Yes, that is the problem. But it is not supposed to be happening if you can compile other things with g++. Like for example, can you do R CMD build package_dir where package_dir is the directory where your package is.
asael, I wasnāt suggesting to use the files from ctsem, the readme on github (first page, below the files) describes various steps to take. I expect you need rtools in your path, like so:
In case of compile errors like g++ not found , ensure the devtools package is installed:
install.packages('devtools')
and include the following in your .Rprofile
library(devtools)
Sys.setenv(PATH = paste("C:/Rtools/bin", Sys.getenv("PATH"), sep=";"))
Sys.setenv(PATH = paste("C:\\Rtools\\mingw_64\\bin", Sys.getenv("PATH"), sep=";"))
Sys.setenv(BINPREF = "C:/Rtools/mingw_$(WIN)/bin/")
@Charles_DriverThank you so much that actually works! Thanks, @bgoodri and @jonah for all the patience and help! Yeiiiiiiiiiii!!!
@Charles_Driver Thanks for helping out. Is this something that most people using Windows/rtools are going to encounter when using rstantools to develop a package (I am no Windows expert)? If so we should either include your solution somewhere in our docs or figure out a way to avoid it being necessary for the user to figure that out.
@jonah, itās definitely a problem when people install from github, and I donāt think itās at all specific to ctsem. Only a few ctsem users are making use of the models that require compilation (nonlinear sde) so far as Iām aware, so Iām still not certain whether it is an issue when installing from CRAN but I assume so. I only recently realised it was such a widespread problem, definitely need to find a better way to deal with itā¦
Ok. I just spent a bit of time fixing up an install script that is hopefully fairly robust, found here: https://github.com/cdriveraus/ctsem/raw/master/installctsem.R
I tried out various incarnations (removed rtools, rstan, makevars, no rtools in path) but couldnāt reproduce the g++ not found error, though had it a lot at a workshop I taught recently.