similar to: anoncvs for R should work again

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]

(cc'ing R-devel as well). >>>>> "RH" == Rich Heiberger <rmh@surfer.sbm.temple.edu> writes: RH> using Mx-R on ntemacs, when I type RH> help(lm) RH> I get an emacs *help[R](lm)* buffer with "^M" on each line. I RH> tried a few ways to make them go away and none worked. RH> When I tried help.start() then internet

I know it isn't done often, so apologies if this posting offends. But after posting the announcement of the open position to a number of other software lists, I was thinking, why not to R-help? (esp since knowing R is one option for meeting the requirements). best, -tony We are looking for an masters level statistician. If interested, feel free to apply. I am available to answer

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.

(please keep sebastian cc'd) Doug - we are managing to trigger, both on R 1.1.1 and R 1.2.0 (proper versions of NLME), the following: > warnings(model1) Warning message: Singular precision matrix in level -1, block 1 Error in cat(list(...), file, sep, fill, labels, append) : argument 2 not yet handled by cat > suggesting either an "inappropriate" use of cat(), or a

So, lots of comments about PICT. If you have to go there (EPS to PICT, but vectorized PICT for an investigator to use), does anyone have suggestions? I've found EPStoPICT (www.artage.com), any other thoughts? (these are real Mac people, so while I could go the WMF route, it's about as palatable as EPS). best, -tony -- A.J. Rossini Rsrch. Asst. Prof. of Biostatistics UW

Not having a Mac in front of me, but having a consultee with one, I pose the following question: Is there an easy way to transfer a Mac excel data set (columns, etc) into R (only on a Mac)? Verified existence of such a thing would be great; I've been uncertain as to the best way to get data in/out of R (on the mac, of course). An actual algorithm/procedure would be met with great rejoicing.

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.

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

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

> > Date: Fri, 11 Aug 2000 15:34:01 -0400 > From: Rodney Sparapani <rodney at stat.duke.edu> > Subject: Emacs Speaks Statistics (ESS) BOF at JSM > > Statisticians: > > For those of you going to the JSM in Indy next week, we are having an > Emacs Speaks Statistics (ESS) BOF Monday, August 14th, 5:30 at the Hyatt > in Salon A. Please bring along your comments,

>>>>> "BR" == B Rowlingson <B.Rowlingson at lancaster.ac.uk> writes: >> Could you check and see if the variable is global or >> buffer-local? ("C-h v ess-dump-filename-template" should do >> it). BR> ess-dump-filename-template's value is "rowlings.%s.S" Local in BR> buffer *S+5*; global value is

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

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 <-

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

What about "Rtips" at http://lark.cc.ukans.edu/~pauljohn/R/statsRus.html ? Regards, Heinrich. > -----Urspr?ngliche Nachricht----- > Von: rossini at blindglobe.net [mailto:rossini at blindglobe.net] > Gesendet: Mittwoch, 26. Juni 2002 14:48 > An: r.hankin at auckland.ac.nz > Cc: r-help at stat.math.ethz.ch > Betreff: Re: [R] sapply() and Monte Carlo > > >

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

I've been having fun with crashes recently, both with Blackdown JDK and the IBM JDK, while building up the ROrca interface. It's just not stable for me, and my lack of grokking the intimate details of lazy evaluation is not helping. Brushing works, though, which is nice. I need a bit more stability before I get the "data add/change" features in line, though. Anyway, other

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