similar to: New version of eha

A few days ago I uploaded to CRAN a new package called 'eha', which stands for 'Event History Analysis'. Its main focus is on proportional hazards modeling in survival analysis, and in that respect eha can be regarded as a complement and an extension to the 'survival' package. In fact eha requires survival. Eha contains three functions for proportional hazards

A few days ago I uploaded to CRAN a new package called 'eha', which stands for 'Event History Analysis'. Its main focus is on proportional hazards modeling in survival analysis, and in that respect eha can be regarded as a complement and an extension to the 'survival' package. In fact eha requires survival. Eha contains three functions for proportional hazards

Now I try to add some C and Fortan code to my package, so the NAMESPACE file is useDynLib(eha) importFrom(survival, Surv) export(mlreg.fit, risksets) but I get ..... * checking R files for library.dynam ... OK * checking S3 generic/method consistency ... WARNING Error in .try_quietly({ : Error in library(package, lib.loc = lib.loc, character.only = TRUE, verbose = FALSE) :

I noticed the same thing as I reported below, when I upgraded from R-2.0.0-alpha-20040918 to today's beta version: 'R CMD check' failed as before; but after cleaning 'src' from .o and .so files, it worked again. So the two-day-old '.so' file is obviously different from today's. Has compiler directives changed in the two last days? Or what is happening? G?ran On

I'm trying to learn how to use namespaces, so I have created a small package with two functions. My NAMESPACE file is importFrom(survival, Surv) export(mlreg.fit, risksets) because 'mlreg.fit' uses 'Surv' from 'survival'. However I get * checking package dependencies ... WARNING Namespace dependencies not required: survival Why? (If I remove 'importFrom' I

When I try 'install-packages' (from menu) in R-1.7.0 (Windows), I get > install.packages(choose.files('',filters=Filters[c('zip','All'),]), .libPaths()[1], CRAN = NULL) Error in file(file, "r") : unable to open connection In addition: Warning message: cannot open file `eha060/DESCRIPTION' > but it works with R-1.6.2. It also works if I

A new version (0.96.3) of the package 'eha' is now on CRAN. Apart from some bug fixes, there are some news. The most noteworthy are: (i) 'plot.Surv' has a few new options, (ii) 'weibreg' now can fit null models (i.e., only scale and shape) and covariates are no longer automatically centered (although this is still the default), (iii) there is a new function,

When reading an spss file, I get the following message: Warning message: ../totmorH.sav: Unrecognized record type 7, subtype 13 encountered in system file. What does it mean? Do I have to worry? The result looks as expected. I do not have (easy) access to SPSS, so I cannot check exactly that I got what I should have. Thanks, G??ran And, I'm using R-2.0.1 on Debian testing (from source).

I will be giving a course in survival analysis using R (of course!) for people who know nothing about the subject (including R), but know basic statistics. I'm looking for a suitable course book. Therneau & Grambsch (2000) is an excellent book, but too much for this course. I need somthing more elementary. I have a vague memory saying that such books exist, but I cannot find any for the

Is it possible to let a function be known under two names without having two identical copies of the function body? --- G?ran Brostr?m tel: +46 90 786 5223 Department of Statistics fax: +46 90 786 6614 Ume? University http://www.stat.umu.se/egna/gb/ SE-90187 Ume?, Sweden e-mail: gb at stat.umu.se

Guido, > thanks for signaling this. The --mno-cygwin was a my typos error > but I have to correct the readme.packages (--no-cygwin is a survival > from the about two year ago version of cygwin). The flag is set correctly in MkRules, though. > Anyway, note that in readme.packages we suggest to edit 'MkRules' > not MakeDll and to set at the beginning of that file

Dear R-helpers, In a data set I got from a medical doctor there are six treatment groups and (about) 5 bivariate responses in each group. Using 'manova', it is easy to see significant differences in treatment effects, but the doctor is more interested in the correlation between the two responses (within groups). I'm willing to assume a common value over groups, and one way of

I have two versions of R on my Debian testing: R-1.9.0 (precompiled) and R-1.9.1-alpha of 2004-06-07 (built on my system). In the latter, I get > system("emacs sim.R &") > emacs: standard input is not a tty while with R-1.9.0 it works as expected. I have absolutely no idea about what's going on, so I could use some help. Anyone? Thanks, G?ran -- G?ran Brostr?m

I have built today's R-devel without problem on a fresh installation of debian testing. However, 'make check' fails. In 'tests/Examples/tools-Ex.Rout.fail': > ### * vignetteDepends > > flush(stderr()); flush(stdout()) > > ### Name: vignetteDepends > ### Title: Retrieve Dependency Information for a Vignette > ### Aliases: vignetteDepends > ### Keywords:

I have been bitten by what is clearly described on the help page for 'sample', namely sampling from a population of size one. I agree that it is convenient to have an exception if 'length(x) == 1', but my suggestion is to enforce the exception only if 'x' is numeric. In any case, if x is not numeric and of length 1, all you get is an error message. I would like

I am calling 'vmmin' several times from a C function (which is called via .C). It works very well, except for memory consumption. The cause is that vmmin allocates memory via R_alloc, and this memory is not freed as vmmin exits. Instead all the allocated memory is freed on return of the .C call. In one application, I have 2000 functions of 500 variables each to minimize. In each call to

Reading about 'R_max_col' in "Writing R extensions", Version 2.1.0,(2005-01-03), I find: "Given the nr by nc matrix matrix in row ("FORTRAN") order, ..." Looks like a contradiction to me, since FORTRAN stores matrices columnwise. So is this a documentation bug? -- G?ran Brostr?m tel: +46 90 786 5223 Department of Statistics

I find 'How do I convert factors to numeric?' in the FAQ (7.12), but not the other way around. Trivial maybe, but > codes(factor(c(2, 10)) [1] 2 1 > codes(factor(c(2, 9))) [1] 1 2 How do I get the levels attached to the codes in numeric order? G?ran --- G?ran Brostr?m tel: +46 90 786 5223 Department of Statistics fax: +46 90 786 6614 Ume? University

I use Gnu emacs 21, Linux RH9 and R-1.7.0. I have succeeded to get the recommended indentation of 4 in C thru customization as described in 'R-exts', p.73, but I can't get it to work in R code. Can someone help, for instance by sending me the appropriate lines in '.emacs'? Thanks! --- G?ran Brostr?m tel: +46 90 786 5223 Department of Statistics

I need to order a long vector of integers with rather few unique values. This is very slow: > x <- sample(rep(c(1:10), 50000)) > system.time(ord <- order(x)) [1] 189.18 0.09 190.48 0.00 0.00 But with no ties > y <- sample(500000) > system.time(ord1 <- order(y)) [1] 1.18 0.00 1.18 0.00 0.00 it is very fast! This gave me the following idea: Since I don't care