similar to: Figure margins too large

Seems to be a day for finding peculiar little things. There is an inconsistency in the behavior of lm vis a vis glm: > x <- rnorm(15) > y <- 1 + 10*x + rnorm(15) > z <- as.factor(rep(c("A","B","C"),rep(5,3))) > xyz <- data.frame(x,y,z) > fit.lm <- lm(y ~ x + z, data=xyz, subset=(z != "C")) > fit.glm <- glm(y ~ x +

I am getting the following: *** Weave Errors *** Error in driver$runcode(drobj, chunk, chunkopts) : Error in eval(expr, envir, enclos) : couldn't find function "pacf.mts" *** Source Errors *** Error in eval(expr, envir, enclos) : couldn't find function "pacf.mts" make[1]: *** [checkVignettes] Error 1 I don't really understand the new namespace mechanism,

I have now tested R 1.3.1 compiled with GCC 3.0.2 on Solaris and the example I previously posted (repeated below) still gives a segmentation fault, as it does with GCC 3.0.1 but not with 2.95.2. The eigen calculation sometimes needs to be repeated several times before the segmentation fault occurs. La.eigen does not cause a problem. (The problem occurred with R-devel too, but I have not tested

I don't seem to be able to initialize a vector of NAs. I hope this is a bug and not a new feature (I realize I have not paid much attention to a lot of recent discussion about NA): platform sparc-sun-solaris2.6 arch sparc os solaris2.6 system sparc, solaris2.6 status major 1 minor 5.0 year

I've been trying to build a bundle with the R-devel snapshot from yesterday. (Yes I know it is unstable and just tell me if I shouldn't be trying to do this yet.) It is failing with the message tar : /dev/rmt/0 : Permission denied It happens after "removing junk files" and just after the message "building dse_2000.9-1.tar.gz" It looks to be coming from the line in

What is the Rd file alias and usage syntax for an S4 assignment method? I have been trying variations on \alias{TSdoc<-,default-method} \usage{ \S4method{TSdoc}{default}(x) <- value but so far I have not got it right according to various codoc, etc, checks. Paul Gilbert ==================================================================================== La version fran?aise

I had a couple of problems getting R 0.90.0. In case anyone has similar problems, I though I would mention them. 1/ I can't use "ftp cvs.r-project.org" from either home or work. The ftp service does not seem to be available. Perhaps this is intended. I only discovered it because I thought the file I got through Netscape might be corrupted. (I don't understand why Netscape

I've noticed that the arguments and result list of prcomp in the mva package (with 61.1) are not quite the same as in the Blue Book and in Splus. Is this intentional or can I change it? If I change it who should I send the code to? Paul Gilbert -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read

There was at one time some discussion about supporting documentation of multiple functions in one Rd file. Is that supported or still in the plans? Paul Gilbert -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the

One of the functions I would like to clean out of my syskern package is a program to send mail. This is a common feature other programs (e.g. bug.report) may use so I would like it to be included in R/base. I will volunteer to do this as it looks like a fairly straightforward dissection of bug.report but I have some questions: Has anyone done this already? Is "mail" a good name or

I was hoping fatal: relocation error: symbol not found: pow_ii: referenced in .../dsefor.so would disappear in R 0.6x, but it has not. I believe this is suppose to come from the f2c lib, but I seem to need just a couple of routines: nm -u dsefor.so Undefined symbols from dsefor.so: .umul pow_dd pow_ii pow_ri For my own purposes I could compile with fortran but that makes it much

I have a generic function test.equal <- function(obj1, obj2, ...) UseMethod("test.equal") however, the only argument that specific methods use is "fuzz" so I would like to change this to test.equal <- function(obj1, obj2, fuzz=???) UseMethod("test.equal") The problem is that some methods use slightly different default values for fuzz than others. How should

Something has changed in the S4 default for coercing. Am I now suppose to use setAs, or is there something else I should do to make this work: R version 2.7.0 (2008-04-22) .... > require("methods") > setClassUnion("OptionalPOSIXct", c("POSIXct", "NULL")) [1] "OptionalPOSIXct" > setClass("TSmetax",

I have a class "tframe" with more specific classes indicating how time is being represented, such as > class(tframe(data)) [1] "ts" "tframe" but now "[.ts" produces warning messages > tframe(data)[2] Warning: Not returning a time series object [1] 2006.25 Even my simplest tests produce hundreds of lines of warnings, so I've commented out

I am having trouble installing cluster from R-1.5.0-recommended : R CMD INSTALL *.tar.gz ... g77 -fPIC -g -O2 -c pam.f -o pam.o g77 -fPIC -g -O2 -c spannel.f -o spannel.o g77 -fPIC -g -O2 -c twins.f -o twins.o cp: cannot access *.so ERROR: compilation failed for package 'cluster' [34] /home/com1/gilp/Rlibs/R-1.5.0-recommended : This is with R-1.5.0 on Solaris 5.8 using gcc

I'm trying to do cvs -d :pserver:anoncvs@cvs.r-project.org:/home/rcvs/src login and I get "authorization failed...rejected access" with the previously advertised password (anoncvs). Has the access been changed or is anyone else having this problem? I'm not very familiar with cvs so I may be making some mistake. Paul Gilbert

o .Platform$OS.type is standardized to be `unix', `mac' or `windows' However, machine() still returns "Unix" rather than "unix". (I'm not sure what it returns on other platforms. Should I be thinking of machine() as standardized already, or is .Platform$OS.type the preferred standard way to check this? Paul Gilbert

When I configure and make R-devel in a directory other than the one where I keep the source, there are a few top level files which are not copied to the build directory. The ones I use are VERSION and date-stamp, to help keep track of what I am testing. Paul Gilbert

The documentation for diff says: If `x' is a vector of length `n' and `differences=1', then the computed result is equal to the successive differences `x[(1:n-lag)] - x[(lag:n)]'. ... This last line should say: `x[(1+lag):n] - x[1:(n-lag)]'. -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-devel mailing list -- Read

Hi, I am using color2D.matplot (...) function of "plotrix" package. I used a matrix of size around 20*20 However, apparently it failed to visualize the matrix and gave the following exception, which I don't have any idea about possible source of this error. "Error in plot.new() : figure margins too large" It would be appreciated if someone points me to the right origin