GWmodel3 is the next generation of the R package for geographically weighted models. It is built on top of a kernel written by C++ and interfaces for R.
Build and installation
This package currently support Windows, Linux and macOS. Before installing this package, please make sure the library Armadillo and GSL are installed. And also install the R package sf first. The way of installing them vary depends on the OS.
Windows
To install Armadillo, please install R package RcppArmadillo together with sf package.
install.packages(c("RcppArmadillo", "sf"))
For different versions of RTools, there are different ways to install GSL. Please check that whether directory include/gsl/
and files lib/libgsl.a
and lib/libgslcblas.a
exist in one of the following paths under the installation directory of RTools:
usr
-
mingw32
ormingw64
-
clang32
orclang64
x86_64-w64-mingw32.static.posix
If they exist, this package can be properly compiled. If they don’t, please use the library files provided by Professor Ripley. Download and uncompress it to the installation directory of R, which is R_HOME
. Make sure the include
directory is merged with R_HOME/include
, and put the whole lib
directory under R_HOME
.
Then use the following command to install this source package.
Linux and macOS
. Firstly, install R package RcppArmadillo together with sf package.
install.packages(c("RcppArmadillo", "sf"))
Then install GSL via package manager. For example, on Ubuntu, use the following command.
Then use the following command to install this source package.
Usage
GWmodel3 provides a set of R interface for the kernel library libgwmodel. Thus, the usage is different from the original package GWmodel. To avoid conflicts, use the following code to import this package.
For the usage of each function, please turn to their tutorials or documents.
Report Problems
Please go to the “Issues” page to submit any issues.