Area-to-point parameter estimation with geographically weighted regression

Murakami, Daisuke; Tsutsumi, M.

doi:10.1007/s10109-015-0212-8

Cited by 12 publications

(7 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is also one aspect of the modifiable areal unit problem (Openshaw 1984), where, as the areal size changes (up‐ or down‐scaling), the correlation between a given pair of variables changes, conditional upon the scales of spatial variation that exist in the same variables. Gotway and Young (2002), Zhang, Atkinson, and Goodchild (2014) and Murakami and Tsutsumi (2015) describe the effects of varying support of the observation units.…”

Section: Semantic Issues In Detailmentioning

confidence: 99%

The Forgotten Semantics of Regression Modeling in Geography

Comber

Harris

Lü

et al. 2019

Geographical Analysis

View full text Add to dashboard Cite

This article is concerned with the semantics associated with the statistical analysis of spatial data. It takes the simplest case of the prediction of variable y as a function of covariate(s) x, in which predicted y is always an approximation of y and only ever a function of x, thus, inheriting many of the spatial characteristics of x, and illustrates several core issues using “synthetic” remote sensing and “real” soils case studies. The outputs of regression models and, therefore, the meaning of predicted y, are shown to vary due to (1) choices about data: the specification of x (which covariates to include), the support of x (measurement scales and granularity), the measurement of x and the error of x, and (2) choices about the model including its functional form and the method of model identification. Some of these issues are more widely recognized than others. Thus, the study provides definition to the multiple ways in which regression prediction and inference are affected by data and model choices. The article invites researchers to pause and consider the semantic meaning of predicted y, which is often nothing more than a scaled version of covariate(s) x, and argues that it is naïve to ignore this.

show abstract

Section: Semantic Issues In Detailmentioning

confidence: 99%

The Forgotten Semantics of Regression Modeling in Geography

Comber

Harris

Lü

et al. 2019

Geographical Analysis

View full text Add to dashboard Cite

show abstract

“…[1,5,10,11]. Some research [12][13][14] show the effects of varying spatial support: as support changes (up-scaling and down-scaling or aggregation and disaggregation) statistical and other relationship changes. MAUP is a core geographical consideration.…”

Section: Introductionmentioning

confidence: 99%

The Importance of Scale and the MAUP for Robust Ecosystem Service Evaluations and Landscape Decisions

Comber

Harris

2022

Land

View full text Add to dashboard Cite

Spatial data are used in many scientific domains including analyses of Ecosystem Services (ES) and Natural Capital (NC), with results used to inform planning and policy. However, the data spatial scale (or support) has a fundamental impact on analysis outputs and, thus, process understanding and inference. The Modifiable Areal Unit Problem (MAUP) describes the effects of scale on analyses of spatial data and outputs, but it has been ignored in much environmental research, including evaluations of land use with respect to ES and NC. This paper illustrates the MAUP through an ES optimisation problem. The results show that MAUP effects are unpredictable and nonlinear, with discontinuities specific to the spatial properties of the case study. Four key recommendations are as follows: (1) The MAUP should always be tested for in ES evaluations. This is commonly performed in socio-economic analyses. (2) Spatial aggregation scales should be matched to process granularity by identifying the aggregation scale at which processes are considered to be stable (stationary) with respect to variances, covariances, and other moments. (3) Aggregation scales should be evaluated along with the scale of decision making (e.g., agricultural field, farm holding, and catchment). (4) Researchers in ES and related disciplines should up-skill themselves in spatial analysis and core paradigms related to scale to overcome the scale blindness commonly found in much research.

show abstract

“…GWR itself has been refined to accommodate extensions found in standard regression, such outlier-resistant (Fotheringham, Brunsdon, and Charlton 2002;, Harris, Fotheringham, and Juggins 2010), heteroskedastic (Fotheringham, Brunsdon, and Charlton 2002;Páez, Uchida, andMiyamoto 2002a, 2002b), ridge (D. C. Wheeler 2007;Gollini et al et al 2015), LASSO (D. C. Wheeler 2009) and elastic net form (K. Li and Lam 2018;. Further extensions include time in the form of geographically and temporally weighted regression (GTWR) (Huang, Wu, and Barry 2010;Fotheringham, Crespo, and Yao 2015), area to point regression (Murakami and Tsutsumi 2015), multiple scales of analysis (Yang 2014;Fotheringham, Yang, and Kang 2017), spatially variable model specification ) and the use of different distance metrics (Lu et al 2016).…”

Section: Introductionmentioning

confidence: 99%

gwverse: a template for a new generic Geographically Weighted Rpackage

Comber¹,

Brunsdon²,

Callaghan³

et al. 2021

Preprint

View full text Add to dashboard Cite

Geographically weighted regression (GWR) is a popular approach for investigating the spatial variation in relationships between response and predictor variables, and critically for investigating and understanding process spatial heterogeneity. It has been refined to accommodate outliers, hetroskedasticity and local collinearity and extended to LASSO and elastic net forms. The geographically weighted (GW) framework is increasingly used to accommodate different types of models and analyses reflecting a wider desire to explore spatial variation in model parameters or components and to move away from global, "whole map" approaches. However the growth in the use of GWR and different GW models has only been partially supported by package development in both R and Python, the major coding environments for spatial analysis. The result is that refinements have been inconsistently included (if at all) within GWR and GW functions in any given package. As an example, the GWmodel R package, which despite including the greatest number of GWR and GW related tools, has been developed and extended on a piecemeal basis with no overarching schema. This paper outlines the structure of a new gwverse package, that will over time replace GWmodel, that takes advantage of recent developments in the composition of complex, integrated packages. It conceptualises gwverse as having a modular structure, that separates core GW functionality and applications such as GWR. It adopts a function factory approach, in which bespoke functions are created and returned to the user based on user-defined parameters. Function factory approaches and the functions they generate have the advantage of enclosing environments that are execution environments of the function. The paper introduces and demonstrates two modules (written as linked packages) that can be used to undertake GWR as an initial transect through the proposed gwverse schema, and to support user defined GW modules, before discussing a number of key considerations and next steps.

show abstract

Area-to-point parameter estimation with geographically weighted regression

Cited by 12 publications

References 41 publications

The Forgotten Semantics of Regression Modeling in Geography

The Forgotten Semantics of Regression Modeling in Geography

The Importance of Scale and the MAUP for Robust Ecosystem Service Evaluations and Landscape Decisions

gwverse: a template for a new generic Geographically Weighted Rpackage

Contact Info

Product

Resources

About