Spatial data analysis with R: wrangling, visualization and econometric models

Spatial Analysis with R: wrangling, visualization and econometric models (Análisis Espacial con R: manejo, visualización y modelos econométricos)is a full open course in Spanish for the analysis of spatial information in R software. The course is intended for Spanish-speaking economics undergraduate students interested in acquiring technical skills for quantitative analysis required by the regional sciences and the spatial approach to economics, but it could be useful for other students of social sciences attracted by the topic.

The objective of the book is to guide the student, from a basic practical approach, in the knowledge and handling of spatial information exploration, analysis and modeling techniques through the use of R and RStudio, which are, respectively, a program computer science and programming language focused on statistical analysis and information visualization, on the one hand, and an integrated development environment (IDE), on the other.

The course is presented in the form of an electronic book and it is structured in five chapters of gradual learning. In Chapter 1, the basics of R and RStudio are introduced using the exploratory data analysis approach not only with the basic R package, but also with the popular set of tools provided by tidyverse, that is, the student is completely introduced to the use of the software to propose and solve questions related to the information structure. Chapter 2 shows how to create various types of choropleth maps and the enormous styling customization flexibility for this purpose via the tmap package. Chapter 3 presents how to carry out an exploratory analysis of spatial data where the student will find how to define the interrelationships that take place in space through the construction of spatial weight matrices and she will learn about spatial autocorrelation and its implications in the information analysis, through various packages such as spdep  and rgeoda. Meanwhile, in Chapter 4, a very synthetic review of simple regression models is presented, emphasizing the autocorrelation problem that could occur when estimating a linear model with spatial data. Finally, in Chapter 5, two of the different spatial econometric modeling alternatives available in R it is shown, with spatialreg.

The logic of each chapter integrates three elements: i) explanation of the fundamental concepts covered, ii) the use of real information in the software that serves to illustrate the highlighted concepts, iii) exercises proposed for the student to delve into the topics exposed.

The examples and exercises presented in the course are based on a database on the situation of the COVID19 pandemic between March and September 2020 in the Metropolitan Zone of the Valley of Mexico (Zona Metropolitana del Valle de México), the largest metropolitan area in Mexico, made up of 76 local administrative units or municipalities with more than 21 million inhabitants. In addition, the database used provides information on the economic structure at the municipal level and a set of variables of sociodemographic characteristics from the population census.