Function Fittings by Regressions and Application in Analyzing Urban Density Patterns

• Published in: Computational Methods and GIS Applications in Social Science - Lab Manual pp. 130 - 145
• Main Authors: Liu, Lingbo, Wang, Fahui
• Format: Book Chapter
• Language: English
• Published: United Kingdom CRC Press 2024; Taylor & Francis Group
• Edition: 1
• Subjects: Weighted Liner Regressions; Column Selection; Census Tract Centroids; Euclidean Distance; Python Script; Related Statistical Issues; Kernel Density Estimation; Origin Id; Top Input; Table Row; Destination Id; Monocentric Models; Tract Centroids; CMSA; Bottom Input; Health Informatics; 2SFCA Method; CSV File; Regular Linear Regression; Census Tract; Left Table; Population Density Patterns; Group Column; Pivot Column; Output Port
• ISBN: 9781032302430; 1032302437
• DOI: 10.1201/9781003304357-6

This chapter discusses how to identify the best fitting function to capture urban and regional population density patterns. Such an approach emphasizes the influence of a center or multiple centers on areawide density patterns in a city or across a region. By examining the change of density function over time, one can detect the growth pattern for urban and regional structures. The methodological focus is on function fittings by regressions and related statistical issues. This chapter discusses how to identify the best fitting function to capture urban and regional population density patterns. The methodological focus is on function fittings by regressions and related statistical issues. Chicago has been an important study site for urban studies attributable to the legacy of so-called "Chicago school". The project analyzes the density patterns at both the census tract and survey township levels to examine the possible modifiable areal unit problem. The chapter implements spatial analysis tasks such as distance computation, areal interpolation, and various regressions such as Ordinary Least Squares, nonlinear and weighted regressions. The output table contains all columns of the two inputs, with many empty values in most columns except for geometry.