Closed jlopezpena closed 5 years ago
Isn't this version of R mostly optimized for intel chips?
I am using a computer with an intel processor, that has nothing to do with my issue.
Woops sorry i though you were trying to compile for the raspberry pi!
I would try using setMKLthreads(1) in your R code to keep the MKL from trying to use multiple threads; see if this helps.
Cheers, Rich Calaway Microsoft R Product Team
I am trying to use MRO from python using
rpy2
, and encountering some issues.First,
rpy2
does not find the R libraries out of the box; this can be solved by setting up the environment variablesLD_LIBRARY_PATH
andR_HOME
before installingrpy2
after doing this,
rpy2
installs and run mostly fine (if it was installed before setting the environment libraries, it must be removed and installed again!). Perhaps this should be added to the documentation somewhere?Secondly, after installing
rpy2
I get constant warnings coming from the R side, concerning theRevoUtilsMath
pacakge and somegetThreads
method:I don't know if it is related to this warning, but when running long R processes from python they randomly die (from jupyter I just get a
Kernel has died
message, without any error I can trace back). I am not sure what is causing this, as it seem to happen randomly. I am guessing some race condition, or the multithreaded libraries not playing nice with rpy2.Any ideas on how I could debug/fix this issue?