Re: [R] R-devel and gcc 3.4
Prof Brian Ripley [EMAIL PROTECTED] writes: On 20 Aug 2004, Peter Dalgaard wrote: Samuelson, Frank* [EMAIL PROTECTED] writes: I encountered a similar problem with 1. Intel compilers. 2. gcc when I turn off optimizations (no -Ox) Ooops. I didn't notice #2 the first time around. If we're failing the checks with no optimization, then we surely have set the tolerances too tight. He may be, but I am not (i686, x86_64, Solaris 32- and 64-bit, Windows), all gcc 3.4.1 with -O2 and without (in most cases). Yes, I couldn't reproduce it on x86_64 either. What was the exact platform again? -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] more on apply on data frame
On Fri, Aug 20, 2004 at 11:40:16PM -0500, Laura Holt wrote: Hi R People: Several of you pointed out that using tapply on a data frame will work on the iris data frame. I'm still having a problem. The iris data frame has 150 rows, 5 variables. The first 4 are numeric, while the last is a factor, which has the Species names. I can use tapply for 1 variable at a time: tapply(iris[,1],iris[,5],mean) setosa versicolor virginica 5.006 5.936 6.588 but if I try to use this for all of the first 4, I get an error: tapply(iris[,1:4],iris[,5],mean) Error in tapply(iris[, 1:4], iris[, 5], mean) : arguments must have same length Any ideas of what I'm doing wrong, please? You are not reading the help page: Usage: tapply(X, INDEX, FUN = NULL, ..., simplify = TRUE) Arguments: X: an atomic object, typically a vector. iris[, 1:4] is a data frame of length 4; iris[, 5] is a vector of length 150. You probably need to loop over the four first columns and apply tapply(!) four times, but I'm sure there is a smarter way. Others will tell you. Göran -- Göran Broströmtel: +46 90 786 5223 Department of Statistics fax: +46 90 786 6614 Umeå University http://www.stat.umu.se/egna/gb/ SE-90187 Umeå, Sweden e-mail: [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] sweave and post rpart
Hi, have anybody positive experience how it is possible get the post output from a rpart-object in a large sweave file or is only the less nicely plot function instead of post possible??? \begin{figure}[htbp] \begin{center} fig=true,eps=T= post(dtreeEB, title=Tree,digits=getOption(digits) - 0,use.n=TRUE) @ \caption{Tree1} \end{center} \end{figure} (1) if the rpart object result contain only the root node, i get this error: Error: chunk 21 Error in if (is.leaf[lson]) left - list(left = x[lson], right = x[lson], : missing value where TRUE/FALSE needed (2) if the rpart object contains further nodes sweave finished fine, but the plot-part contains no tree and corupted the tex-document, so a tex2pdf isn't possible. Many Thanks, Christian __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] paired t-test vs pairwise t-test
John Christie [EMAIL PROTECTED] writes: Maybe I'm missing something but I don't see an easy way to help with this ad hoc. Perhaps someone (hey, maybe me) could volunteer as a filter for ad hoc help modifications. Although, I certainly do not feel qualified. I collect any errors reported in the printed versions of the manuals and will periodically submit suitable patches for them to r-bugs. If someone finds a section of the manual that needs more description they should let me know -- I will either submit a patch myself or by hiring an R expert (as sales permit). -- Brian Gough Network Theory Ltd, Publishing the R Reference Manuals --- http://www.network-theory.co.uk/R/ __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] sweave and post rpart
On Sat, 21 Aug 2004, Christian Schulz wrote: Hi, have anybody positive experience how it is possible get the post output from a rpart-object in a large sweave file or is only the less nicely plot function instead of post possible??? \begin{figure}[htbp] \begin{center} fig=true,eps=T= post(dtreeEB, title=Tree,digits=getOption(digits) - 0,use.n=TRUE) @ \caption{Tree1} \end{center} \end{figure} (1) if the rpart object result contain only the root node, i get this That's not a tree so an error is probably correct if not so unfriendly a one. Perhaps you would like to submit a patch to the *authors*, cc the maintainer. error: Error: chunk 21 Error in if (is.leaf[lson]) left - list(left = x[lson], right = x[lson], : missing value where TRUE/FALSE needed (2) if the rpart object contains further nodes sweave finished fine, but the plot-part contains no tree and corupted the tex-document, so a tex2pdf isn't possible. post.rpart() does not produce graphics output on the open device. If you trouble to read its help page, you will see the solution in the Details: section. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272860 (secr) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] error bars
Dear all, is there an easy way to create error bars for the following types of plots: a) barplots b) interaction plots Many other statistics packages (e.g. Statistica) offer very nice interaction plots with error bars, and I´d like to be able to do the same in R. Best regards Christoph __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] (no subject)
Hi! I would like to unsubscribe this mailing list, but I do not know how to do that. It is my humble request to mailing list administrator to delete my email address from the list. Thanks a lot. with regards; Santosh __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] slackware packages
Carlos Henrique Grohmann wrote: Hello all, I'm new in this list and I'd like to know if someone knows about R packages for slackware linux. I don't know slackware linux, but what about simply installing the packages using install.packages()? Uwe Ligges thanks all. __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] unsubscribing R-help (no subject)
Santosh Kumar wrote: Hi! I would like to unsubscribe this mailing list, but I do not know how to do that. It is my humble request to mailing list administrator to delete my email address from the list. Thanks a lot. with regards; Santosh Arrgh! If you were on this list before, have you never noticed the lines below all the messages: __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html The second one tells you how to unsubscribe (please use that way in order to reduce the load of work of the list administrator - thanks again to Martin!). The third one tells you not to post questions like this on the list, as well as using a sensible subject line! Uwe Ligges __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] more on apply on data frame
Laura Holt lauraholt_983 at hotmail.com writes: Hi R People: Several of you pointed out that using tapply on a data frame will work on the iris data frame. I'm still having a problem. The iris data frame has 150 rows, 5 variables. The first 4 are numeric, while the last is a factor, which has the Species names. I can use tapply for 1 variable at a time: tapply(iris[,1],iris[,5],mean) setosa versicolor virginica 5.006 5.936 6.588 but if I try to use this for all of the first 4, I get an error: tapply(iris[,1:4],iris[,5],mean) Error in tapply(iris[, 1:4], iris[, 5], mean) : arguments must have same length This is a job for aggregate: R data(iris) R aggregate(iris[,1:4], list(Species = iris[,5]), mean) Species Sepal.Length Sepal.Width Petal.Length Petal.Width 1 setosa5.006 3.4281.462 0.246 2 versicolor5.936 2.7704.260 1.326 3 virginica6.588 2.9745.552 2.026 The by command would also work using colMeans: R by(iris[,1:4], list(Species = iris[,5]), colMeans) Species: setosa Sepal.Length Sepal.Width Petal.Length Petal.Width 5.0063.4281.4620.246 Species: versicolor Sepal.Length Sepal.Width Petal.Length Petal.Width 5.9362.7704.2601.326 Species: virginica Sepal.Length Sepal.Width Petal.Length Petal.Width 6.5882.9745.5522.026 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] paired t-test vs pairwise t-test
John Christie jc at or.psychology.dal.ca writes: : : On Aug 20, 2004, at 6:16 PM, Berton Gunter wrote: : : The fact is that, while certainly desirable, it is very difficult and : time-consuming to write the sort of extensively exampled, : instructional Help : files that you desire. : : It would be great if one could easily submit a modified help file for : inclusion in the default distribution. I had mentioned this some time ago but another possibility would be to turn the whole help system into a wiki or have a link from each help page to a supplementary wiki page that users could add additional comments to. This would not require any coordination and could harness the growing R community making it easy for more people to contribute. There is already an unused wiki at: http://fawn.unibw-hamburg.de/cgi-bin/Rwiki.pl?RwikiHome but having the links might provide a more structured approach and further encouragement to use the wiki. __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] A troubled state of freedom: generalized linear models where number of parameters number of samples
Good morning, Thank you all for your help so far. I really appreciate it. The crux of my problem is that I am generating a generalized linear model with 1 dependent variable, approximately 50 training samples and 100 parameters (gene levels). Essentially, if I have 100 genes and 50 samples, this results in coefficients for the first 49 samples, and NAs for the rest, with an ultra low residual deviance (usually approx. 10^-27). This seems to have something to do with the number of degrees of freedom (since as the number of genes increases up to 49, the number of residual degrees of freedom drops to 0) What kind of methods can I use to make sense of this? I have a subsequent set of samples to work on to validate the results of this glm, so I am not sure if overfitting is really a problem. Background: this is a microarray study, where I have divided the samples in the training set into 2 groups, and generated a number of genes to differentiate between both groups. I am going to use the GLM in a subsequent regression analysis to determine survival. For this purpose, I need to generate some kind of score for each individual case using the coefficients of each gene level * gene expression level. I am not a statistician (but a clinician) - many apologies if I am not conveying myself very clearly here! Thanks. Min-Han Tan __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] relative frequencies for hist()
I have problems getting a histogram with relative frequencies on the y-axis. Here is an example data set: a - c(4.626, 4.627, 4.627, 4.628, 4.629, 4.629, 4.630, 4.631, 4.632, 4.632) d = hist(a,freq=F) d$density [1] 299. 100. 200. 100. 100. 200. The obtained densities are given by counts/(total n * bin width), with bin width being 0.001 in this case. Is there any way to get a histogram with relative frequencies irrespective of bin width, i.e. counts/total n ? It doesn't seem to work for truehist() either. Thanks. Steffen __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] relative frequencies for hist()
Hi! d$counts/length(a) And of course you can if it is what you want. d$density-d$counts/length(a) plot(d,freq=F) Sincerely Eryk *** REPLY SEPARATOR *** On 8/21/2004 at 7:10 PM Steffen Katzner wrote: I have problems getting a histogram with relative frequencies on the y-axis. Here is an example data set: a - c(4.626, 4.627, 4.627, 4.628, 4.629, 4.629, 4.630, 4.631, 4.632, 4.632) d = hist(a,freq=F) d$density [1] 299. 100. 200. 100. 100. 200. The obtained densities are given by counts/(total n * bin width), with bin width being 0.001 in this case. Is there any way to get a histogram with relative frequencies irrespective of bin width, i.e. counts/total n ? It doesn't seem to work for truehist() either. Thanks. Steffen __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Dipl. bio-chem. Eryk Witold Wolski@MPI-Moleculare Genetic Ihnestrasse 63-73 14195 Berlin 'v' tel: 0049-30-83875219 / \ mail: [EMAIL PROTECTED]---W-Whttp://www.molgen.mpg.de/~wolski __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Convergence code in nlm function
Dear R users, I am using the nlm function for minimization of the very non-linear function of four parameters. I am running 100 simulations and almost always I get the convergence code =2 (Successive iterates within tolerance. Current iterate is probably solution.) [about 75 times of 100]. Frequently, 3 of 4 relative gradients are close to zero and the fourth is huge but there are also cases where all four gradients are far from zero. Interestingly, that the estimates are close enough to the true values in most of cases. I tried to change the optional parameters in the nlm call but this does not help to achieve code=1. What may be a reason for such situation? Could I think that the obtained estimates are what I am looking for? I am using R 1.9.1: nlm(N.Loglik,initial.values,fscale=1,print.level = 1, gradtol = 1e-6,steptol = 1e-6,iterlim = 500,y=y) Much thanks, Victoria. [[alternative HTML version deleted]] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] loadhistory() in .Rprofile ?
dear wizards: my .Rprofile has just one command for testing, loadhistory(~/.Rhistory) but this gives me an error on R startup: Error: couldn't find function loadhistory Invoking loadhistory() as the first interactive command works fine; incidentally, I believe loadhistory() in the .Rprofile worked in earlier or other platform R releases, too. Is the .Rprofile now loaded before loadhistory()? if so, this seems to make it impossible to set up one big history file across directories (save history prior to exit, then reload automatically on entry). if not, I probably screwed something up---again. regards, /iaw --- ivo welch PS: my earlier amd64 gentoo problem was caused by my use of f2c, which apparently is not happy on amd64 platforms. thanks to those who let me know the problem. __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] review by statistician/ Heiberger and Holland
System Info: OS: linux Mandrake 9.1 R Version 1.8.1 (2003-11-21), GNU Emacs 21.2.93.1 Browser Mozilla firefox - Colleagues Has anyone written a review of this book? Statistical Analysis and Data Display: An Intermediate Course with Examples in S-PLUS, R, and SAS. Richard M. Heiberger and Burt Holland http://www.insightful.com/news_events/webcasts/pharm04/heiberg er.asp I can't find a review on the web yet, except for publisher type reviews, e.g. http://bookwebpro.kinokuniya.co.jp/booksea.cgi?ISBN=038740270 5USID=. I'm really interested in a review by a statistician. I'm particularly interested with how it compares with Statistical Computing: An Introduction to Data Analysis using S-Plus Michael J. Crawley I've found MJ Crawleys book very useful for both myself and grad students, and got our library http://www.sardi.sa.gov.au/aquatic/index.html to buy it. Best fishes Sam -- Sam McClatchie Elena Turin 22 McKenzie St, Coromandel Valley, Adelaide, South Australia 5051 email [EMAIL PROTECTED] Telephone: (61-8) 8270 1022 cell. 0431 304 497 Research home page http://www.members.iinet.net.au/~s.mcclatchie/ /\ ...xX(? ?)Xx / \\ (((? (((? ...xX(?O?)Xx __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] loadhistory() in .Rprofile ?
find(loadhistory) [1] package:utils so see the comments at the top of the NEWS for 1.9.0. You will need to use utils::loadhistory() in .Rprofile, or a loadhook. On Sat, 21 Aug 2004 [EMAIL PROTECTED] wrote: dear wizards: my .Rprofile has just one command for testing, loadhistory(~/.Rhistory) but this gives me an error on R startup: Error: couldn't find function loadhistory Invoking loadhistory() as the first interactive command works fine; incidentally, I believe loadhistory() in the .Rprofile worked in earlier or other platform R releases, too. Is the .Rprofile now loaded before loadhistory()? if so, this seems to make it impossible to set up one big history file across directories (save history prior to exit, then reload automatically on entry). if not, I probably screwed something up---again. regards, /iaw --- ivo welch PS: my earlier amd64 gentoo problem was caused by my use of f2c, which apparently is not happy on amd64 platforms. thanks to those who let me know the problem. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html