Playground

Six concrete scenarios to get started with GISPulse in 5 minutes, on real IGN BD TOPO V3 + DVF Etalab data (Toulouse, Clermont-Ferrand, Versailles).

Each scenario is a self-contained JSON pipeline: 2-4 steps, runnable with a single CLI command and rendered step by step on an interactive map (demo backend required).

Install

pip install gispulse
# or from source
git clone https://github.com/imagodata/gispulse && cd gispulse
pip install -e ".[dev]"

Download data

# All 3 cities (recommended)
python examples/prepare_playground_data.py

# Or a single city
python examples/prepare_playground_data.py --city toulouse
python examples/prepare_playground_data.py --city clermont-ferrand
python examples/prepare_playground_data.py --city versailles

City	Buildings	Roads	Facilities	Specificity
Toulouse	~31,000	3,680	393	Garonne corridor, dense center
Clermont-Ferrand	~5,000	2,272	590	Compact network, relief
Versailles	~5,000	1,617	405	509 vegetation zones

Scenarios

IntermediaireCLI + Map

S1 — Flood Risk

Toulouse — low-rise buildings (<=15 m) in the 250 m Garonne corridor, ground 0-15 m above water level (altitude_minimale_sol, BD TOPO V3). 4 steps.

filter

DebutantCLI + Map

S2 — Commercial Buildings along Arterials

Toulouse — commercial buildings (usage_1 or usage_2 == 'Commercial et services') within 50 m of an IGN arterial road (importance 2-4 — national, departmental, main urban). 3 steps — each step (filtered roads, buildings in buffer, commercial) is a visible layer.

filter

AvanceCLI + Map

S3 — Health Facility Accessibility

Clermont-Ferrand — 10-min walking isochrones (833 m) on the BD TOPO network from health facilities (categorie == 'Santé'), multi-source Dijkstra, metric coverage in m². (Pro.)

filterisochronearea_length

IntermediaireCLI + Map

S4 — Main Road Network + Urban Setback

Clermont-Ferrand — filter by importance (1-4), Lambert93 length, indicative 100 EUR/m cost. DML trigger: draw a building inside the visible 50 m setback zone around motorways/national roads and see the L111-6-inspired cascade fire (FLAG, LOG, NOTIFY).

filterarea_lengthcalculate

IntermediaireCLI + Map

S5 — Park accessibility per building

Versailles — BD TOPO vegetation + buildings: parks ≥ 1 ha (SCoT IdF), nearest_neighbor distance from each residential building to the closest park, manual classification against WHO / SCoT / ADEME thresholds (300 / 600 / 1000 m). Green-to-red building choropleth. Weekly cron inside the pipeline.

area_lengthfilternearest_neighborclassify

IntermediaireCLI + Map

S6 — Price-per-m² Map (DVF)

Versailles — Etalab DVF mutations 2022-2024, filter residential sales, compute price/m², quintile classes + YlOrRd palette. Color gradient by price-per-m² rendered live on the map.

filtercalculateclassify

Ready-to-use pipelines

Pipeline v2 format (DAG with steps, triggers, ref_layers), validated by JSON Schema on load.

Scenario	Pipeline	Trigger
S1 Flood risk (4 steps)	scenario-1-rules.json	—
S2 Commercial / arterial roads (3 steps)	scenario-2-rules.json	—
S3 Health accessibility (3 steps)	scenario-3-rules.json	—
S4 Road network (3 steps)	scenario-4-rules.json	scenario-4-trigger.json
S5 Park accessibility (5 steps, 2 branches)	scenario-5-rules.json	weekly cron (inside rules)
S6 Price-per-m² DVF (8 steps)	scenario-6-rules.json	—

Capabilities out of scope here

spatial_aggregate, dissolve, connectivity_check, shortest_path, zonal_stats, ndvi, etc. are covered by the templates gallery and the capabilities guide. The 6 playgrounds favour short, linear workflows (2-8 steps each).