Dashboard Overview
Real-time monitoring of stove installations & impact
Loading...
Loading...
Loading...
Loading...
Location Classification
Inactive Stoves
Customer Registry
Detailed database of beneficiaries with GPS tracking.
Registered Beneficiaries
| ID | Customer / GPS Location | Context | Stove Model | Sale Date | Price | Usage | CO₂ Saved | Status |
|---|
Stove Inventory
Current stock and sales aggregates by model
Loading Inventory...
System Settings
Configuration and Visibility Controls
🏘️ Classification Thresholds
Adjust thresholds for location classification. Changes require re-running the classification script.
📊 Kontur Population Density (people per 400m hexagon)
🌍 WorldPop Density (people per 100m pixel)
🛰️ GHSL SMOD Classification (1km grid, scientific standard)
Understanding the Classification Methodology
📐 Why These Default Values?
The default thresholds are calibrated for Senegal's settlement patterns, balancing scientific definitions with local context:
- Kontur 1,000 people/hex — Matches urban density where infrastructure (paved roads, electricity grid, water access) is typically present.
- Kontur 300 people/hex — Captures peri-urban zones: transitional areas with mixed infrastructure, often on city outskirts.
- WorldPop 50/10 people/100m — Finer resolution for validating Kontur estimates in ambiguous locations.
GHSL SMOD uses satellite imagery + population models at 1km resolution — it's the scientific gold standard for urban/rural classification and cannot be adjusted (fixed global methodology by EU Joint Research Centre).
🔗 Why Use Multiple Data Sources?
No single dataset is perfect. Our multi-source approach provides cross-validation:
| GHSL SMOD | Primary authority — scientifically validated, globally consistent |
| Kontur | High resolution (400m hex) — captures small settlements GHSL may miss |
| WorldPop | Disaggregated census data — validation layer for population estimates |
When sources agree, confidence is high. When they disagree, the classification reasoning shows which factors contributed to the final decision.
⚠️ What Happens When You Change Thresholds?
More locations classify as Rural. Use this if you want to focus resources on truly dense urban centers only.
More locations classify as Urban or Peri-Urban. Use this for programs that need broader urban coverage.
💡 Example Adjustment Scenarios
- Rural Electrification Project: Raise urban thresholds to exclude electrified areas and focus on off-grid communities.
- Urban Health Initiative: Lower peri-urban threshold to include transitional settlements that may lack clinics.
- Market Expansion Study: Keep defaults — balance between high-density commercial zones and emerging markets.
python classify_locations.py
to apply changes to the dataset. The saved thresholds will be passed as CLI arguments.
📡 Data Sources
Installed datasets and update status.
Version Control
App Version: v1.4.0 (Data Source Check)