Road-Network Imports for METROPOLIS2

This repository contains Python scripts to import a road network from various data providers (although only OpenStreetMap is supported for now).

Compatible METROPOLIS2 version: 1.0.0

Requirements

To install the required Python packages, you can either use pip (it is recommended to use a virtual environment):

pip install -r requirements.txt

Or create a new conda environment with:

conda env create -f environment.yml

The code has been tested with Python version 3.12.2 only.

Configuration

The config.toml file is used to specify all the configuration variables (paths to the input / output files, default values for some variables, etc.).

By default, the script will read the config.toml file in the current working directory. To specify a different configuration file, you can use the --config command-line option. For example:

python python/write_metropolis_edges.py --config my-config.toml

Step 1: Download the input data

OpenStreetMap

To import a road network from OpenStreetMap data, you need to obtain a .osm.pbf file of your study area. One way to do so is to download an extract for your region on Geofabrik. Check the OSM wiki for more solutions.

Step 2: Import the data

This step consists in reading the input data and creating a GeoDataFrame with the imported edges.

OpenStreetMap

For OpenStreetMap, you need to run the script import_osm.py:

python python/import_osm.py

The config.toml file needs to contain the following key-value pairs:

• crs: The projected coordinate system to use for metric operations (e.g., computing edges's length). The value can be anything that is accepted by the pyproj package. You need to make sure to use a projection that is adapted for you study area. For France, the recommanded CRS is "EPSG:2154" (Lambert).
• raw_edges_file: Path to the file where the imported edges will be stored. Accepted formats: Parquet, GeoJSON, FlatGeobuf, GeoPackage, Shapefile.
• osm.input_file: Path to the .osm.pbf file to import from.
• osm.highways: List of highway tags to import.
• osm.urban_landuse: Optional list of landuse tags to be classified as urban areas. If ommitted, the edges will not be classified between urban and rural.

The ouput file is a GeoDataFrame with the following columns:

• geometry: LineString of the edge (using EPSG:4326 coordinates).
• source: OSM id of the source node.
• target: OSM id of the target node.
• length: Length of the edge, in meters.
• speed: Speed limit of the edge, in km/h (if specified in OSM).
• lanes: Number of lanes on the edge (if specified in OSM).
• urban: Whether the edge is in a urban area.
• osm_id: OSM id of the corresponding way in OSM (not unique).
• name: Name of the edge (if specified in OSM).
• road_type: OSM highway tag of the edge.
• id: Unique identifiers for the edges (created by the script).

Step 3: Post-processing

This step consists in some additional computations, common to all data providers to make the road network ready to use with METROPOLIS2.

To run this step, execute the script postprocess_network.py:

python python/postprocess_network.py

The following operations are performed:

• Default values are set for the speed, lanes and capacity variables so that there is no null values.
• If parallel edges are found, only the one with the largest capacity (smallest travel time in case of tie) is kept.
• The edges which are not part of the largest strongly connected component of the graph are removed (unless the postprocess_network.ensure_connected configuration parameter is set to false, in which case some origin-destination pairs might not be feasible in METROPOLIS2).
• The edges are re-indexed so that all ids are unique and range between 0 and n-1 where n is the number of edges.
• Lower bounds are enforced for the number of lanes, the speed-limit and the length.
• The target_count variable is computed: number of incoming edges for the target node of the edge.

The config.toml file needs to contain the following key-value pairs:

• raw_edges_file: Path to the file where the imported edges are stored. Accepted formats: Parquet, GeoJSON, FlatGeobuf, GeoPackage, Shapefile.
• clean_edges_file: Path to the file where the edges will be stored after postprocessing. This path can be the same as raw_edges_file if you just want to replace the imported file. Accepted formats: Parquet, GeoJSON, FlatGeobuf, GeoPackage, Shapefile.
• postprocess_network.default_speed: Default speed limits for urban and rural areas, for each road_type.
• postprocess_network.default_nb_lanes: Default number of lanes, for each road_type.
• postprocess_network.default_capacity: Default bottleneck capacity, for each road_type.

Step 4: Writing METROPOLIS2 input

This step will write a CSV or Parquet file of the road network, compatible with the METROPOLIS2 simulator.

To run this step, execute the script write_metropolis_edges.py:

python python/write_metropolis_edges.py

The config.toml file needs to contain the following key-value pairs:

• clean_edges_file: Path to the file where the edges will be stored after postprocessing. This path can be the same as raw_edges_file if you just want to replace the imported file. Accepted formats: Parquet, GeoJSON, FlatGeobuf, GeoPackage, Shapefile.
• metropolis.input_directory: Path to the directory where the edges input file will be stored.
• metropolis.format: Format to be used for the input file (either "CSV" or "Parquet").