Explicitly initializing all c++ objects in attempt to avoid valgrind error
This still has valgrind errors on my local Docker container.
However, with track-origins=yes, the source of the uninitialized value is a stack allocation in the function dlatrs_ (in /usr/lib/libopenblasp-r0.2.19.so), not inside phyr code. I was also very careful to avoid non-initialized items in any matrices/vectors. Given that this error is occurring in the likelihood function, the output from cor_phylo would very likely be influenced by it; no problems like this show up in the tests. Thus, I believe the source of this error is in the /usr/lib/libopenblasp-r0.2.19.so library.
Hopefully this explanation is sufficient for the CRAN maintainers.
This still has valgrind errors on my local Docker container.
However, with
track-origins=yes
, the source of the uninitialized value is a stack allocation in the functiondlatrs_
(in/usr/lib/libopenblasp-r0.2.19.so
), not insidephyr
code. I was also very careful to avoid non-initialized items in any matrices/vectors. Given that this error is occurring in the likelihood function, the output fromcor_phylo
would very likely be influenced by it; no problems like this show up in the tests. Thus, I believe the source of this error is in the/usr/lib/libopenblasp-r0.2.19.so
library.Hopefully this explanation is sufficient for the CRAN maintainers.