Dashboard Overview

Real-time monitoring of stove installations & impact

Admin User
A
Active Stoves

Loading...

Total Revenue

Loading...

CO₂ Saved

Loading...

Customers

Loading...

Location Classification

Regional Installations

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)

≥ value = Urban
≥ value = Peri-Urban

🌍 WorldPop Density (people per 100m pixel)

≥ value = Urban
≥ value = Peri-Urban

🛰️ GHSL SMOD Classification (1km grid, scientific standard)

30 = Urban Centre
23 = Dense Urban Cluster
22 = Semi-Dense Urban
21 = Suburban
13 = Rural Cluster
11-12 = Low Density Rural

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?

⬆️ Raising thresholds = Stricter definition of "urban"
More locations classify as Rural. Use this if you want to focus resources on truly dense urban centers only.
⬇️ Lowering thresholds = More inclusive "urban" definition
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.
Note: After saving new thresholds, you must re-run 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.

Loading data source info...
View Full Documentation

Version Control

App Version: v1.4.0 (Data Source Check)