Things not required to be there needs to be placed in inst/, so they will
make it into the binary package. This is where I had to put the NEWS file
for my package, as I learned here a while ago.
Andy
From: Gabor Grothendieck
I have NEWS, WISHLIST and THANKS files in the 'dyn' package
in
Let me add to the pot:
I think Robert and Brian are against imposing additional _requirement_ on
packages to provide an overview in .Rd, and I tend to agree with that
sentiment.
However, if such a facility is made optional (like vignettes) for package
author/maintainer, then I have no problem
From: Kurt Hornik
Brian D Ripley writes:
As far as I know R 2.0.1 was built on AIX. A quick diff
gannet% diff ~/R/R-2.0.1/src/modules/X11/Makefile
~/R/R-2.1.0/src/modules/X11/Makefile
54c54
$(SHLIB_LINK) -o $@ $(R_X11_la_LDFLAGS) $(R_X11_la_OBJECTS)
$(R_X11_la_LIBADD)
I believe that's been fixed in R-2.1.0. Please check.
I got the same segfault with R-2.0.1 on SuSE Linux x86_64, but on both that
machine and my WinXP latop, I get syntax error. E.g.,
R : Copyright 2005, The R Foundation for Statistical Computing
Version 2.1.0 Patched (2005-05-12), ISBN
Please read the R FAQ on how to report bugs on contributed packages: Not to
R-bug: as that's not a bug in R itself.
example(varImpPlot) ran just fine in R-2.1.0. Please show _the_ example
where you got the error.
Andy
From: [EMAIL PROTECTED]
Full_Name: Daniel Normolle
Version: 2.0.1
OS:
From: [EMAIL PROTECTED]
Hello,
I found a potential problem in R 2.1.0 (and R 2.0.1)
I expect that
tmp - FUN(x1, x2, x3, x4)
as.data.frame(tmp)
is the same as
as.data.frame(FUN(x1, x2, x3, x4))
since the tmp variable in this case is unnecessary.
However, below I will
From: Lars Schouw
Can anyone explain what this means the NEWS for for
2.1 beta?
R is now linked against ncurses/termlib/termcap only
if
readline is specified (now the default) and that
requires it.
I get an error while running configure:
checking for history_truncate_file... no
From: Ross Darnell
Liaw, Andy wrote:
From: Liaw, Andy
From: Ross Darnell
A good point but what is the value of storing a large set of
predicted
values when the values of the explanatory variables are lost
(predicted
values of what?). I thought the purpose of objects
Please consider adding the following to the next R release (I understand
that it's too late for R-2.1.0). It adds the menu item `Search R Site' in
the `Help' menu in Rgui (which calls RSiteSearch() on the input string).
Best,
Andy
--- R-beta/src/gnuwin32/rui.c 2005-03-06 09:41:40.0
From: Prof Brian Ripley
On Tue, 12 Apr 2005, Liaw, Andy wrote:
Please consider adding the following to the next R release
(I understand
that it's too late for R-2.1.0). It adds the menu item
`Search R Site' in
the `Help' menu in Rgui (which calls RSiteSearch() on the
input
From: Peter Dalgaard
Liaw, Andy [EMAIL PROTECTED] writes:
From: Peter Dalgaard
Liaw, Andy [EMAIL PROTECTED] writes:
** data
** moving datasets to lazyload DB
** inst
** preparing package for lazy loading
/SFS/user/ry/liawand/R/Rbuild/bin/INSTALL[325
From: [EMAIL PROTECTED]
Liaw, Andy [EMAIL PROTECTED] writes:
OK, after re-building R with the debug flags, I don't think
I'm getting
anything useful. GDB gives me:
gdb) run --vanilla /ltmp/try.R
Starting program: /SFS/user/ry/liawand/R/Rbuild/bin/exec/R
--vanilla
/ltmp
I thought I'd give this another shot before the official release. I tried
building R-beta_2004-04-07 on the AIX system that I have access to, and it
seemed to failed at lazy-loading survival. I'd very much appreciate any
pointer on what to try or look for next.
1. I set OBJECT_MODE to 64 for
Dear R-devel,
A colleague was recently surprised by the behavior of the : operator (as
sequence generator), so I decided to dig a little bit. Here are a couple of
things I found.
- ?: seems a bit imprecise. The Details section says:
`The operator : and the seq(from, to) form generate the
From: Thomas Lumley
On Thu, 7 Apr 2005, Liaw, Andy wrote:
- It seems counter-intuitive to allow non-integer values as
operand for :.
It also seems a bit odd that 1:2 returns something with storage.mode
integer, whereas 0.5:2 gives doubles. Would it make sense
to disallow
non
I saw the following in a Springer catalog recently:
Developing Statistical Software in Fortran 95
Lemmon, David R., Schafer, Joseph L.
2005, XVI, 328 p., Softcover
ISBN: 0-387-23817-4
which has the following in the description:
Through detailed examples, readers are shown how to call Fortran
[Re-directing to R-devel, as I think this needs changes to the code.]
Can I suggest a modification to stats:predict.princomp so that it will check
for column (variable) names?
In src/library/stats/R/princomp-add.R, insert the following after line 4:
if (!is.null(cn - names(object$center)))
OK, this is really a new one for me. Just thought I'd try building R-devel
(2005-03-15) on our AIX power machines and see if there are any problems. I
used the settings in the Appendix B of R-admin, using the set of variables
that Tim Hoar used (i.e., trying to build 64-bit R using native
Given the confusions that some users have regarding package bundles, I'd
like to suggest some changes to how package bundles are handled.
My understanding is that a package bundle is essentially a convenient way to
distribute related packages, so that users can download/install them in one
shot.
Hi all,
I was playing with R-devel 2005-03-08 on winXP Pro (installed from Duncan's
installer). My .Rprofile contains:
options(defaultPackages=c(getOption(defaultPackages), mypkg),
chmhelp=TRUE)
but mypkg has not yet been installed in the library/ directory of this
installation. When
From: Gorjanc Gregor
! Look after character !
From: Prof Brian Ripley [mailto:[EMAIL PROTECTED]
You too have not give an reproducible example!
! Yes, I was not able to do it from my data. But bellow is one. It is
! a stupid one, but it works. The problem is use of as.data.frame in
!
From: John Marsland
I am having no luck compiling R-2.0.1 on a Windows XP
platform. I have not had these problems when compliling
previous versions of R.
I've installed all the recommended software and tools. But I
cannot get round this error message:
make
make[1]: `Rpwd.exe' is
It really pays to RTFM sometimes:
http://cran.r-project.org/bin/windows/base/rw-FAQ.html#The-output-to-the-con
sole-seems-to-be-delayed
Andy
From: [EMAIL PROTECTED]
Full_Name: Markus Liedgens
Version: Ever since 1.6
OS: Windows 98 / Windows Xp
Submission from: (NULL) (81.63.111.226)
From: Douglas Bates
Patrick Burns wrote:
The following is at least as much out of intellectual curiosity
as for practical reasons.
On reviewing some code written by novices to R, I came
across:
crossprod(x, y)[1,1]
I thought, That isn't a very S way of saying that, I wonder
I don't know the issue, but a search for texi2dvi under my TeXLive tree
found a shell script by that name in the directory
texmf\doc\graphics\texdraw. Can that be made to work, in conjunction with
the shell in the Rtool bundle? Just curious. (This is fpTeX installed from
the TeXLive 2004 CD.)
You seem not to have followed the instructions carefully enough. Note that
the `make' that is picked up is the wrong one. You probably need to move
path to R tools, etc. to the beginning of the PATH variable.
Also, you might want to install a suitable TeX system (e.g., fpTeX).
Andy
From:
From: [EMAIL PROTECTED]
Dear R Developers:
I have been playing with R, release 2.0.1 for a week now and
have detected =
that all stat functions related to distribution probabilities
have the same=
problem:
1.- According to the manual the log.p parameter is always the
last one.
Mick,
Questions about packaging are better suited for R-devel (so I'm redirecting
there).
My understanding is that R CMD build mypkg, w/o other flags, does a bit more
than tar/gzip: it skips junk files (e.g., .o, .a, .so, etc.). Also, the
--force flag will update the INDEX file.
On Windows,
Gabor,
Here are my takes on these:
1: Not that I know of. Only the package developer(s) see the code this way,
and is really not different from 3. (A developer would not need 3, as R CMD
check can be done with 1.)
2 4: As Uwe said, the location is `whereever you like'. You would create
them
absolute path and where the various directories wind up also in
terms of example absolute paths it would clear up a lot.
Liaw, Andy andy_liaw at merck.com writes:
:
: Gabor,
:
: Here are my takes on these:
:
: 1: Not that I know of. Only the package developer(s) see
the code
From: Vadim Ogranovich
Hi,
It seems that a formal function argument can not default to an outer
variable of the same name:
x - foo
This is just a red herring... Even w/o this you get the same thing:
ff - function(x=x) x
ff()
Error in ff() : recursive default argument reference
From: Prof Brian Ripley
On Sun, 7 Nov 2004, Kjetil Brinchmann Halvorsen wrote:
Prof Brian Ripley wrote:
install/update.packages() in R-devel has been enhanced in
a few ways,
notably to handle multiple repositories, and to allow
packages to be
reinstalled under the current
Gabor,
I guess is that you did not try to run R CMD INSTALL before R CMD check. R
CMD check will try to install the package first (in pkg.Rcheck), and only if
that's successful would checks be done.
The installation process will concatenate all R files in R/ to a single file
and essentially
From: Gabor Grothendieck
I have question regarding package installation.
What is the difference between check, INSTALL, build and
build --binary, which imply which and what order does one
normally perform them? I have been trying:
R CMD build /mypackage
R CMD check /mypackage
R CMD
functions, bypassing the Free().
Best,
Andy
From: Duncan Murdoch
On Mon, 25 Oct 2004 22:26:35 -0400, Liaw, Andy [EMAIL PROTECTED]
wrote :
Dear R-devel,
Apologies for bothering y'all with this seemingly perennial
question. A
user reported problem with my most recent version
Dear R-devel,
Apologies for bothering y'all with this seemingly perennial question. A
user reported problem with my most recent version of randomForest (4.4-1),
and I was able to reproduce it with his data with R-2.0.0 patched
(2004-10-24) on WinXP Pro. The problem is that it crashes R on
In the 2.0.0 version of the R FAQ (html format), the links at the end of
most sections look wrong in many places (but not all), starting with Section
1.1:
[...]
You can also obtain it by writing to the Free Software Foundation, Inc., 59
Temple Place - Suite 330, Boston, MA 02111-1307, USA.
From: Prof Brian Ripley
On Sun, 24 Oct 2004, Liaw, Andy wrote:
In the 2.0.0 version of the R FAQ (html format), the links
at the end of
most sections look wrong in many places (but not all),
starting with Section
1.1:
[...]
You can also obtain it by writing to the Free
Please give a _reproducible_ example on the current version of R (2.0.0).
It works fine for me using the example in ?ppr:
rock.ppr - ppr(log(perm) ~ area1 + peri1 + shape,
+ data = rock, nterms = 2, max.terms = 5, optlevel=0)
rock.ppr
rock.ppr - ppr(log(perm) ~
[I'm diverting this to R-devel, as I believe my questions are more
appropriate for that.]
From: Prof Brian Ripley
[snip]
Luke Tierney recommends removing the NAMESPACE file during
development of
a package if you need frequent access to debug/change its functions.
But isn't that a bit
From: Duncan Murdoch
[snip]
For 2.1.0, would it be reasonable to extend debug() the way ? was
extended? E.g. allow
debug(plot(x))
and have the debugging flag be set on the appropriate function,
whatever and wherever it happens to be? There's some ambiguity with
S3 methods (did you
From: Gustavo Corral
Hello:
I want to write my thesis about the design and implementation
of R, but I have
not enough information and that's why I'm looking for the
article Lexical
Scope and Statistical Computing from Gentleman, but I don't
want for the
moment
subscribe me to the
In ?options:
'warning.expression': an R code expression to be called if a
warning is generated, replacing the standard message. If
non-null is called irrespective of the value of option
'warn'.
Is there a missing `it' between `non-null' and `is'?
Best,
Andy
Dear R-devel,
I've noticed that if the method for formula is not exported in NAMESPACE,
then update() would fail. As an example:
library(e1071)
data(iris)
iris.svm - svm(Species ~ ., data=iris)
update(iris.svm)
Error in eval(expr, envir, enclos) : couldn't find function svm.formula
The
From: Prof Brian Ripley
On Tue, 14 Sep 2004, Liaw, Andy wrote:
Dear R-devel,
I've noticed that if the method for formula is not exported
in NAMESPACE,
then update() would fail. As an example:
library(e1071)
data(iris)
iris.svm - svm(Species ~ ., data=iris)
update
Thanks to Peter and Brian for the info. Also, thanks to Fritz and Kurt for
the fantastic service they provided!!
Best,
Andy
-Original Message-
From: Peter Dalgaard [mailto:[EMAIL PROTECTED]
Sent: Thursday, September 09, 2004 1:16 PM
To: Prof Brian Ripley
Cc: Liaw, Andy; [EMAIL
The `problem', I think, is your expectation that the output of apply(a, 2,
var) to be of the same dimension as apply(a, 2, sd) if a has dimensions 2.
Note that:
sd(matrix(1:9, 3, 3))
[1] 1 1 1
var(matrix(1:9, 3, 3))
[,1] [,2] [,3]
[1,]111
[2,]111
[3,]11
This is _not_ a bug. Please read `R Language Manual', section 4.3.2 (as
well as the R FAQ on what constitute a bug).
Andy
From: [EMAIL PROTECTED]
The following code prints [1] 2, as it should
temp-function(ab,...){
print(ab)
}
temp(2,s=3)
However, this code prints [1] 3:
From: Prof Brian Ripley
On Tue, 24 Aug 2004, Peter Kleiweg wrote:
Liaw, Andy schreef:
This is _not_ a bug. Please read `R Language Manual',
section 4.3.2
I don't have the `R Language Manual'. Is it explained somewhere
in the docs that come with R?
Amongst the docs
Dear R-devel,
May I suggest that a new feature be added to a couple of unique() methods?
Sometimes it's useful to have the indices of the original data that the
unique elements come from, so that the original data can be recreated from
the unique()ed data. I suggest that an `index' argument be
How is this a bug? You get exactly what you _should_ get (or otherwise it
might have been a legitimate bug report).
I suspect what you really should use is assign().
Andy
From: [EMAIL PROTECTED]
I am trying to assign the values in a matrix to an object
whose name is
created
From: Thomas Baier
I am confused as to what I need and must do.
Your message said
ftp://cran.r-project.org/pub/R/src/contrib/Devel/rcom_0.97.tar.gz
The DESCRIPTION file inside that download says
Package: rcom
Version: 0.95
Date: 2004-07-14
That's too bad. Sorry for that.
Dear R-devel,
May I suggest the following, or something to its effect (taken from the R
Data Import Export manual, item #8 in section 2.1) be added to the
colClasses argument in ?read.table:
Note that colClasses and as.is are specified per column, not per variable,
and so
include the column of
)],
collapse=_)))
}
}
Best,
Andy
From: Prof Brian Ripley
Use the inst directory - it's documented in Writing R
Extensions, and has
lots of examples on CRAN.
On Thu, 24 Jun 2004, Liaw, Andy wrote:
I've noticed that when I place a `NEWS' file in the
top-level directory
Just want to point out what seems to be a typo in R-admin.texi (still there
in R-devel snapshot of 2004-06-24). Line 1698 of R-admin.texi probably
should read:
@section Java Virtual Machines on Mac
Best,
Andy
Andy Liaw, PhD
Biometrics Research PO Box 2000, RY33-300
Merck Research
From: Friedrich Leisch
[snip]
You might want to take a look at the system.file() function, that
reduces the above quite considerably :-)
Indeed:
news - function(package=NULL, lib.loc=NULL) {
if (is.null(package)) {
newsfile - file.path(R.home(), NEWS)
title - paste(NEWS
From: Toralf Kirsten
Hi Uwe,
thanks for your fast response.
The content of the file izbi/R/First.lib.R of the source
package is as
follows:
*
.First.lib - function(libname, pkgname) {
library.dynam(izbi, package = pkgname, lib.loc = libname)
data(COLS,
IMHO this should be done by the package author(s), without resorting to
changing behavior of library(). I.e., I write a function, say, startGUI(),
that does just that, and print a message in .First.lib() that tells the user
that that's what he/she should execute to startup the GUI. I believe
I believe the problem that Zednek described is specific to Windows, and had
been reported and fixed in the current R-patched.
Cheers,
Andy
From: Roger D. Peng
I have occasionally noticed this too in 1.9.0 but have been unable to
come up with a simple reproducible example. I'll try to look
version of the package, and probably the best qualified person to do
so, too. There's no use filing to R-bugs for bugs in contributed packages.
Do you really expect the R Core developers to fix bugs in contributed
packages?
Andy
thanks
Fred J.
--- Liaw, Andy [EMAIL PROTECTED] wrote
From: [EMAIL PROTECTED]
On Fri, 23 Apr 2004 01:29:18 +0200 (CEST), [EMAIL PROTECTED] wrote:
2. The old one.
When no device has been opened, then q() cleanly closes R.
When a graphics device has been opened, and either is still
open or was
manually closed, then q() hangs. I need to
For those interested:
I compiled R-1.9.0beta_2004-04-02 with pgf77/gcc/g++ and g77/gcc/g++, both
_without_ fast BLAS. Both passed make check-all. Simple crossprod(m) for
fairly large `m' shows that the version compiled with pgf77 is about 15%
faster:
[EMAIL PROTECTED]:scratch]% echo
Dear R-devel,
I'm trying to compile R-1.9.0beta (2004-03-31) with the PGI compilers
(15-day free evaluation: after the 15 days the binaries will stop working).
I'm running into some problems, that I hope some can offer pointers.
1. After configure, the FLIBS include /usr/local/lib' --- note the
From: Prof Brian Ripley [mailto:[EMAIL PROTECTED]
On Fri, 2 Apr 2004, Liaw, Andy wrote:
Dear R-devel,
I'm trying to compile R-1.9.0beta (2004-03-31) with the PGI
compilers
(15-day free evaluation: after the 15 days the binaries
will stop working).
I'm running into some
When running make check-all on R-1.9.0beta (2004-03-31) built with
pgf77/gcc/g++ on the Opteron (running SLES8), I get four warnings about not
finding the DESCRIPTION files for the packages in the VR bundle; e.g.,
cannot open file `/u1/scratch/R-1.9.0/tests/Packages/VR/spatial/DESCRIPTION'
Error
Ivo,
Let me address your points in reverse order:
1. There is a `wishlist' category for bug reports, which I guess you've
overlooked.
2. There is also a `Contributed Documentation' section on the R web site,
which you can submit your contribution. As well, there are a few
introductory level
I was actually thinking along the same line. How about a `live'
installation with Windows binary (including contrib packages)? I know it's
possible for Windows (running live from CD). Maybe the same can be done on
the same CD for other OS's (Mac and Linux)? How about T-shirts and baseball
You seem to have not read the message shown at R startup:
R : Copyright 2003, The R Foundation for Statistical Computing
Version 1.8.1 (2003-11-21), ISBN 3-900051-00-3
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type
Dear R-devel,
A while ago Prof. Ripley had mentioned that g77 on AMD64 isn't the greatest,
and the Salford Fortran is likely to be better. My question is, has anyone
tried compiling R with Salford compilers (or other commercial compilers)?
If so, do they offer performance advantages over GCC?
have been benchmarking as up to 50% faster than g77.
http://www.pgroup.com/
If you (Andy) want to email [EMAIL PROTECTED] she can
put you in
touch with our local expert.
On Thu, 25 Mar 2004, Liaw, Andy wrote:
A while ago Prof. Ripley had mentioned that g77 on AMD64
isn't
(cr system.time(crossprod(m))
[1] 79.36 0.11 85.53 0.00 0.00
I have not check the docs, but my guess is that the library is not threaded.
BTW, this is on dual Opteron 244 with 16GB of PC2100 and ATA100 EIDE HDD.
Best,
Andy
From: Liaw, Andy
Dear R-devel,
Has anyone played with this?
http
Dear all,
Has anyone seen the same problem? I tried compiling R-1.9.0 alpha
(2004-03-16) on our Opteron box running SUSE Linux ES8. I ran:
./configure --enable-R-shlib --with-blas=goto
and got:
Source directory: .
Installation directory:/usr/local
C compiler:
Which Linux is that, and is R compiled from source? I tested both on SLES8
for amd64 and Mandrake 9 on Xeon, compiled R from source in both. I just
get NaN and a warning about the NaN. (I tried qnorm(2) as the subject
says.)
Andy
From: [EMAIL PROTECTED]
Full_Name: Xiong Guanglei
) of memory modules on the motherboard can
be critical for speed with the opterons; well known issue on the
beowulf lists.
best,
-tony
Peter Dalgaard [EMAIL PROTECTED] writes:
Liaw, Andy [EMAIL PROTECTED] writes:
Hi Martin,
When I attended the LinuxWorld Expo in NYC back
Wrap your data frame in data.matrix() and you should be fine; e.g.,
apply(data.matrix(df), 1, all)
Don't think this qualifies as bug...
Andy
From: [EMAIL PROTECTED]
Full_Name: Roland Puntaier
Version: 1.8.1
OS: Windows XP
Submission from: (NULL) (62.99.238.78)
I have a data frame
Dear R-devel,
Can anyone give me some hints on how to go about debugging a strange
segfault in my randomForest package? Here's the scoop:
A user reported segfault when running predict() in the randomForest package.
I asked for the data and code. The combination runs fine under WinXPPro,
but
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
AndyL == andy liaw [EMAIL PROTECTED]
on Fri, 15 Aug 2003 04:08:59 +0200 (MET DST) writes:
AndyL Here's one possible fix (may not be very efficient).
AndyL Change lines 82-83 in $R_HOME/src/base/R/plot.lm.R
to the
Here's one possible fix (may not be very efficient).
Change lines 82-83 in $R_HOME/src/base/R/plot.lm.R to the following:
if (id.n 0) {
qqx - rep(NA, n)
qqy - rep(NA, n)
qqx[!is.na(rs)] - qq$x
qqy[!is.na(rs)] - qq$y
Prof. Ripley,
I grabbed it this morning (EST):
-rw-r--r--1 andy andy 9207525 Mar 27 08:02
R-1.7.0beta_2003-03-27.tar.gz
Cheers,
Andy
-Original Message-
From: Prof. Brian Ripley [mailto:[EMAIL PROTECTED]
Sent: Friday, March 28, 2003 11:01 AM
To: Liaw, Andy
Cc: '[EMAIL
Dear R-devel,
I grabbed 1.7.0 beta from today. make check still fails at base-Ex.R.
The last few lines of the output are:
lsf.str()#- how do the functions look like which I am using?
Error in exists(nam - nms[i], envir = envir, mode = mode) :
F used instead of FALSE
Execution halted
Please, do read the documentation! Try write.table(..., col.names=NA)
Andy
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Sent: Thursday, March 06, 2003 12:42 PM
To: [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
Subject: [Rd] write.table row.names and col.names
with the same shape as. This sort of language would be
helpful.
Jon
On 01/31/03 12:42, Liaw, Andy wrote:
If I'm not mistaken, matrices and arrays are just vectors with dim
attribute. so I'm not surprised that pmax and pmin work as
they do. You
get the same answer by the following
I know that Bert has used something like the following in his S code in the
past:
myfunction - function(..., requiredArg = stop(requiredArg not supplied,
...)
Wouldn't this work? This would be a problem if somehow the requiredArg is
not evaluated, but then it wouldn't be a required agrument...
83 matches
Mail list logo
