Mobility constraints in segregation models

Citing
Packages
Abstract
Data Availability
Structure of the repository
Analysis

Citing

In this repository you can find the code for running our model and to replicate the analysis conducted in our paper. If you use the code in this repository, please cite our paper:

Daniele Gambetta, Giovanni Mauro, and Luca Pappalardo. "Mobility constraints in segregation models." arXiv preprint arXiv:2305.10170 (2023).*

@article{gambetta2023mobility,
  title={Mobility constraints in segregation models},
  author={Gambetta, Daniele and Mauro, Giovanni and Pappalardo, Luca},
  journal={arXiv preprint arXiv:2305.10170},
  year={2023}
}

Packages

For running notebooks and scripts of this project you must install the following Python packages:

  mesa
  pandas
  matplotlib
  numpy
  altair
  scikit-mobility

Abstract

Since the development of the original Schelling model of urban segregation, several enhancements have been proposed, but none have considered the impact of mobility constraints on model dynamics. Recent studies have shown that human mobility follows specific patterns, such as a preference for short distances and dense locations. This paper proposes a segregation model incorporating mobility constraints to make agents select their location based on distance and location relevance. Our findings indicate that the mobility-constrained model produces lower segregation levels but takes longer to converge than the original Schelling model. We identified a few persistently unhappy agents from the minority group who cause this prolonged convergence time and lower segregation level as they move around the grid centre. Our study presents a more realistic representation of how agents move in urban areas and provides a novel and insightful approach to analyzing the impact of mobility constraints on segregation models. We highlight the significance of incorporating mobility constraints when policymakers design interventions to address urban segregation.

Data Availability

The data generatd and analyzed in our work are the ones gnerated by our simulations that you can replicate by running the scripts and notebooks of this repository.

Structure of the repository

In the main level of the repo you can find four folders:

notebooks: contains Jupyter Notebooks to generate analysis and visualizations of the paper
results_data: contains the data generated by our models for the analysis of the main paper and for the analysis of the supplementary notes
models: contains python scripts for MESA implementation of the Model Agent classes of our simulation and pyhon scripts to generate results of simulations
figures: contains figures of main paper, generated with notebooks of notebooks and data in results_data

In particular, inside models folder, you can find:

schellingmob.py: python script containing the MESA implementation of the Model and Agent classes of our simulation
generator_gravity_puagents.py: python code for replicating the analysis of the mobility of the persistently unhappy agents in gravity model
run.py: python code to run simulations of the paper using MESA Batchrunner class
measures_individual_schelling.py: python code used to calculate jump length statistics

Inside results_data there are two subfolders:

results_data_main contains the data generated by our models for the analysis of the main paper.
results_data_supplementary contains the data generated by our models

Analysis

To run our model for replicate the analysis use three jupyter notebooks of this repository. In particular:

Figure 1 is generated by fig1.ipynb without any supplementary data
Figure 2 is generated by fig2.ipynb and data contained in subfolders of data_main
Figure 3 is generated by fig3.ipynb and data contained in data_main (agensts_pu.csv and suburbia_analysis.csv)
Supplementary images are generated by supplementary.ipynb and data contained in results_data/results_data_supplementary

Regarding data generation process:

Data in subfolders of results_data/results_data_main and data of results_data/results_data_supplementary are generated by run.py code, using different configuration parameters for each model
agensts_pu.csv and suburbia_analysis.csv are generated by generator_gravity_puagents.py

figure_1 figure_2 figure_3

dgambit/mobility_schelling