Skip to content

Latest commit

 

History

History
127 lines (95 loc) · 5.5 KB

toulouse.md

File metadata and controls

127 lines (95 loc) · 5.5 KB
Raw

Toulouse / Occitanie

The pipeline makes it easy to create synthetic populations and simulations for other regions than Île-de-France. In any case, we recommend to first follow instructions to set up a synthetic population for Île-de-France and (if desired) the respective simulation. The following describes the steps and additional data sets necessary to create a population and simulation for Toulouse.

Additional data

A) Regional census data

Toulouse is not included in the census data set that is uesd for Île-de-France (Zone A). Instead, Zone D needs to be obtained from the same source. Download the dbase version of Zone D and put the respective file (FD_INDCVIZD_2015.dbf) into the data/rp_2015 folder.

B) Address database (BD-TOPO)

You need to download the region-specific address database. Go to IGN Open data database, scroll down until you see BD TOPO® Décembre 2020 Tous Thèmes par région édition Décembre 2020 format shapefile. Click on the download link under Région Occitanie - R 76. Open the downloaded archive and copy the files ADRESSE.* from the folder ADDRESSES in shape file format into data/bdtopo (overriding the data for Île-de-France if you had set up that scenario before).

C) OpenStreetMap data

Only if you plan to run a simulation (and not just generate a synthetic population), you need to obtain additional data from OpenStreetMap. Geofabrik does not provide a continuous cut-out for Occitanie. Instead, the former regions of Midi-Pyrennées and Languedoc-Roussilon are available. Download both regions in .osm.pbf format and put the files into the folder data/osm.

D) GTFS data

Again, only if you want to run simulations, the digital transit schedule is required. Unfortunately, there is no consolidated GTFS schedule avaiable for the Occitanie region. Hence, it is necessary to collect all relevant GTFS schedules one by one. Here, we provide a selection of links, which is not necessarily exhaustive:

Download all the zip'd GTFS schedules and put them into the folder data/gtfs.

Overview

Afterwards, you should have the following additional files in your directory structure:

  • data/rp_2015/FD_INDCVIZD_2015.dbf

Only for simulation:

  • osm/midi-pyrenees-latest.osm.pbf
  • osm/languedoc-roussillon-latest.osm.pbf
  • gtfs/tisseo.zip
  • gtfs/TAM_MMM_GTFS.zip
  • gtfs/RESEAU_LR_GTFS_20200706.zip
  • gtfs/export_gtfs_voyages.zip
  • gtfs/export-intercites-gtfs-last.zip
  • gtfs/export-ter-gtfs-last.zip

Note that the file names may change slightly over time as GTFS schedule are updated continuously.

Generating the population

To generate the synthetic population, the config.yml needs to be updated. While the relevant code points to the Île-de-France data sets by default, you can adjust the paths inidividually. To let the pipeline use the Zone D census data set, add the following to config.yml in the config section:

config:
  # ...
  census_path: rp_2015/FD_INDCVIZD_2015.dbf
  # ...

Furthermore, by default the pipeline will filter all other data sets for the Île-de-France region. To make it use the Occitanie region, adjust the configuration as follows:

config:
  # ...
  regions: []
  departments: ["09", 82, 81, 11, 31, 32] # 12 30 34 46 48 65 66
  # ...

This will make the pipeline filter all data sets for the departments noted in the list above, which is a set of the closest departments around Toulouse. If you want to generate the whole Occitanie region, add the commented out department identifiers to the list.

Finally, to not confuse output names, we can define a new prefix for the output files:

config:
  # ...
  output_prefix: toulouse_
  # ...

You can now enter your Anaconda environment and call the pipeline with the synthesis.output stage activated. This will generate a synthetic population for Toulouse and surroundings.

Running the simulation

To prepare the pipeline for a simulation of Toulouse, the paths to the OSM data sets and to the GTFS schedule must be adjusted explicitly:

config:
  # ...
  gtfs_path: gtfs/tisseo.zip;gtfs/TAM_MMM_GTFS.zip;gtfs/export_gtfs_voyages.zip;gtfs/export-intercites-gtfs-last.zip;gtfs/export-ter-gtfs-last.zip;gtfs/RESEAU_LR_GTFS_20200706.zip
  osm_path: osm/midi-pyrenees-latest.osm.pbf;osm/languedoc-roussillon-latest.osm.pbf
  # ...

Note that the pipeline will automatically cut GTFS and OpenStreetMap data to the relevant area (defined by the filter above) if you run the simulation.

To test the simulation and generate the relevant MATSim files, run the pipeline with the matsim.output stage enabled.