This repository contains works that I did for Spatial Statistics class at MIT, Fall 2023. Everything here owes very much to Eric Huntley's teaching materials. Please check the credit and sourse for each file.
Spatial statistics is a branch of statistics that focuses on the analysis of data that have a spatial component, allowing for the understanding and interpretation of spatial patterns, relationships, and dynamics.
It encompasses techniques for analyzing spatial autocorrelation, where observations nearby in space may be correlated with each other, and spatial heterogeneity, where relationships between variables may vary across space.
Through methods like spatial autocorrelation (Moran's I, hotspot analysis) and spatial regression, it provides powerful tools for urban planners and geographers to explore and model the complexity of spatial data, revealing insights into the spatial structure of phenomena.
-
Global Spatial Autocorrelation
- Understanding and measuring spatial autocorrelation on a global scale.
-
Local Spatial Autocorrelation
- Identifying local patterns of spatial association.
-
Spatial Regression (Error and Lag Models)
- Modeling spatial relationships accounting for spatial autocorrelation and spatial heterogeneity.
-
Spatial Regression (Geographically Weighted Regression)
- Adapting regression coefficients to vary across space to reflect local conditions.