Sample Dashboards

Sample Data and Demo Dashboards

GeoSet ships with a self-contained demo stack that loads real geospatial data and pre-built dashboards so you can explore the platform without setting up your own data sources.

What's Included

The demo stack loads three datasets from public APIs:

Dataset	Source	Table
US State Boundaries	Census Bureau API	census_state_boundaries
Active Wildfire Locations	NIFC ArcGIS API	nifc_wildfire_locations
Hurricane Best Track	NHC/NOAA API	nhc_best_track

These datasets power two pre-built dashboards:

Hurricane Tracker

A multi-layer map combining:

• Storm Track Points — NHC best track observation points colored by max wind speed (yellow → red gradient)
• Storm Track Lines — aggregated best track lines per storm per year
• Census State Boundaries — US state boundary polygons

Includes native dashboard filters for Hurricane Season (year) and Hurricane Name (cascades from Season).

Wildfire Tracker

A multi-layer map combining:

• NIFC Wildfire Locations — active wildfire points colored by fire cause (Human, Natural, Undetermined)
• Census State Boundaries — US state boundary polygons

Includes a native dashboard filter for Fire Cause.

Running the Demo Stack

The demo uses a separate Docker Compose file (docker-compose-geoset.yml) that is fully independent from the main docker-compose.yml. It runs on different ports and separate volumes so the two stacks don't interfere.

# Start the demo stack
docker compose -f docker-compose-geoset.yml up

# Start with a full rebuild
docker compose -f docker-compose-geoset.yml up --build

# Stop
docker compose -f docker-compose-geoset.yml down

# Stop and wipe all data (full reset)
docker compose -f docker-compose-geoset.yml down -v

Access the demo at http://localhost:9001 with credentials admin / admin.

The data ingest container runs once and is idempotent — if the tables already have data, the scripts skip them.

Demo Stack Architecture

Browser (:9001)
    │
    ▼
superset-node-geoset       ← Webpack dev server (hot reload)
    │
    ▼ API calls
superset-geoset (:8088)    ← Flask backend
    │
    ├──▶ db-geoset          ← PostgreSQL (Superset metadata)
    └──▶ postgis (:5433)    ← PostGIS (geospatial data)
              │
              ▼
    geoset-sample-data-ingest  ← One-shot data loader

Adding a New Example Dataset or Dashboard

1. Add the Data

1. Add a CREATE TABLE statement to sample-data/init.sql
2. Write an ingest script in sample-data/ that fetches and inserts the data. Use utils.py for the DB connection and call skip_if_populated() to keep it idempotent
3. Register the script in sample-data/entrypoint.sh

2. Create and Export the Chart

1. Build the chart in the Superset UI using your new dataset
2. Export it from Charts → Actions → Export (downloads a ZIP with YAMLs)
3. Copy the YAMLs into:
   • superset/examples/geoset_configs/charts/GeoSet/
   • superset/examples/geoset_configs/datasets/geoset/
4. Strip the query_context field from the chart YAML (it contains instance-specific IDs)

3. Export and Add the Dashboard

1. Build the dashboard in the Superset UI
2. Export it from Dashboards → Actions → Export
3. Copy the dashboard YAML into superset/examples/geoset_configs/dashboards/

Multi-Layer Map Charts

Multi Map charts reference sub-layer charts by integer ID in deck_slices. Since IDs vary per instance, use deck_slice_uuids with stable UUIDs in the YAML instead. The post-import hook resolves these to runtime IDs automatically.

4. Rebuild

docker compose -f docker-compose-geoset.yml up --build

The importer walks the entire geoset_configs/ directory — all dashboards, charts, and datasets are picked up automatically.