nssac / simba_schelling

SIMBA_Schelling

Summary

This git repository contains code for demonstrating the SIMBA framework in the context of the Schelling model for segregation. It is applied to Richmond, VA for which there is accompanying data.

Preliminaries

The following preparatory steps are required:

• Download the git repository: https://github.com/NSSAC/SIMBA_Schelling into a base diretory .
• From the folder https://net.science/files/0c7ae1b6-13d3-4a97-8851-95e3b0e909f8/ download:
  • The person/household synthetic_richmond.csv
  • The boundary data file geo_reference.csv

Requirements

The code minimally requires Python 3.8, Pandas 1.3.5, Geopandas 0.10.2, and Requests 2.28.1. Currently, it has only been tested under Linux using the slurm job submission system.

Notes

13 Feb 2023: we are exploring ways to execute the model without the need for using slurm. Updates will be posted when ready.

Running the Simba-Schelling model

• Define module paths and execution order in simba_schelling/data/config.json
• Edit the configuration script to reference Schelling environment path and execution variables as defined in simba_schelling/data/config.json.
• Edit configuration script for slurm execution parameters - account details, run times, etc.
• Run SIMBA/Schelling through run.sh launcher script; SIMBA assumes the default path to the configuration script to be under data/config.json. In the event other configuration files are created, they can be passed as an argument during execution.

To verify the successful run of the SIMBA Schelling integration, a visualization of the segregated houshold locations can be generated through schelling/visualize_run.py.

Licenses

The code in the SIMBA_Schelling repository uses the Apache 2.0 license, see https://www.apache.org/licenses/LICENSE-2.0.html. The two data files (a and b) are made available under the CC-BY-4.0 license, see https://creativecommons.org/licenses/by/4.0/.

Citation

Stefan Hoops, Ian Le, Dustin Machi, Henning S. Mortveit, Sami Saliba, Samarth Swarup (2023). SIMBA: A Framework for Rapid and Extensible Agent-based Simulation Development. Submitted.

Further Reading

• Thomas C. Schelling: Dynamic Models of Segregation, Journal of Mathematical Sociology 1(2), 143-186 (1971).
• Anderson, T., Leung, A., Dragicevic, S., Perez, L.: Modeling the geospatial dynamics of residential segregation in three Canadian cities: An agent-based approach. Transactions in GIS 25(2), 948–967 (2021)
• The geographic boundary data geo_reference.csv was derived from TIGER/Line data, see https://catalog.data.gov/dataset/tiger-line-shapefile-2017-2010-state-virginia-2010-census-block-state-based