Argentina Transport Risk World Bank project

Requirements

Python and libraries

Python version 3.6 is required to run the scripts in this project. We suggest using miniconda to set up an environment and manage library dependencies.

Create a conda environment from the environment.yml definition:

conda env create -f environment.yml
conda install python-igraph

See http://igraph.org/python/ for instructions on Windows installation of python-igraph.

Activate the environment:

conda activate argentina-transport

Set up the atra package (this project) for development use:

python setup.py develop

GAMS

The macroeconomic loss model uses GAMS (General Algebraic Modeling System) via its python API. GAMS provide installation and licensing instructions.

Configuration

The location of data and output files are configured by a config.json file. To point the scripts to the shared folder locations:

• copy config.template.json to config.json
• edit config.json to provide the paths to your working copy of your system directories (scripts may assume that file locations within the shared folders are consistent)
  • incoming_data: /incoming_data
  • data: /data
  • figures: /figures
  • output: /results

Note that on Windows, you will need to use double backslashes (\\) in the file paths, for example:

"data": "C:\\Users\\Username\\projects\\atra\\data"

Acknowledgements

This project has been developed by Oxford Infrastructure Analytics as part of a project funded by the World Bank and the Global Facility for Disaster Reduction and Recovery (GFDRR).

All code is copyright Oxford Infrastructure Analytics, licensed MIT (see the LICENSE file for details) and is available on GitHub at oi-analytics/argentina-transport.