High-performance solutions of geographically weighted regression in R
ABSTRACTAs an established spatial analytical tool, Geographically Weighted Regression (GWR) has been applied across a variety of disciplines. However, its usage can be challenging for large datasets, which are increasingly prevalent in today’s digital world. In this study, we propose two high-perfor...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2022-10-01
|
| Series: | Geo-spatial Information Science |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/10095020.2022.2064244 |
| _version_ | 1811205507590914048 |
|---|---|
| author | Binbin Lu Yigong Hu Daisuke Murakami Chris Brunsdon Alexis Comber Martin Charlton Paul Harris |
| author_facet | Binbin Lu Yigong Hu Daisuke Murakami Chris Brunsdon Alexis Comber Martin Charlton Paul Harris |
| author_sort | Binbin Lu |
| collection | DOAJ |
| description | ABSTRACTAs an established spatial analytical tool, Geographically Weighted Regression (GWR) has been applied across a variety of disciplines. However, its usage can be challenging for large datasets, which are increasingly prevalent in today’s digital world. In this study, we propose two high-performance R solutions for GWR via Multi-core Parallel (MP) and Compute Unified Device Architecture (CUDA) techniques, respectively GWR-MP and GWR-CUDA. We compared GWR-MP and GWR-CUDA with three existing solutions available in Geographically Weighted Models (GWmodel), Multi-scale GWR (MGWR) and Fast GWR (FastGWR). Results showed that all five solutions perform differently across varying sample sizes, with no single solution a clear winner in terms of computational efficiency. Specifically, solutions given in GWmodel and MGWR provided acceptable computational costs for GWR studies with a relatively small sample size. For a large sample size, GWR-MP and FastGWR provided coherent solutions on a Personal Computer (PC) with a common multi-core configuration, GWR-MP provided more efficient computing capacity for each core or thread than FastGWR. For cases when the sample size was very large, and for these cases only, GWR-CUDA provided the most efficient solution, but should note its I/O cost with small samples. In summary, GWR-MP and GWR-CUDA provided complementary high-performance R solutions to existing ones, where for certain data-rich GWR studies, they should be preferred. |
| first_indexed | 2024-04-12T03:32:48Z |
| format | Article |
| id | doaj.art-ef6fab6a9cd448b7a8c0440be9dce91a |
| institution | Directory Open Access Journal |
| issn | 1009-5020 1993-5153 |
| language | English |
| last_indexed | 2024-04-12T03:32:48Z |
| publishDate | 2022-10-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Geo-spatial Information Science |
| spelling | doaj.art-ef6fab6a9cd448b7a8c0440be9dce91a2022-12-22T03:49:30ZengTaylor & Francis GroupGeo-spatial Information Science1009-50201993-51532022-10-0125453654910.1080/10095020.2022.2064244High-performance solutions of geographically weighted regression in RBinbin Lu0Yigong Hu1Daisuke Murakami2Chris Brunsdon3Alexis Comber4Martin Charlton5Paul Harris6School of Remote Sensing and Information Engineering, Wuhan University, Wuhan, ChinaSchool of Remote Sensing and Information Engineering, Wuhan University, Wuhan, ChinaDepartment of Data Science, Institute of Mathematical Statistics, Tokyo, JapanNational Centre for Geocomputation, Maynooth University, Maynooth, IrelandSchool of Geography, University of Leeds, Leeds, UKNational Centre for Geocomputation, Maynooth University, Maynooth, IrelandSustainable Agriculture Sciences North Wyke, Rothamsted Research, Okehampton, UKABSTRACTAs an established spatial analytical tool, Geographically Weighted Regression (GWR) has been applied across a variety of disciplines. However, its usage can be challenging for large datasets, which are increasingly prevalent in today’s digital world. In this study, we propose two high-performance R solutions for GWR via Multi-core Parallel (MP) and Compute Unified Device Architecture (CUDA) techniques, respectively GWR-MP and GWR-CUDA. We compared GWR-MP and GWR-CUDA with three existing solutions available in Geographically Weighted Models (GWmodel), Multi-scale GWR (MGWR) and Fast GWR (FastGWR). Results showed that all five solutions perform differently across varying sample sizes, with no single solution a clear winner in terms of computational efficiency. Specifically, solutions given in GWmodel and MGWR provided acceptable computational costs for GWR studies with a relatively small sample size. For a large sample size, GWR-MP and FastGWR provided coherent solutions on a Personal Computer (PC) with a common multi-core configuration, GWR-MP provided more efficient computing capacity for each core or thread than FastGWR. For cases when the sample size was very large, and for these cases only, GWR-CUDA provided the most efficient solution, but should note its I/O cost with small samples. In summary, GWR-MP and GWR-CUDA provided complementary high-performance R solutions to existing ones, where for certain data-rich GWR studies, they should be preferred.https://www.tandfonline.com/doi/10.1080/10095020.2022.2064244Non-stationaritybig dataparallel computingCompute Unified Device Architecture (CUDA)Geographically Weighted models (GWmodel) |
| spellingShingle | Binbin Lu Yigong Hu Daisuke Murakami Chris Brunsdon Alexis Comber Martin Charlton Paul Harris High-performance solutions of geographically weighted regression in R Geo-spatial Information Science Non-stationarity big data parallel computing Compute Unified Device Architecture (CUDA) Geographically Weighted models (GWmodel) |
| title | High-performance solutions of geographically weighted regression in R |
| title_full | High-performance solutions of geographically weighted regression in R |
| title_fullStr | High-performance solutions of geographically weighted regression in R |
| title_full_unstemmed | High-performance solutions of geographically weighted regression in R |
| title_short | High-performance solutions of geographically weighted regression in R |
| title_sort | high performance solutions of geographically weighted regression in r |
| topic | Non-stationarity big data parallel computing Compute Unified Device Architecture (CUDA) Geographically Weighted models (GWmodel) |
| url | https://www.tandfonline.com/doi/10.1080/10095020.2022.2064244 |
| work_keys_str_mv | AT binbinlu highperformancesolutionsofgeographicallyweightedregressioninr AT yigonghu highperformancesolutionsofgeographicallyweightedregressioninr AT daisukemurakami highperformancesolutionsofgeographicallyweightedregressioninr AT chrisbrunsdon highperformancesolutionsofgeographicallyweightedregressioninr AT alexiscomber highperformancesolutionsofgeographicallyweightedregressioninr AT martincharlton highperformancesolutionsofgeographicallyweightedregressioninr AT paulharris highperformancesolutionsofgeographicallyweightedregressioninr |