Note: this is a package forked from the original KRLS2 package, edited with minor fixes by Xinkun Nie and further modified by Pawel Morzywolek to remove the pre-scaling of the data in the function krls2.R. The version number has been updated to 1.1.2. This package provides methods for fitting flexible functional forms for continuous and binary outcomes.
You can install the latest version by running:
devtools::install_github('pmorzywolek/KRLS')This version uses Rcpp extensively for speed reasons. These means you need to have the right compilers on your machine.
If you are on Windows, you will need to install RTools if you haven't already. If you still are having difficulty with installing and it says that the compilation failed, try installing it without support for multiple architectures:
devtools::install_github('pmorzywolek/KRLS', args=c('--no-multiarch'))In order to compile the C++ in this package, RcppArmadillo will require you to have compilers installed on your machine. You may already have these, but you can install them by running:
xcode-select --installIf you are having problems with this install on Mac OSX, specifically if you are getting errors with either lgfortran or lquadmath, then try open your Terminal and try the following:
curl -O http://r.research.att.com/libs/gfortran-4.8.2-darwin13.tar.bz2
sudo tar fvxz gfortran-4.8.2-darwin13.tar.bz2 -C /Also see section 2.16 here