Skip to content

matuteiglesias/spatial-coexistence-metrics

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

📘 The Overlooked Insights from Correlation Structures in Economic Geography

A computational exploration and open codebase supporting the findings of Matias Nehuen Iglesias, Utrecht University.

Keywords: spatial economics, co-location, cosine similarity, MAUP, economic geography, industry clustering, employment density, continuous space modeling, spatial analysis, kernel density, economic complexity, GIS.


🧠 What This Project Is

This repository contains the source code, computational experiments, and data structures supporting the research paper:

📄 “The Overlooked Insights from Correlation Structures in Economic Geography”
Papers in Evolutionary Economic Geography, Working Paper #21.05
Download PDF

This project studies the mathematical and empirical conditions under which continuous spatial models (e.g., radial influence around establishments) and discrete areal data models (e.g., county-level employment statistics) yield comparable measures of industrial co-location.


🔍 Motivation

Traditional spatial analyses rely on administrative units that may distort the real patterns of economic co-location due to the Modifiable Areal Unit Problem (MAUP). This repository:

  • Formalizes coexistence measures in continuous space.
  • Compares them with cosine similarity metrics computed on discrete areal data.
  • Offers code to replicate analytical derivations and computational simulations.
  • Helps identify when discrete spatial indicators are reliable proxies for true spatial proximity.

🧪 Contents


/notebooks/           # Exploratory and replication notebooks
/models/              # Density function models and simulation engines
/data/                # Sample geospatial datasets (synthetic and real)
/experiments/         # Setup for controlled spatial displacement simulations
/utils/               # Math helpers, kernel generators, area integrators


🧰 Methods Implemented

  • Kernel density estimation (Gaussian & Laplace)
  • Cosine similarity over employment vectors
  • Numerical integration of joint density functions
  • Systematic decomposition of co-location pairs (A/B/C/D types)
  • Simulation of influence decay and spatial shifting
  • Comparative experiments across area sizes (e.g., 10km², 100km²)

💬 Citation

If you use this codebase or ideas in your work, please cite:


@techreport{iglesias2021correlation,
title={The Overlooked Insights from Correlation Structures in Economic Geography},
author={Matias Nehuen Iglesias},
year={2021},
institution={Utrecht University},
series={Papers in Evolutionary Economic Geography},
number={21.05}
}


🌐 Discover More


👨‍🔬 Author

Matias Nehuen Iglesias
PhD in Economics – SSSUP Univ di Pisa
Researcher in economic geography, spatial modeling, and applied AI.