Bug 16568 - rpart (x ~ x) segfaults
Summary: rpart (x ~ x) segfaults
Status: CLOSED CONTRIBUTED PACKAGE
Alias: None
Product: R
Classification: Unclassified
Component: Add-ons (show other bugs)
Version: R 3.2.1
Hardware: x86_64/x64/amd64 (64-bit) Windows 64-bit
: P5 normal
Assignee: R-core
URL:
Depends on:
Blocks:
 
Reported: 2015-10-15 22:50 UTC by Sam Buttrey
Modified: 2018-01-11 19:09 UTC (History)
3 users (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Sam Buttrey 2015-10-15 22:50:52 UTC
Greetings. The following code produces a segmentation fault in my installation of R 3.2.1 under Windows 7 (64-bit). Since rpart is "recommended" I figured I should report this here.

library (rpart)
set.seed (123)
x <- rnorm (100)
rpart (x ~ x)
Comment 1 Benjamin Tyner 2018-01-11 12:19:32 UTC
Perhaps these are a clue?

Warning messages:
1: In model.matrix.default(attr(frame, "terms"), frame) :
  the response appeared on the right-hand side and was dropped
2: In model.matrix.default(attr(frame, "terms"), frame) :
  problem with term 1 in model.matrix: no columns are assigned
Comment 2 Luke Tierney 2018-01-11 16:36:05 UTC
The segfault occurs in a C routine in the contributes rpart package.
Please report to the package maintiner.
Comment 3 Benjamin Tyner 2018-01-11 17:17:42 UTC
Isn't this a "Core-supported package" ?
Comment 4 Luke Tierney 2018-01-11 17:54:42 UTC
It is not. Look at the result of

packageDescription("rpart")

or the CRAN web page for rpart.
Comment 5 Peter Dalgaard 2018-01-11 18:01:13 UTC
Specifically, it ia a _recommended_ package, which is not quite the same thing as a core-maintained package. In particular, it is the maintainer who builds the package and submits to CRAN. Of course, anyone is welcome to help with debugging.
Comment 6 Benjamin Tyner 2018-01-11 19:09:59 UTC
Thanks. So then I take it that for bugzilla reports, the maintainer really does need to be "R Core Team" rather than an individual member of said team? In that case, should other such reports be closed as well, for example:

https://bugs.r-project.org/bugzilla/show_bug.cgi?id=16900