baseline` | UN WPP 2024 medium variant projections with minor stochastic variation |
| `accelerated_urbanization` | Faster urbanization (+60% urban growth rate), slightly lower fertility, higher GDP growth |
| `fertility_decline` | Faster fertility decline, higher life expectancy, age-structure shift toward working age |

## Countries (20)

Nigeria, Ethiopia, DR Congo, Tanzania, Kenya, Uganda, Sudan, Mozambique, Ghana, Angola, Cameroon, Niger, Mali, Malawi, Zambia, Senegal, Chad, Somalia, Zimbabwe, Rwanda

## Variables (17)

| Variable | Description | Unit |
|---|---|---|
| `record_id` | SHA-256 hash (16 char) of country+year+scenario | — |
| `country` | Country name | — |
| `year` | Projection year | 2020–2050 |
| `scenario` | Projection scenario | — |
| `total_population_millions` | Total population | millions |
| `urban_population_pct` | Urban population share | % |
| `fertility_rate` | Total fertility rate | children/woman |
| `mortality_rate` | Crude death rate | per 1,000 |
| `life_expectancy` | Life expectancy at birth | years |
| `dependency_ratio` | (youth + elderly) / working age × 100 | % |
| `youth_pct` | Population aged 0–14 | % |
| `working_age_pct` | Population aged 15–64 | % |
| `elderly_pct` | Population aged 65+ | % |
| `net_migration_rate` | Net migration | per 1,000 |
| `population_growth_rate` | Annual population growth | % |
| `median_age` | Median age | years |
| `gdp_per_capita_usd_ppp` | GDP per capita (PPP) | USD |
| `urbanization_rate_annual` | Annual change in urban share | % |

## Dataset Structure

- **Records**: 1,860 (20 countries × 3 scenarios × 31 years)
- **Format**: CSV (`data/african_demographic_projections.csv`)
- **Seed**: `random.seed(42)` for reproducibility

## Calibration Sources

Projections are calibrated to the **UN World Population Prospects 2024 Revision** medium variant for baseline trajectories. Scenario deviations are modeled as proportional adjustments to urbanization, fertility, life expectancy, and GDP growth trajectories.

Key references:
- UN DESA (2024). *World Population Prospects 2024*. https://population.un.org/wpp/
- UN DESA (2018). *World Urbanization Prospects 2018*. https://population.un.org/wup/

## Usage

```python
from datasets import load_dataset
ds = load_dataset("electricsheepafrica/african-demographic-projections")
df = ds["train"].to_pandas()
```

## Generation

```bash
pip install -r requirements.txt
python generate_dataset.py
python validate_dataset.py
```

## License

CC-BY-4.0. This dataset is synthetic; the underlying calibration parameters are derived from publicly available UN statistics.