similar to: seek helppage. Possible clarification?

In order to quickly install lots ("all") of CRAN packages on a nice new machine here, I did the following: install.packages(CRAN.packages()[,1]) 1. Is this sensible (other than the obvious, "who needs everything"?) 2. Should this be made easier to do (added as a possible option)? best, -tony -- A.J. Rossini Rsrch. Asst. Prof. of Biostatistics U. of Washington

We have two lectureship positions advertised at the moment. One (or both) of these is intended for a specialist in statistical computation. The salary doesn't look great by international standards but there are compensations: housing is cheap, the air is clean, and the worlds problems are far away. We also have a track record of getting tenure for computing people.

In a recent post, I reported that help.start worked with Mozilla, although it had not worked before. I now have discovered the difference. I'm using RH Linux 7.1 and Mozilla 0.9.5 and R 1.3.1. When I tried it before, I had use ESS to start R. If I run R without ESS, help.start() will start the browser (Mozilla), but help(apply), for example, will produce help in ESS but not in the browser.

I'd like to get summary statistics (really just a mean would be fine) for a vector in a data frame, but split based on the value of another vector. That is, I have a data frame (hcd.df) with variables datecat (which is always 1 or 2) and auth.sum (-8..+8). I've used xtabs to get chi-square comparisons, but what I need now is a simple mean of auth.sum where datecat is 1 and another where

Probably wrong group for this, but a quick question. I've just switched from emacs to Xemacs. In reinstalling ESS 5.1.19 I keep getting the following error when loading Xemacs: "Error in init file: Symbol's function definition is void: w32-using-nt" I've debugged the ess-site.el file, which is where the error originates from. The line causing the difficulty is:

could we add an argument to suppress printing of levels for print.factor()? i.e.: print.factor <- function (x, quote = FALSE, print.levels = TRUE, ...) ## MODIFIED { if (length(x) <= 0) cat("factor(0)\n") else print(as.character(x), quote = quote, ...) if (print.levels) { ## ADDED cat("Levels:

Thanks to Tony Plate for letting me know what the abbreviate.arg option does. I think this could be made more flexible (I.e. =TRUE, =FALSE, =#, where # would be passed to the abbreviate min.length argument). But it follows the example I was given. "print.factor" <- function (x, quote = FALSE, max.levels=5, print.levels = {if (max.levels==0) FALSE else TRUE},

Suppose one wanted to build an R wizard for packaging... One would need to specify and create: package-name/ package-name/Description package-name/INDEX package-name/R/ package-name/man/ package-name/src/ >From this, we'd "probably" like to: - script-up the Description - provide a script for regenerating INDEX by calling out to the

One thing that has driven me batty to no end is the fact that ESS doesn't handle R's "exit" well. Any thoughts as to the default? I'm thinking of something like: current: do what currently is done C-u Y C-c C-q : quit and save workspace C-u N C-c C-q : quit and don't save workspace. Comments? best, -tony -- A.J. Rossini Rsrch. Asst.

On RedHat linux 7.2, I upgraded (?) Emacs and R at the same time and now I'm getting some funny business with R and ESS. I reinstalled ESS from the tarball and re-byte-compiled. In particular, help.start() does work, and ESS works to send text regions to the R process, but ?function does not return anything, and the status line says "ESS process not ready. Finish your command

There's a bit of a symmetry issue, which may or may not be important (led to 15 seconds of confusion until I got my bearings straight): row.names, rownames; colnames, no col.names? I _NOW_ realize the difference between row.names and rownames, but is there any reason not to have col.names for re-naming columns in a data.frame? (if there is, I don't particular need to know it, but it

for Spam. In the process of setting up a more effective spam filtering system, I just noticed that bogofilter, which implements extensions of the (a?) "Naive Bayes" text classification approach, will dump out R data frames; the man page suggests how to "integrate" it with R for verification. (sort of, that is). Anyway, for those of you looking for silly and perhaps

Duncan - (but sent to R-devel, for any other thoughts?) How do you envision RSPerl being used? Without having seen the details, I can think of playing with strings; is there a preferred incantation for stringification of R objects (serialization)? (actually, the main problem I'm having is that I know how I want to code something like: Robject <-

After installing nlme_3.1-7.tar.gz under the latest R (1.2.0, via anonymous CVS), I'm unable to get it to work, i.e.: rossini 94 > /usr/local/bin/R R : Copyright 2000, The R Development Core Team Version 1.2.0 Under development (unstable) (2000-12-06) R is free software and comes with ABSOLUTELY NO WARRANTY. You are welcome to redistribute it under certain conditions. Type

Is there a simple way to determine at run-time the directory with HTML version of help pages? (to see where I'm heading here, if you've got w3 installed under (X)Emacs on Unix _or_ a Microsoft OS, do w3-find-file and load $(R_HOME)/library/html/whateverfunction.html It's slightly nicer than the standard ESS text-help. best, -tony -- A.J. Rossini Rsrch. Asst.

Hi to all who suffer from rterm not shutting down in xemacs/ESS on windows NT or 2000. Also hi to those who could eventually help. Here is some more information which could help and some ENCOURAGEMENT to contribute to a solution. 1. It may be an xemacs problem but it is more likely an interaction between rterm/comint/and xemacs. In fact, the problem started occurring around version R 1.6.0. I

What's the current best approach to dealing with the nlminb problem? Are there other similar constrained optimizers, or is current wisdom just to code it in C (or similar...)? best, -tony -- A.J. Rossini Research Assistant Professor of Biostatistics Biostatistics/Univ. of Washington (Th) Box 357232 206-543-1044 (3286=fax) Center for AIDS Research/HMC/UW (M/F) Box 359931

(thanks to thomas for assistance; I'm posting this for archival reasons). Question: My silly dataset looks like "00223423", "00234234", ... and when I read it in, I get a factor which looks like and integer. How do I convert it back? (i.e. I've got something like: > hivplasma [1] 00050400 00219000 00136000 00000906 00142000 00002150 00000400 00001880 [9]

I've been having problems building vignettes in bioconductor packages with R-devel. Turns out that Rdevel/share/texmf/hyperref.cfg wants Blue and Red predefined, when only blue and red are defined (as of rsync Rdevel, Sept 10th). This is on a Debian unstable system (Sept 9th version). Might not apply to all other tetex systems. Seems to have bitten the bioconductor build system, though.

Okay, so I've got a set of directories containing unpacked R packages (unpacked, since for various sensible reasons, I'm editing them in warped ways). Supposing that these are in: /home/rossini/Runpacked I thought I could do something like: install.packages("MyUnPackedRPackage", # top package name lib="/home/rossini/lib/R", # where