Functional SAR models: With application to spatial econometrics

Simultaneous autoregressive (SAR) models have been extensively used for the analysis of spatial data in areas as diverse as demography, economy and geography. These are linear models with a scalar response, scalar explanatory variables and autoregressive errors. In this work we extend this modeling...

Full description

Saved in:
Bibliographic Details
Published inSpatial statistics Vol. 29; pp. 145 - 159
Main Authors Pineda-Ríos, Wilmer, Giraldo, Ramón, Porcu, Emilio
Format Journal Article
LanguageEnglish
Published Elsevier B.V 01.03.2019
Subjects
Autoregressive errors
Functional linear models
Spatial dependence
Spatial weight matrix
Spatial dependence
Spatial weight matrix
Functional linear models
Autoregressive errors
Online AccessGet full text

Cover

More Information
Summary:Simultaneous autoregressive (SAR) models have been extensively used for the analysis of spatial data in areas as diverse as demography, economy and geography. These are linear models with a scalar response, scalar explanatory variables and autoregressive errors. In this work we extend this modeling approach from scalar to functional covariates. Least squares and maximum likelihood are used as estimation methods of the parameters. A simulation study is considered for evaluating the performance of the proposed methodology. As an illustration, the model is used to establish the relationship between unsatisfied basic needs and curves of gross domestic product obtained in 32 departments of Colombia (districts of the country).
ISSN:2211-6753
2211-6753
DOI:10.1016/j.spasta.2018.12.002