Hello,
I encountered this problem while working on my Rcpp package. I don't
know if it can be related to Rcpp, but I post it here looking for some
advices. I'll try to make myself as clear as possible:
from an R function in my Rcpp package I call a C++ function where huge
datasets (~ 2GB) are loaded in memory, parsed and used for computation.
Clearly, every object and every data structure is handled on the C++
side. The computation advances successfully and produces a graph that is
returned to R in terms of an edges list.
On the R side I can use this list in many ways, but first it gets
converted into a data.table, whose size is about 2MB. What is surprising
is that the 'rsession' process keeps holding 3.2 GB of physical memory,
as it can be easily checked with the 'top' command from shell.
When I decide to run again the R function above, the rsession crashes
soon afterwards, and this is likely due to a memory overflow (despite my
PC has 8GB RAM).
it seems like the C++ function, once it terminates its execution, does
not free the memory used. instead, rsession keeps holding these objects
even though C++ has terminated. is it possible?
Thanks in advance,
Fabio
_______________________________________________
Rcpp-devel mailing list
Rcpp-devel@lists.r-forge.r-project.org
https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcpp-devel
