TY - BOOK
T1 - Reducing the running time for gwr.multiscale in GWmodel
AU - Höglund, Hjalmar
AU - von Brömssen, Claudia
PY - 2025
Y1 - 2025
N2 - GWmodel is an R package for geographically weighted regression. It allows researchers to consider environmental data collected at different geographic locations and study connections between them. As the number of observations grow, the time to fit these models increases to the point where it is no longer practical for the researcher to run it on their own laptop, instead having to offload the computation to a supercomputer. By reducing memory copies, vectorising the calculations and using memoization the running time is reduced from days to minutes, increasing the size of data sets that can feasibly be handled on a researcher’s laptop.
AB - GWmodel is an R package for geographically weighted regression. It allows researchers to consider environmental data collected at different geographic locations and study connections between them. As the number of observations grow, the time to fit these models increases to the point where it is no longer practical for the researcher to run it on their own laptop, instead having to offload the computation to a supercomputer. By reducing memory copies, vectorising the calculations and using memoization the running time is reduced from days to minutes, increasing the size of data sets that can feasibly be handled on a researcher’s laptop.
UR - https://res.slu.se/id/publ/143310
M3 - Report
T3 - Rapport (Institutionen för energi och teknik, SLU)
BT - Reducing the running time for gwr.multiscale in GWmodel
PB - Department of Energy and Technology, Swedish University of Agricultural Sciences
ER -