Uncovering Kenya's invisible digital deserts — county-level Digital Desert Index (DDI) for 47 counties | ITU Data Hackathon 2026
`linka-digital-desert-index`
`Uncovering Kenya's invisible digital deserts — county-level Digital Desert Index (DDI) for 47 counties | ITU Data Hackathon 2026`
---
## Folder structure
```
linka-digital-desert-index/
│
├── README.md
├── LICENSE
├── .gitignore
│
├── data/
│ ├── raw/
│ │ └── linka_kenya_county_data.csv
│ └── processed/
│ ├── linka_ddi_results.csv
│ └── linka_typology_summary.csv
│
├── analysis/
│ └── linka_analysis.py
│
├── visualisation/
│ └── linka_dashboard.html
│
├── docs/
│ ├── linka_policy_brief.md
│ └── linka_etl_diagram.png
│
└── presentations/
├── linka_step1_etl.pptx
└── linka_step2_intermediate.pptx
```
---
`# LINKA — Kenya Digital Desert Index`
Add an ITU hackathon badge and a Kenya flag emoji for instant context.
Kenya has ~74% 4G coverage but only 35% actual internet use. Linka builds a Digital Desert Index (DDI) for all 47 counties to reveal where connectivity exists on paper but not in practice — and why. Built for the ITU Data Hackathon 2026.
**The DDI formula**
```
DDI = (coverage_gap + affordability_gap + electricity_gap + skills_gap) / 4
```
One sentence explaining each gap. Scores range from 0 (connected) to 1 (excluded).
**Key findings** — three bullet points maximum:
- West Pokot scores 0.67 (worst), Nairobi 0.17 (best)
- 42 of 47 counties exceed the ITU 2% GNI affordability target
- Education index (r = 0.91) is the strongest predictor of internet use
**Data sources table** — a simple markdown table with five rows: ITU DataHub, KNBS 2023/24, World Bank WDI, Kenya Shapefiles (HDX), VIIRS/WorldPop
**How to run the analysis**
```bash
git clone
github.com
cd linka-digital-desert-index
pip install pandas numpy geopandas
python analysis/linka_analysis.py
```
**File guide** — one line per file explaining what it does
**Team and attribution**
Team Linka · ITU Data Hackathon 2026 · Organised by ITU BDT with EU support
---
##
```
kenya digit …