[Rd] Error description not prompted but R crashes instead (PR#14181)
Full_Name: Philippe Demoulin Version: R version 2.10.0 (2009-10-26) OS: OSX 10.4.11 on PPC Submission from: (NULL) (80.236.193.215) Test: At the prompt, just type (without quotes) "> x=1:3". Please note the redundant ">" explictly keyed in for simulating the error On my daughter MacMini G4 PPC 10.4.11 JSE 1.5, an normal error "Unexpected character >" is issued. But on my iMac 10.4.11 PPC JSE 1.5 (so, same environment), R crashes: Exception: EXC_BAD_ACCESS (0x0001) Codes: KERN_PROTECTION_FAILURE (0x0002) at 0x0020 Thread 0 Crashed: 0 libobjc.A.dylib 0x90a410f8 objc_msgSend + 24 1 org.R-project.R 0xdae4 Re_RBusy + 60 (Rcallbacks.m:178) 2 libR.dylib 0x003cdcf0 R_ReplDLLdo1 + 64 3 org.R-project.R 0x00018358 run_REngineRmainloop + 368 (Rinit.m:521) 4 org.R-project.R 0x00010a0c -[REngine runREPL] + 484 (REngine.m:181) 5 org.R-project.R 0x2a38 main + 1040 (main.m:140) 6 org.R-project.R 0x25ac _start + 760 7 org.R-project.R 0x22b0 start + 48 Regards Philippe Demoulin, Tilff, Belgium __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] seq.int broken (seq as well) (PR#14169)
Petr, Aside of the fact that the argument about someting bad being good because documented is strongly overused. I think this does NOT behave as documented, because a) the behaviour cannot be explained by rounding error on double precision. b) 1e7 is not even outside the range of integer calculation Up to the limit of a) or at least upto b) any expression of the type seq(a, b, by=b) should only return a but not b. Also something like seq.int should ONLY use and return integer, for performance reasons, but even more so for reliability: the reported behaviour is not just a little bit wrong. Since seq is used for looping in R, the looping of the language is broken. This can have severe consequences like accessing beyond the limits of an array. If C-code is involved, this can crash R. In the worst case algorithms can silently do wrong. Being an admirer of R since its early days, I was shocked to see this, and as a consequence, I suggest we do our homework and suspend -- for a year or two -- any claims that R can be used productive such as SAS. Yours regretfully Jens Oehlschlägel __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] interfacing R with RFC2445
R Developers, I cannot seem to find an R package that can read/write iCalendar (RFC2445) files. I have found the libical library at sourceforge. I've used it briefly so it may be what I need, but again no R interface. I may have to connect the dots myself. I'm guessing this might be useful for someone other than myself (scrubbing files, performing temporal analysis on events, etc). Since I need to perform these tasks, I will probably do it using R if I can get away with it. Can anyone provide some basic advice for me? A list object of entries? The original data will be coming from an SQL table/view result. Is there any interest in helping with funding such a small project? Thanks, Jeff. __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] (PR#14181) Error description not prompted but R crashes instead
Is this from the Mac GUI R.app. You are asked to send bug reports on that to the R-sig-mac list, as this is almost certainly not a bug in R but R.app, a separate project. On Mon, 11 Jan 2010, pdemou...@gmail.com wrote: Full_Name: Philippe Demoulin Version: R version 2.10.0 (2009-10-26) OS: OSX 10.4.11 on PPC Submission from: (NULL) (80.236.193.215) Test: At the prompt, just type (without quotes) "> x=1:3". Please note the redundant ">" explictly keyed in for simulating the error On my daughter MacMini G4 PPC 10.4.11 JSE 1.5, an normal error "Unexpected character >" is issued. But on my iMac 10.4.11 PPC JSE 1.5 (so, same environment), R crashes: Exception: EXC_BAD_ACCESS (0x0001) Codes: KERN_PROTECTION_FAILURE (0x0002) at 0x0020 Thread 0 Crashed: 0 libobjc.A.dylib 0x90a410f8 objc_msgSend + 24 1 org.R-project.R 0xdae4 Re_RBusy + 60 (Rcallbacks.m:178) 2 libR.dylib 0x003cdcf0 R_ReplDLLdo1 + 64 3 org.R-project.R 0x00018358 run_REngineRmainloop + 368 (Rinit.m:521) 4 org.R-project.R 0x00010a0c -[REngine runREPL] + 484 (REngine.m:181) 5 org.R-project.R 0x2a38 main + 1040 (main.m:140) 6 org.R-project.R 0x25ac _start + 760 7 org.R-project.R 0x22b0 start + 48 Regards Philippe Demoulin, Tilff, Belgium __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel -- Brian D. Ripley, rip...@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] rgl: bogus configure[.ac] (PR#14183)
Full_Name: Jens Elkner
Version: 2.10.1
OS: Solaris
Submission from: (NULL) (141.44.24.14)
The configure.ac is bogus, since it does not include additional library pathes
at the right place and thus fails. E.g.:
$R_HOME/bin/R CMD INSTALL -l $PROTO/R/library \
--configure-args='--with-gl-includes=/usr/X11/include/NVIDIA
--with-gl-libs=/usr/X11/lib/NVIDIA/amd64' $TMPBUILD
And thus configure tries: cc ... -lGL -L/usr/X11/lib/NVIDIA/amd64
which obviously fails: the path needs to be placed before -l...
So the correct way is to add the path to LD_FLAGS instead of LIBS.
To get the path at the correct place into the to PKG_LIBS limitted Makevars,
one
needs to add the path, when the LIB expansion has been done. Suggested patch:
--- rgl/configure.ac.orig Mon Feb 2 14:31:24 2009
+++ rgl/configure.acTue Jan 12 06:16:52 2010
@@ -157,9 +157,12 @@
AC_ARG_WITH(gl-libs,
[ --with-gl-libs=DIR specify location of OpenGL libs],
-[LIBS="${LIBS} -L${withval}"]
+[LDFLAGS="${LDFLAGS} -L${withval}"
+L_LIB="-L${withval}"
+]
)
+
AC_ARG_WITH(gl-libname,
[ --with-gl-libname=NAME specify Library name (defaults to "GL")],
[lGL=${withval}], [lGL=GL]
@@ -179,6 +182,9 @@
if test "x$this" != xyes; then
AC_ERROR([missing required library ${lGLU}])
fi
+if test x$L_LIB != x; then
+ LIBS="${L_LIB} ${LIBS}"
+fi
## --- FTGL --
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
[Rd] MCMCpack + SunStudio12 patch
Hi, FYI: made some patches to get MCMCpack compiled with Sun StudioCompiler 12 (on Solaris). If somebody wanna try/comment on it: http://iws.cs.uni-magdeburg.de/~elkner/R/MCMCpack-studio.patch Regards, jel. -- Otto-von-Guericke University http://www.cs.uni-magdeburg.de/ Department of Computer Science Geb. 29 R 027, Universitaetsplatz 2 39106 Magdeburg, Germany Tel: +49 391 67 12768 __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] (PR#14181) Error description not prompted but R crashes instead
On Jan 12, 2010, at 4:04 , Prof Brian Ripley wrote: > Is this from the Mac GUI R.app. You are asked to send bug reports > on that to the R-sig-mac list, as this is almost certainly not a bug > in R but R.app, a separate project. > Even worse, this is a known bug long fixed in R 2.10.1 and hence a doubly pointless report ... Cheers, Simon > On Mon, 11 Jan 2010, pdemou...@gmail.com wrote: > >> Full_Name: Philippe Demoulin >> Version: R version 2.10.0 (2009-10-26) >> OS: OSX 10.4.11 on PPC >> Submission from: (NULL) (80.236.193.215) >> >> >> Test: >> At the prompt, just type (without quotes) "> x=1:3". Please note >> the redundant >> ">" explictly keyed in for simulating the error >> >> On my daughter MacMini G4 PPC 10.4.11 JSE 1.5, an normal error >> "Unexpected >> character >" is issued. >> >> But on my iMac 10.4.11 PPC JSE 1.5 (so, same environment), R crashes: >> Exception: EXC_BAD_ACCESS (0x0001) >> Codes: KERN_PROTECTION_FAILURE (0x0002) at 0x0020 >> >> Thread 0 Crashed: >> 0 libobjc.A.dylib 0x90a410f8 objc_msgSend + 24 >> 1 org.R-project.R 0xdae4 Re_RBusy + 60 >> (Rcallbacks.m:178) >> 2 libR.dylib 0x003cdcf0 R_ReplDLLdo1 + 64 >> 3 org.R-project.R 0x00018358 run_REngineRmainloop + 368 >> (Rinit.m:521) >> 4 org.R-project.R 0x00010a0c -[REngine runREPL] + 484 >> (REngine.m:181) >> 5 org.R-project.R 0x2a38 main + 1040 (main.m:140) >> 6 org.R-project.R 0x25ac _start + 760 >> 7 org.R-project.R 0x22b0 start + 48 >> >> >> Regards >> Philippe Demoulin, Tilff, Belgium >> >> __ >> R-devel@r-project.org mailing list >> https://stat.ethz.ch/mailman/listinfo/r-devel >> > > -- > Brian D. Ripley, rip...@stats.ox.ac.uk > Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ > University of Oxford, Tel: +44 1865 272861 (self) > 1 South Parks Road, +44 1865 272866 (PA) > Oxford OX1 3TG, UKFax: +44 1865 272595 > > __ > R-devel@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel > > __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] rgl: bogus configure[.ac] (PR#14183)
This isn't an R bug; please don't submit it to the R bug address. You
should email such things to the package maintainer. In this case that's
me, so I'll take a look.
rgl (like many packages) also has its own bug tracker (on R-forge in the
case of rgl); that's another good place to send bugs for such packages.
Duncan Murdoch
On 12/01/2010 10:15 AM, je...@cs.uni-magdeburg.de wrote:
Full_Name: Jens Elkner
Version: 2.10.1
OS: Solaris
Submission from: (NULL) (141.44.24.14)
The configure.ac is bogus, since it does not include additional library pathes
at the right place and thus fails. E.g.:
$R_HOME/bin/R CMD INSTALL -l $PROTO/R/library \
--configure-args='--with-gl-includes=/usr/X11/include/NVIDIA
--with-gl-libs=/usr/X11/lib/NVIDIA/amd64' $TMPBUILD
And thus configure tries: cc ... -lGL -L/usr/X11/lib/NVIDIA/amd64
which obviously fails: the path needs to be placed before -l...
So the correct way is to add the path to LD_FLAGS instead of LIBS.
To get the path at the correct place into the to PKG_LIBS limitted Makevars,
one
needs to add the path, when the LIB expansion has been done. Suggested patch:
--- rgl/configure.ac.orig Mon Feb 2 14:31:24 2009
+++ rgl/configure.acTue Jan 12 06:16:52 2010
@@ -157,9 +157,12 @@
AC_ARG_WITH(gl-libs,
[ --with-gl-libs=DIR specify location of OpenGL libs],
-[LIBS="${LIBS} -L${withval}"]
+[LDFLAGS="${LDFLAGS} -L${withval}"
+L_LIB="-L${withval}"
+]
)
+
AC_ARG_WITH(gl-libname,
[ --with-gl-libname=NAME specify Library name (defaults to "GL")],
[lGL=${withval}], [lGL=GL]
@@ -179,6 +182,9 @@
if test "x$this" != xyes; then
AC_ERROR([missing required library ${lGLU}])
fi
+if test x$L_LIB != x; then
+ LIBS="${L_LIB} ${LIBS}"
+fi
## --- FTGL --
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
__
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Documentation: format of read.table help text (PR#14180)
> -Original Message- > From: r-devel-boun...@r-project.org > [mailto:r-devel-boun...@r-project.org] On Behalf Of > henrik.p...@bio.ntnu.no > Sent: Monday, January 11, 2010 7:10 AM > To: r-de...@stat.math.ethz.ch > Cc: r-b...@r-project.org > Subject: [Rd] Documentation: format of read.table help text (PR#14180) > > Dear R developers, > > I find the format of descriptions of the arguments in the read.table > help text slightly inconsistent. > > For example, the logical arguments comes in seven different formats, > more or less explicit about the consequences of a TRUE (or FALSE): > > 1. check.nameslogical. If TRUE then the names... > 2. blank.lines.skip logical: if TRUE blank lines... > 3. flush logical: if TRUE, scan... > > 4. header a logical value indicating whether... > > 5. allowEscapes logical. Should C-style [...] ? > 6. stringsAsFactors logical: should character [...] ? > > 7. strip.whitelogical. Used only when... > > > First, I believe that a consistent style would make the help > text easier > to follow. Second, I believe that an explicit style (sth > like: "logical. > If TRUE/FALSE then...") is more intuitive It would also be nice if the argument description indicated if the argument should have length one or if it could be a general vector. E.g. a style guide could say the default is that arguments are general vectors but that stringsAsFactors logical scalar. If TRUE ... would tell us that stringsAsFactors had to be of length one (so you cannot say that some columns are to be factors and others not). The code phrase could be ' scalar' or 'a ' but having a common phrase would help. Bill Dunlap Spotfire, TIBCO Software wdunlap tibco.com > and thus preferable > over the > other styles. This might apply to other help pages as well. > > I fully understand that the choice of style in help text is a > matter of > taste and not of very high priority. Nevertheless, I just wish to > mention it as a potential for improvement. > > > Thank you for your great work with R! > > > Best regards, > > Henrik > > > R version 2.10.1 (2009-12-14) > i386-pc-mingw32 > > locale: > [1] LC_COLLATE=English_United Kingdom.1252 LC_CTYPE=English_United > Kingdom.1252LC_MONETARY=English_United Kingdom.1252 LC_NUMERIC=C > > [5] LC_TIME=English_United Kingdom.1252 > > attached base packages: > [1] grDevices datasets splines graphics stats tcltk utils >methods base > > other attached packages: > [1] svSocket_0.9-48 TinnR_1.0.3 R2HTML_1.59-1 Hmisc_3.7-0 > survival_2.35-7 > > loaded via a namespace (and not attached): > [1] cluster_1.12.1 grid_2.10.1 lattice_0.17-26 svMisc_0.9-56 > tools_2.10.1 > > -- > Henrik Pärn > Centre for Conservation Biology > Department of Biology > Norwegian University of Science and Technology > NO-7491 Trondheim > Norway > > Office: +47 73596285 > Fax: +47 73596100 > Mobile: +47 90989255 > > E-mail: henrik.p...@bio.ntnu.no > > __ > R-devel@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel > __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
Re: [Rd] Documentation: format of read.table help text (PR#14180)
On 11/01/2010 10:10 AM, henrik.p...@bio.ntnu.no wrote: Dear R developers, I find the format of descriptions of the arguments in the read.table help text slightly inconsistent. "Slightly inconsistent" is not a bug. Your bug report is quite likely to be ignored because of that and because of the lack of an offer to actually do the work it entails, regardless of the validity of your observations. A way to have attention paid to it is to choose a consistent style, then edit every man page in one of the base packages to follow it, and submit the patches to all those files, with an offer to do the rest once the style is accepted. Expect that there will be one or more rounds of discussion about whether your style really is usable. Writing a style guide would really help. Sound like too much work? I agree. Duncan Murdoch For example, the logical arguments comes in seven different formats, more or less explicit about the consequences of a TRUE (or FALSE): 1. check.names logical. If TRUE then the names... 2. blank.lines.skip logical: if TRUE blank lines... 3. flushlogical: if TRUE, scan... 4. header a logical value indicating whether... 5. allowEscapes logical. Should C-style [...] ? 6. stringsAsFactors logical: should character [...] ? 7. strip.white logical. Used only when... First, I believe that a consistent style would make the help text easier to follow. Second, I believe that an explicit style (sth like: "logical. If TRUE/FALSE then...") is more intuitive and thus preferable over the other styles. This might apply to other help pages as well. I fully understand that the choice of style in help text is a matter of taste and not of very high priority. Nevertheless, I just wish to mention it as a potential for improvement. Thank you for your great work with R! Best regards, Henrik R version 2.10.1 (2009-12-14) i386-pc-mingw32 locale: [1] LC_COLLATE=English_United Kingdom.1252 LC_CTYPE=English_United Kingdom.1252LC_MONETARY=English_United Kingdom.1252 LC_NUMERIC=C [5] LC_TIME=English_United Kingdom.1252 attached base packages: [1] grDevices datasets splines graphics stats tcltk utils methods base other attached packages: [1] svSocket_0.9-48 TinnR_1.0.3 R2HTML_1.59-1 Hmisc_3.7-0 survival_2.35-7 loaded via a namespace (and not attached): [1] cluster_1.12.1 grid_2.10.1 lattice_0.17-26 svMisc_0.9-56 tools_2.10.1 __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
