Skip to content

Quick Start

This page uses the bundled Testkistan dataset so it works without access to the private data repository.

Prerequisites

  • Install DisruptSC with pip install -e ..
  • Activate the environment, for example conda activate dsc.
  • Leave DISRUPT_SC_DATA_PATH unset if you want to use bundled Testkistan.

1. Validate Inputs

validate-inputs Testkistan

The validator checks that the configured scope folder and required files exist.

2. Run A Baseline Simulation

disruptsc Testkistan

The default Testkistan configuration runs the simulation type configured in config/user_defined_Testkistan.yaml.

3. Override Runtime Options

# Run a disruption scenario
disruptsc Testkistan --simulation_type disruption

# Override duration and input coverage
disruptsc Testkistan --duration 12 --input_coverage 0.95

# Reuse all cached initialization stages
disruptsc Testkistan --cache same_logistic_routes

# Show all CLI options
disruptsc Testkistan --help

4. Configure A Disruption

Disruptions are configured in the scope YAML file:

simulation_type: "disruption"
disruptions:
  - type: "transport_disruption"
    description_type: "edge_attributes"
    attribute: "name"
    values: ["road_1"]
    start_time: 1
    duration: 4

Use disruptions, not events, for new configuration. The legacy events key is still accepted for compatibility.

5. Check Outputs

If export_files: true, results are saved under:

output/<scope>/<timestamp>/

Useful outputs include:

  • parameters.yaml - resolved configuration snapshot.
  • firm_table.geojson - firm locations and attributes.
  • household_table.geojson - household locations and attributes.
  • transport_edges_with_flows_*.geojson - transport flows.
  • summary.csv - aggregate loss summary when available.

Full Data Scopes

For scopes such as Cambodia, Ecuador, or Gulf, clone the private data repo as a sibling folder:

DisruptSC/
+-- disrupt-sc/
+-- disrupt-sc-data/

Then run:

validate-inputs Cambodia
disruptsc Cambodia