Hi,
I am working with the package nlme, and I tried creating a new correlation
class (which, according to the help pages, is possible if you write a few
new method functions). Anyways, I think I am 99% of the way there, but I
have a recurring problem with R crashing on seemingly innocuous statements
(I have set debug() on nearly every function, so I can see where it is
failing).
For instance, sometime the code fails on the simple statement:
object[[i]]
Two things are odd about this:
1. It's not the first time I call object[[i]]
2. If I run save() on the object just before the crash, open the object in a
new workspace and run object[[i]], then no crash or error occurs.
"[[" is not the only function that has been giving me grief. Certain C
routines that get called multiple time have displayed very similar behavior
(ie. they work fine several times before they crash, and if the object is
exported just before the crash and the routine is called in a new workspace
there is no error).
The behavior kind of reminds me of the craziness that occur with an
uninitialized pointer (thinking back to coding in C in college). The other
thing that points me in that direction is that Windows returns an error code
ending ...00005, which I believe indicates a segmentation fault. I was also
wondering whether there might be some rogue global variable, but I have no
idea what.
I have struggling with this for a week. Does anyone have any insights on
what might be wrong or what I might try?
Thanks so much,
Mike
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
