INSTALL

This is a very concise listing, for developers, detailing how to install and work with the source files in this github repository. If you are a regular user, you should see the README.md file, best read on github: https://github.com/DormanLab/kmodes.git

# to compile standalone executable only
git clone https://github.com/DormanLab/kmodes.git
cd kmodes/src
cmake .
make
sudo cp run_kmodes /usr/local/bin
cd ../..	# for further development
make clean	# you should clean up the compile detritus

# to install R package
## the easiest way is from R using devtools
devtools::install_github("DormanLab/kmodes")
## but you can also install from source
git clone https://github.com/DormanLab/kmodes.git
R CMD build kmodes
sudo R CMD INSTALL kmodes*.tar.gz

# to work with both packages at once, it is best to separate the compiles and clean often
git clone https://github.com/DormanLab/kmodes.git
cd kmodes/src
cmake -S . -B build
cd build
make
sudo cp run_kmodes /usr/local/bin
cd ../../
make clean	# clean up any left-over compile detritus in case you did not use a build directory
Rscript --vanilla -e "devtools::check()"	# check package
make clean	# critical to clean up check detritus
sudo Rscript --vanilla -e "devtools::install()"	# install from source
# or, another way:
cd ..
R CMD build kmodes
sudo R CMD INSTALL kmodes*.tar.gz
# or another way: but this appears to NOT respect the build process, compiling for example, cmdline.c
cd kmodes
sudo R CMD INSTALL .
# if all the builds and R scripts, you can hand-compile and test the R shared library
R CMD SHLIB -o kmodes.so cluster.c entropy.c error.c hash.c kmodes.c kmodes_init.c kmodes_r.c matrix_exponential.c order.c r8slib.c run_kmodes.c sample.c timing_mach.c
# then test
scripts/test_kmodes.R