Robert_Keefe@umit.maine.edu
2005-Jan-07 04:44 UTC
[Rd] R 2.1 dies when some text() arguments are NULL (PR#7477)
[My very deepest apologies in advance if this is not plain text; I am recently at a new job in a windows-based environment and have not yet set up my favorite old (pine) system. I have tried my best to strip all formatting from the message..] I'm overlaying plots of tree locations mapped before and after logging activities and checking my matched individuals using several arguments to text() which plot tree attributes by or near their (x,y) locations. R 2.1 dies when a coding error on my part results in a null data.frame... Example: plot(1) text(NULL ~ NULL, label=NULL) This happens for me under Windows 98 and XP. Andrew Robinson was kind enough to try it for me on FreeBSD with the same result. The code returns nothing for me under R 1.9 on Windows 98 and XP (same win. machines). The version details of the three are as follows: # FreeBSD:>version_ platform i386-portbld-freebsd5.3 arch i386 os freebsd5.3 system i386, freebsd5.3 status major 2 minor 0.1 year 2004 month 11 day 15 language R # Windows XP:> version_ platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major 2 minor 0.1 year 2004 month 11 day 15 language R # Windows 98:> version_ platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major 2 minor 0.1 year 2004 month 11 day 15 language R Please let me know if I can add any more details- Thank you, Rob _________________________________________________________ Rob Keefe Cooperative Forestry Research Unit University of Maine
MSchwartz@MedAnalytics.com
2005-Jan-07 05:11 UTC
[Rd] R 2.1 dies when some text() arguments are NULL (PR#7477)
On Fri, 2005-01-07 at 04:44 +0100, Robert_Keefe@umit.maine.edu wrote:> > [My very deepest apologies in advance if this is not plain text; I am > recently at a new job in a windows-based environment and have > not yet set up my favorite old (pine) system. I have tried my best to > strip all formatting from the message..] > > I'm overlaying plots of tree locations mapped before and after > logging activities and checking my matched individuals using > several arguments to text() which plot tree attributes by or > near their (x,y) locations. R 2.1 dies when a coding error on > my part results in a null data.frame... > > Example: > > plot(1) > text(NULL ~ NULL, label=NULL) > > This happens for me under Windows 98 and XP. Andrew Robinson > was kind enough to try it for me on FreeBSD with the same result. > The code returns nothing for me under R 1.9 on Windows 98 and > XP (same win. machines). The version details of the three > are as follows: > > # FreeBSD: > > >version > _ > platform i386-portbld-freebsd5.3 > arch i386 > os freebsd5.3 > system i386, freebsd5.3 > status > major 2 > minor 0.1 > year 2004 > month 11 > day 15 > language R > > # Windows XP: > > > version > _ > platform i386-pc-mingw32 > arch i386 > os mingw32 > system i386, mingw32 > status > major 2 > minor 0.1 > year 2004 > month 11 > day 15 > language R > > # Windows 98: > > > version > _ > platform i386-pc-mingw32 > arch i386 > os mingw32 > system i386, mingw32 > status > major 2 > minor 0.1 > year 2004 > month 11 > day 15 > language R > > Please let me know if I can add any more details- > > Thank you, > > RobYour subject line and body refer to R version 2.1, which technically does not exist as 2.1.0 is the current development version. However, your output above indicates that R version 2.0.1 is what you have installed on the three platforms. That is the current released version. In either case, using: Version 2.0.1 Patched (2005-01-02) I get the following:> plot(1) > text(NULL ~ NULL, label=NULL)Error in text(xy.coords(x, y, recycle = TRUE), labels, adj, pos, offset, : no coordinates were supplied Same thing with:> plot(1) > text(NULL, NULL, label = NULL)Error in text.default(NULL, NULL, label = NULL) : x and y lengths differ in text(). No crash. There is the following note in the NEWS file for 2.0.1 patched: "text()'s default method could segfault if passed 0-length coordinates." So I believe that this fixes the issue that you have posted about. A patched version for Windows is available via CRAN: http://www.cran.mirrors.pair.com/bin/windows/base/rpatched.html and a source code tarball is available for the *nix versions via the FTP link under the R Sources link: ftp://ftp.stat.math.ethz.ch/Software/R/ HTH, Marc Schwartz