Re: [R] orphaned packages
On Sun, 25 Jan 2004, Neil Eastep wrote: Are there many packages that are orphaned, and have a better half much like bootstrap? Not many are orphaned -- currently precisely two although one previously orphaned has acquired a new maintainer recently. I suspect a few more have non-responsive maintainers. Is there a listing, or directory of recommended packages (other than the packages downloaded with the initial R install)? It is precisely the set that come with the R distribution. The CRAN packages page lists both the priority (possibly recommended) and maintainer (possibly ORPHANED). The daily check page http://cran.r-project.org/src/contrib/checkSummary.html lists those as well as the result of the latest checks (although today's list seems missing results for the last few packages). Thanks, Neil Eastep. An R newbie --- Prof Brian Ripley [EMAIL PROTECTED] writes: Package bootstrap is orphaned, that is no longer maintained. Package boot is much more comprehensive, and a recommended package. That suggests to me that you always use boot: I do. -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] Re-Post: Combining Factors in model.matrix
On Sat, 24 Jan 2004 [EMAIL PROTECTED] wrote: I didn't get any response on this before, leading me to believe I've missed something fundamental. Can anybody guide me in the correct direction for more help on this? Thanks for your reply: You will need to explain to us why the object you list is `the design matrix': have *you* a reference for that? R is doing the conventional thing, and I at least have no idea where your example comes from. Perhaps I have used the wrong terminology? My understanding of a design matrix is that it identifies the factors are present for a given experiment. Here, I have a two factor experiment, where each factor has two levels. In the case I gave: t1 t2 1 1 0 2 1 1 3 0 0 4 0 1 I had expected this to represent four distinct experiments where factor one is present in the first two and absent in the second two. You seem to have coded variables t1 and t2 the opposite ways (the reference level is 2 for t1 and 1 for t2), and your model has the fit at levels t1=2,t1=1 constrained to pass through the origin. I don't think R has a simple syntax for that (although you can fake anything), and I find it hard to believe that is actually what you want. That wasn't my intention, I want to retain the intercept term and not constrain it to pass through the origin. Paul Paul = I want to be able to create a design matrix with two factors. For instance, if I have: t1 - factor(c(1,1,2,2)); t2 - factor(c(1,2,1,2)); design - model.matrix(~ -1 + (t1+t2)); design; t11 t12 t22 1 1 0 0 2 1 0 1 3 0 1 0 4 0 1 1 But the design matrix I want is: t1 t2 1 1 0 2 1 1 3 0 0 4 0 1 Actually, in general I'm struggling with the syntax for formulating a design matrix I can write down on paper. Is there a reference for this beyond the R documentation? Chapter 6 of MASS has the most complete exposition (by Bill Venables) that I know of, and the White Book (Chambers Hastie, 1992) goes well beyind the R documentation (which uses it as the reference). -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] D(COM) with Excel
On Mon, 26 Jan 2004, Joel Pitt wrote: Hi there, I'm currently trying to use R in an automated macro with Excel, and to this effect I've been using the D(COM) server. However I've been having alot of problems with it, because it seems to be limited to only recieving and sending arrays. I've been struggling trying to find a way to receive model summaries from R to put in Excel. I also seem to have some strange errors coming up... I've tried doing everything I want directly in R by hand, and it has been fine, but excel seems to complicate everything :( Any people have suggestions? There are two (D)COM servers for R, on by Thomas Baier that has its own mailing list, and one by Duncan Temple Lang (at www.omegahat.org and in the Omegahat area on CRAN). It is impossible for us to help with a message as vague as `some strange errors coming up', so I suggest you read http://www.chiark.greenend.org.uk/~sgtatham/bugs.html and then post on the appropriate mailing list for whichever server you are using. [The Baier-version list is listed at the bottom of its readme.txt file.] -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] Re-Post: Combining Factors in model.matrix
On Mon, 26 Jan 2004, Paul Boutros wrote: On Sat, 24 Jan 2004 [EMAIL PROTECTED] wrote: I didn't get any response on this before, leading me to believe I've missed something fundamental. Can anybody guide me in the correct direction for more help on this? Thanks for your reply: You will need to explain to us why the object you list is `the design matrix': have *you* a reference for that? R is doing the conventional thing, and I at least have no idea where your example comes from. Perhaps I have used the wrong terminology? My understanding of a design matrix is that it identifies the factors are present for a given experiment. The design matrix is X in the regression usually represented by y = Xb + e and is called a model matrix in S/R. Here, I have a two factor experiment, where each factor has two levels. In the case I gave: t1 t2 1 1 0 2 1 1 3 0 0 4 0 1 I had expected this to represent four distinct experiments where factor one is present in the first two and absent in the second two. You can't have factors that are present/absent. (You can have levels of treatments which are present/absent.) We understand the rows to represent the individuals runs of a single experiment, but what do the columns represent? You seem to have coded variables t1 and t2 the opposite ways (the reference level is 2 for t1 and 1 for t2), and your model has the fit at levels t1=2,t1=1 constrained to pass through the origin. I don't think R has a simple syntax for that (although you can fake anything), and I find it hard to believe that is actually what you want. That wasn't my intention, I want to retain the intercept term and not constrain it to pass through the origin. So why did you use ~ -1 + (t1+t2) ? That explicitly removes the intercept. Paul Paul = I want to be able to create a design matrix with two factors. For instance, if I have: t1 - factor(c(1,1,2,2)); t2 - factor(c(1,2,1,2)); design - model.matrix(~ -1 + (t1+t2)); design; t11 t12 t22 1 1 0 0 2 1 0 1 3 0 1 0 4 0 1 1 But the design matrix I want is: t1 t2 1 1 0 2 1 1 3 0 0 4 0 1 Actually, in general I'm struggling with the syntax for formulating a design matrix I can write down on paper. Is there a reference for this beyond the R documentation? Chapter 6 of MASS has the most complete exposition (by Bill Venables) that I know of, and the White Book (Chambers Hastie, 1992) goes well beyind the R documentation (which uses it as the reference). -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] Re-Post: Combining Factors in model.matrix
You will need to explain to us why the object you list is `the design matrix': have *you* a reference for that? R is doing the conventional thing, and I at least have no idea where your example comes from. Perhaps I have used the wrong terminology? My understanding of a design matrix is that it identifies the factors are present for a given experiment. The design matrix is X in the regression usually represented by y = Xb + e and is called a model matrix in S/R. Right, that's how I understood it. Here, I have a two factor experiment, where each factor has two levels. In the case I gave: t1 t2 1 1 0 2 1 1 3 0 0 4 0 1 I had expected this to represent four distinct experiments where factor one is present in the first two and absent in the second two. You can't have factors that are present/absent. (You can have levels of treatments which are present/absent.) We understand the rows to represent the individuals runs of a single experiment, but what do the columns represent? Yes, I mis-spoke. I thought the columns represent individual factors, with a 0 = level 1 for this factor 1 = level 2 for this factor Hence the encoding I gave above would indicate that factor 1 is at level 1 for the first pair of experiments, but at level 0 for the second pair. You seem to have coded variables t1 and t2 the opposite ways (the reference level is 2 for t1 and 1 for t2), and your model has the fit at levels t1=2,t1=1 constrained to pass through the origin. I don't think R has a simple syntax for that (although you can fake anything), and I find it hard to believe that is actually what you want. That wasn't my intention, I want to retain the intercept term and not constrain it to pass through the origin. So why did you use ~ -1 + (t1+t2) ? That explicitly removes the intercept. Ahh, I had misunderstood the -1 as explicitly specifying an intercept. So now: t1 - factor(c(1,1,2,2)); t2 - factor(c(1,2,1,2)); design - model.matrix(~ t1+t2); design; (Intercept) t12 t22 1 1 0 0 2 1 0 1 3 1 1 0 4 1 1 1 Which is what I had been looking for! Thank you for your patient help, Paul __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Error Loading r
On Sun, 25 Jan 2004, Jerome Swartz wrote: Hi there, This is the first time I have encountered such an error. Error loading r. I did forward the call to our desktop services, but I am just a bit curious on what the actual problem could be and how I could have resolved myself. OS is unix version 4. Not sure of the version of R... As its my first encounter. It's hard to guess without more information. I'd suggest either to ask your IT staff or to try (re)installing a recent version of R. Uwe Ligges Kind regards Jerome Swartz Please note: This e-mail and its contents are subject to a disclaimer which can be viewed at http://www.woolworths.co.za/disclaimer. Should you be unable to access the link please e-mail [EMAIL PROTECTED] and a copy of the disclaimer will be e-mailed to you. [[alternative HTML version deleted]] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] how to keep functions while remove all other commands
Hi On 24 Jan 2004 at 21:47, Yong Wang wrote: Dear all: a quick question: I am used to apply rm(list=()) regularly to remove all old codes in preventing them creeping in current analysis.however, with that application, functions I wrote are also removed. please let me know how to keep the thing you want while remove those you don't. I usually work in some separate home directory (e.g. D:\data\stat\standard) and I store my functions I want to keep in another directory (D:\prog\R\..\library\fun), which can be attached as any other library in the same way like other libraries can. You just have to use some files (I think at least INDEX and DESCRIPTION). More in R-exts manual. Cheers Petr thank you best yong __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Petr Pikal [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] warning associated with Logistic Regression
On Sunday, Jan 25, 2004, at 18:06 Europe/London, (Ted Harding) wrote: On 25-Jan-04 Guillem Chust wrote: Hi All, When I tried to do logistic regression (with high maximum number of iterations) I got the following warning message Warning message: fitted probabilities numerically 0 or 1 occurred in: (if (is.empty.model(mt)) glm.fit.null else glm.fit)(x = X, y = Y, As I checked from the Archive R-Help mails, it seems that this happens when the dataset exhibits complete separation. This is so. Indeed, there is a sense in which you are experiencing unusually good fortune, since for values of your predictors in one region you are perfectly predicting the 0s in your reponse, and for values in another region your a perfectly predicting the 1s. What better could you hope for? However, you would respond that this is not realistic: your variables are not (in real life) such that P(Y=1|X=x) is ever exactly 1 or exactly 0, so this perfect prediction is not realistic. In that case, you are somewhat stuck. The plain fact is that your data (in particular the way the values of the X variables are distributed) are not adequate to tell you what is happening. There may be manipulative tricks (like penalised regression) which would inhibit the logistic regression from going all the way to a perfect fit; but, then, how would you know how far to let it go (because it will certainly go as far in that direction as you allow it to). The key parameter in this situation the dispersion parameter (sigma in the usual notation). When you get perfect fit in a completely separated situation, this corresponds to sigma=0. If you don't like this, then there must be reasons why you want sigma0 and this may imply that you have reasons for wanting sigma to be at least s0 (say), or, if you are prepared to be Bayesian about it, you may be satisfied that there is a prior distribution for sigma which would not allow sigma=0, and would attach high probability to a range of sigma values which you condisder to be realistic. Unless you have a fairly firm idea of what sort of values sigma is likely to havem then you are indeed stuck because you have no reason to prefer one positive value of sigma to a different positive value of sigma. In that case you cannot really object if the logistic regression tries to make it as small as possible! This seems arguable. Accepting that we are talking about point estimation (the desirability of which is of course open to question!!), then old-fashioned criteria like bias, variance and mean squared error can be used as a guide. For example, we might desire to use an estimation method for which the MSE of the estimated logistic regression coefficients (suitably standardized) is as small as possible; or some other such thing. The simplest case is estimation of log(pi/(1-pi)) given an observation r from binomial(n,pi). Suppose we find that r=n -- what then can we say about pi? Clearly not much if n is small, rather more if n is large. Better in terms of MSE than the MLE (whose MSE is infinite) is to use log(p/(1-p)), with p = (r+0.5)/(n+1). See for example Cox Snell's book on binary data. This corresponds to penalizing the likelihood by the Jeffreys prior, a penalty function which has good frequentist properties also in the more general logistic regression context. References given in the brlr package give the theory and some empirical evidence. The logistf package, also on CRAN, is another implementation. I do not mean to imply that the Jeffreys-prior penalty will be the right thing for all applications -- it will not. (eg if you really do have prior information, it would be better to use it.) In general I agree wholeheartedly that it is best to get more/better data! In the absence of such reasons, (cut) All good wishes, David __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Response Surface
Hallo On 23 Jan 2004 at 9:08, Li Xiaolei wrote: Hi, Is there any existing way in R of doing response surface analyses and plotting the response surface, like what minitab can do? Well, I do not know about any package doing response surface design, but if you already made experiments according to design you elaborated by yourself, function contour maybe with help of interp from akima package can do the plotting. Cheers Petr Thanks Xiaolei __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Petr Pikal [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] D(COM) with Excel
Hi Joel, You may want to take a look at the RDCOM implementation at http://www.omegahat.org/RDCOMServer. We've had very good experience with it. -- David Joel Pitt wrote: Hi there, I'm currently trying to use R in an automated macro with Excel, and to this effect I've been using the D(COM) server. However I've been having alot of problems with it, because it seems to be limited to only recieving and sending arrays. I've been struggling trying to find a way to receive model summaries from R to put in Excel. I also seem to have some strange errors coming up... I've tried doing everything I want directly in R by hand, and it has been fine, but excel seems to complicate everything :( Any people have suggestions? Thanks, joel __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] correlation/random effects when using nlme
Hi, I'm trying to use nlme to program a model where, as well as a number of fixed effects, I have a random effect for each subject in my model. I would also like to include a correlation statement in my model, where the grouping factor for the correlation is a subset of the observations on a particular subject. However, when I try and program this, R will not let me use a different grouping structure for the random effects and the correlation structure. I can't seem to solve this problem myself and so I would be grateful for any help! Thanks, John Marioni University of Cambridge __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Fitting compartmental model with nls and lsoda?
DivineSAAM == DivineSAAM [EMAIL PROTECTED] writes: Dear Colleagues, Our group is also working on implementing the use of R for pharmacokinetic compartmental analysis. Perhaps I have missed something, but fit - nls(noisy ~ lsoda(xstart, time, one.compartment.model, c(K1=0.5, k2=0.5)), +data=C1.lsoda, +start=list(K1=0.3, k2=0.7), +trace=T +) Error in eval(as.name(varName), data) : Object C1.lsoda not found What part of the e-mail did I miss? I would like to get this problem up an running. Oscar, There were several problems with the code in my previous posting. I'll append an example that does work. While this example often works, there are cases when nls fails by reducing the step factor below minFactor. It is even less stable for fitting less trivial examples with real data sets. I'll be very interested to keep in touch as we make progress on this problem. Regards, Mike Here's my (more) correct example: require(odesolve) ## Simple one compartment blood flow model: one.compartment.model - function(t, x, parms) { C1 - x[1] # compartment with(as.list(parms),{ input - approx(signal$time, signal$input, t)$y dC1 - K1 * input - k2 * C1 list(c(dC1)) }) } ## vector of timesteps time - seq(0, 100, 1) ## external signal with rectangle impulse signal - as.data.frame(list(time=time, input=rep(0,length(time signal$input[signal$time = 10 signal$time =40] - 0.2 ## Parameters for steady state conditions parms - c(K1=0.5, k2=0.5) ## Start values for steady state xstart - c(C1=0) ## calculate C1 with lsoda: C1.lsoda - as.data.frame(lsoda(xstart, time, one.compartment.model, parms)) ## Add some noise to the output curve C1.lsoda$noisy - C1.lsoda$C1 + rnorm(nrow(C1.lsoda), sd=0.15*C1.lsoda$C1) ## See if I can run a fit to find the parameters that I started with... require(nls) fit - nls(noisy ~ lsoda(xstart, time, one.compartment.model, c(K1=K1, k2=k2))[,2], data=C1.lsoda, start=list(K1=0.3, k2=0.7), trace=T, control=list(tol=1e-2, minFactor=1/1024/1024) ) fit.rk4 - nls(noisy ~ rk4(xstart, time, one.compartment.model, c(K1=K1, k2=k2))[,2], data=C1.lsoda, start=list(K1=0.3, k2=0.7), trace=T, control=list(tol=1e-2, minFactor=1/1024/1024) ) ## Plot what I've got so far: par(mfrow=c(2,2)) plot(noisy ~ time, data=C1.lsoda, main='Input, C1, C1+noise', col='forestgreen') points(input ~ time, data=signal, type='b') points(C1 ~ time, data=C1.lsoda, type='b', pch=16) t - seq(0,100,0.1) plot(noisy ~ time, data=C1.lsoda, main='Input, C1+noise, lsoda', col='forestgreen') lines(t, predict(fit, list(time=t)), col='red',type='l') plot(noisy ~ time, data=C1.lsoda, main='Input, C1+noise, rk4', col='forestgreen') lines(t, predict(fit.rk4, list(time=t)), col='blue', type='l') plot(t, predict(fit.rk4, list(time=t))-predict(fit, list(time=t)), type='l') abline(h=0) print('Coefficients for lsoda solution:') print(coef(fit)) print(vcov(fit)) print('Coefficients for rk4 solution:') print(coef(fit.rk4)) print(vcov(fit.rk4)) -- Michael A. Miller [EMAIL PROTECTED] Imaging Sciences, Department of Radiology, IU School of Medicine __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Openin a file from spss
Hello everybody, I'm trying to open an SPSS (version 11.5) file with R (1.8.1 for windows) and I have some problems. I already used the following syntax to open other files ant it worked fine. (I loaded the foreign package). Would be very nice to understand what happened because I'm going to work a lot on this file and it is a good opportunity to learn R instead of spss and to confrontate results with my group mates. I apologize if I'm asking something that is well know to the entire list, but I browsed briefly through the past articles and I didn't find anything and I am short in time Thanks a lot. dennis. data - read.spss(C:/a.sav, use.value.labels=TRUE, max.value.labels=Inf, to.data.frame=TRUE) Error in read.spss(C:/a.sav, use.value.labels = TRUE, max.value.labels = Inf, : Error reading system-file header. In addition: Warning message: C:/a.sav: File layout code has unexpected value 50331648. Value should be 2, in big-endian or little-endian format. __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] scan() Bug
The \zzz notation is octal (just like C)! I presume you want ASCII character 10, that is LF, not 8 (BS), although using \n would be much easier to remember. On Mon, 26 Jan 2004, Greg Riddick wrote: Thanks for your suggestions on dealing with binary files, Prof Ripley I ended up using this method: PDF = file(file.pdf,a+b) PDFlines = readLines(PDF) . . . (Extract Some Information From PDFlines and create some objects to add back to the PDF file) . . . writeLines(newobjects, PDF, sep = \12) close(PDF) So I opened the file as binary in read/append mode. Works fine now...though I have noticed that the sep character that actually gets written to the file is -2 the value specified. So I wanted \10 and needed to specify \12 to get it. Am I doing something wrong here? I'm working on an R package to add annotations(hyperlinks, popups etc.) to PDF files that I should release in about 2 weeks. Should be useful especially to the bioinformatics people who use R. Incidentally, the uncompressed PDF files that I have seen R produce are actually just plain text files---human-readable ascii characters delimited by CR or CR/LF. They are binary only in the sense that a cross-reference table at the end of the file records byte offsets of individual objects in the file. So insertions and deletions cannot be made without updating the cross-reference table. - Original Message - From: Prof Brian Ripley [EMAIL PROTECTED] To: Greg Riddick [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Thursday, January 22, 2004 4:52 PM Subject: Re: [R] scan() Bug? On Thu, 22 Jan 2004, Greg Riddick wrote: I'm reading a file into a list by: PDF = scan(file,what=character,sep=\10) \10 is the newline character in this file, also tried \n originally On lines that are ended by \13\10, both are dropped from the list entry I want scan to keep the \13 in the list entry. Is this a bug or just a strange feature? Not a strange feature, but the documented behaviour (and useful, too). You have opened the file in text mode. If you want to keep CRs, open and read in binary mode. -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] scan() Bug?
Greg Riddick [EMAIL PROTECTED] writes: Works fine now...though I have noticed that the sep character that actually gets written to the file is -2 the value specified. So I wanted \10 and needed to specify \12 to get it. Am I doing something wrong here? Just overlooking that such codes are specified in octal notation, I think. -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] warning associated with Logistic Regression
Hi All: I am really fascinated by the content and the depth of discussion of this thread. This really exemplifies what I have come to love and enjoy about the R user group - that it is not JUST an answering service for getting help on programming issues, but also a forum for some critical and deep thinking on fundamental statistical issues. Kudos to the group! Best, Ravi. - Original Message - From: David Firth [EMAIL PROTECTED] Date: Monday, January 26, 2004 5:28 am Subject: Re: [R] warning associated with Logistic Regression On Sunday, Jan 25, 2004, at 18:06 Europe/London, (Ted Harding) wrote: On 25-Jan-04 Guillem Chust wrote: Hi All, When I tried to do logistic regression (with high maximum number of iterations) I got the following warning message Warning message: fitted probabilities numerically 0 or 1 occurred in: (if (is.empty.model(mt)) glm.fit.null else glm.fit)(x = X, y = Y, As I checked from the Archive R-Help mails, it seems that this happens when the dataset exhibits complete separation. This is so. Indeed, there is a sense in which you are experiencing unusually good fortune, since for values of your predictors in one region you are perfectly predicting the 0s in your reponse, and for values in another region your a perfectly predicting the 1s. What better could you hope for? However, you would respond that this is not realistic: your variables are not (in real life) such that P(Y=1|X=x) is ever exactly 1 or exactly 0, so this perfect prediction is not realistic. In that case, you are somewhat stuck. The plain fact is that your data (in particular the way the values of the X variables are distributed) are not adequate to tell you what is happening. There may be manipulative tricks (like penalised regression) which would inhibit the logistic regression from going all the way to a perfect fit; but, then, how would you know how far to let it go (because it will certainly go as far in that direction as you allow it to). The key parameter in this situation the dispersion parameter (sigma in the usual notation). When you get perfect fit in a completely separated situation, this corresponds to sigma=0. If you don't like this, then there must be reasons why you want sigma0 and this may imply that you have reasons for wanting sigma to be at least s0 (say), or, if you are prepared to be Bayesian about it, you may be satisfied that there is a prior distribution for sigma which would not allow sigma=0, and would attach high probability to a range of sigma values which you condisder to be realistic. Unless you have a fairly firm idea of what sort of values sigma is likely to havem then you are indeed stuck because you have no reason to prefer one positive value of sigma to a different positive value of sigma. In that case you cannot really object if the logistic regression tries to make it as small as possible! This seems arguable. Accepting that we are talking about point estimation (the desirability of which is of course open to question!!), then old-fashioned criteria like bias, variance and mean squared error can be used as a guide. For example, we might desire to use an estimation method for which the MSE of the estimated logistic regression coefficients (suitably standardized) is as small as possible; or some other such thing. The simplest case is estimation of log(pi/(1-pi)) given an observation r from binomial(n,pi). Suppose we find that r=n -- what then can we say about pi? Clearly not much if n is small, rather more if n is large. Better in terms of MSE than the MLE (whose MSE is infinite) is to use log(p/(1-p)), with p = (r+0.5)/(n+1). See for example Cox Snell's book on binary data. This corresponds to penalizing the likelihood by the Jeffreys prior, a penalty function which has good frequentist properties also in the more general logistic regression context. References given in the brlr package give the theory and some empirical evidence. The logistf package, also on CRAN, is another implementation. I do not mean to imply that the Jeffreys-prior penalty will be the right thing for all applications -- it will not. (eg if you really do have prior information, it would be better to use it.) In general I agree wholeheartedly that it is best to get more/better data! In the absence of such reasons, (cut) All good wishes, David __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting- guide.html __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Learning to use survey package
Being an electrical engineer lent to statistics, I'm now studying the stratified sampling techniques using the survey package and its documentation. I've found somewhat complex the example in the package also for the magnitude of the dataset used and the not-clearly-exposed purposes of the sampling, and my attempts to find something in the net through Google ended up with a flood of useless info. What I would find extremely helpful would be having a very simple example developped step by step, starting from an **easy** dataset, setting the aims of the sampling, and showing in a straightforward way the actions to be taken (of course, better, but not necessarely, using survey). Do you know if there is anything of this kind available in the Internet? Thanks for your help Vittorio __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Trouble with HTML search engine Mozilla Firebird
I'm having trouble with the HTML-based search engine, and I'd be grateful for any advice. I am using Mozilla Firebird 0.7 on Debian 3.0 with R 1.8.0. (I've also had the trouble with Mozilla Firebird 0.6 on Debian 3.0. I also use Mozilla Firebird 0.6 on Windows XP, and I have no trouble with the HTML-based search engine.) When I type help.start() in R, it brings up the proper search page. I follow the link to Search engine keywords, type some text into the search box, and click search. It wipes the text in the search box, and leaves me at the search page. The URL changes, to something like file:///tmp/Rtmp27053/.R/doc/html/search/SearchEngine.html?SEARCHTERM=MySearchTermTITLES=1KEYWORDS=1ALIASES=1 I've looked through the help on the R web site. There are plenty of messages which say to check that Java is correctly installed and that Javascript is turned on, but this does not seem to be my problem... I have installed Java 1.4.2_03. Java is installed in my browser, according to the test applet at http://www.java.com/en/download/help/testvm.jsp Javascript works in my browser, as far as I can test it. I tried modifying the search page, to put in various debugging messages, and they showed up as I expected. When I execute the search, the Mozilla Firebird javascript console comes up with an error message: Error: document.SearchEngine.search is not a function. The line of the web page which causes this error is line = line + document.SearchEngine.search(...) I tried putting in some debugging messages before this, to show me what's going on: alert(I found +document.SearchEngine) alert(I found +document.SearchEngine.search) The first shows me that javascript can find the object document.SearchEngine, and the second suggests that it can't find the method search() from within that object. This suggests there is some trouble in the interface between Java and Javascript in this browser. I looked through all the security settings, but (as far as I can see) I've enabled everything to do with Java and Javascript. Does anyone have any advice? Damon. __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Fortran source code
Hi, I am wondering if it is possible to view the Fortran source code called by R functions. In particular, I am interested in the leaps.setup function in a package called leaps, which calls Fortran functions ssleaps, initr etc. Any help would be greatly appreciated. Ka Yee Yeung Bioinformatics Scientist Dept of Microbiology University of Washington __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Fortran source code
Ka Yee Yeung [EMAIL PROTECTED] writes: Hi, I am wondering if it is possible to view the Fortran source code called by R functions. R is Open Source, and so are most of the packages developed for it. Would be a rare case if the source code was not available. In particular, I am interested in the leaps.setup function in a package called leaps, which calls Fortran functions ssleaps, initr etc. Any help would be greatly appreciated. Look inside http://cran.r-project.org/src/contrib/leaps_2.6.tar.gz -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Question about design matrix
Now, I am working with some design matrices. My problem is to set contrasts option. Now I want to use contr.sum as the option, and it works properly. However, this option sets the last element of a factor as -1. For example, if I have a factor which has 5 elements and want to use the contr.sum option, the 5th elements is always set to -1. I want set -1 to other element such as 1st or 2nd. Is it possible? Thank you in advance. Yongwan Chun [[alternative HTML version deleted]] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] D(COM) with Excel
Hi, as the one who wrote the RExcel package let me add something to the discussion. Thomas Baier's R (D)COM package tries to use native Windows data types to be as fast as possible. My RExcel package has the following philosophy: Offer R functions to people who think spreadsheet. One way of describing it is: You can have an R process living in each cell of the spreadsheet, and the output of R we want to deal with hast to be compatible with the spreadsheet structure. That is the reason for the restriction to arrays. Spreadsheets (NOT the embedded programming language) do not know about objects. Somewhat oversimplifying one might say: RExcel brings R computation to the spreadsheet, and RDCOM brings R objects to VBA. Erich Neuwirth David James wrote: Hi Joel, You may want to take a look at the RDCOM implementation at http://www.omegahat.org/RDCOMServer. We've had very good experience with it. -- David Joel Pitt wrote: Hi there, I'm currently trying to use R in an automated macro with Excel, and to this effect I've been using the D(COM) server. However I've been having alot of problems with it, because it seems to be limited to only recieving and sending arrays. I've been struggling trying to find a way to receive model summaries from R to put in Excel. I also seem to have some strange errors coming up... I've tried doing everything I want directly in R by hand, and it has been fine, but excel seems to complicate everything :( Any people have suggestions? Thanks, joel __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Erich Neuwirth, Computer Supported Didactics Working Group Visit our SunSITE at http://sunsite.univie.ac.at Phone: +43-1-4277-38624 Fax: +43-1-4277-9386 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] conditional assignment
Hi all I want to conditionally operate on certain elements of a matrix, let me explain it with a simple vector example z- c(1, 2, 3) zz - c(0,0,0) null - (z 2) ( zz - z) zz [1] 1 2 3 why zz is not (0, 0, 3) ? the null - assignment is to keep the console silent in the other hand, it curious that null has reasonable values null [1] FALSE FALSE TRUE Thanks in advance Ulisses Debian GNU/Linux: a dream come true - Computers are useless. They can only give answers.Pablo Picasso Humans are slow, innaccurate, and brilliant. Computers are fast, acurrate, and dumb. Together they are unbeatable ---Visita http://www.valux.org/ para saber acerca de la--- ---AsociaciĆ³n Valenciana de Usuarios de Linux --- __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Fortran source code
They are part of the package leaps_2.6.tar.gz on CRAN. Just look at the source package and not (I assume) a binary installation. On Mon, 26 Jan 2004, Ka Yee Yeung wrote: I am wondering if it is possible to view the Fortran source code called by R functions. In particular, I am interested in the leaps.setup function in a package called leaps, which calls Fortran functions ssleaps, initr etc. -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] conditional assignment
On Mon, 26 Jan 2004 [EMAIL PROTECTED] wrote: Hi all I want to conditionally operate on certain elements of a matrix, let me explain it with a simple vector example z- c(1, 2, 3) zz - c(0,0,0) null - (z 2) ( zz - z) zz [1] 1 2 3 why zz is not (0, 0, 3) ? Break it down into bits: z- c(1, 2, 3) (z 2) [1] FALSE FALSE TRUE ( zz - z) [1] 1 2 3 (z 2) ( zz - z) [1] FALSE FALSE TRUE TRUE ( zz - z) [1] TRUE TRUE TRUE rep(TRUE,4) ( zz - z) [1] TRUE TRUE TRUE TRUE Warning message: longer object length is not a multiple of shorter object length in: rep(TRUE, 4) (zz - z) The first part is a logical vector, the second is the result of assigning z to zz, of them isn't terribly meaningful? Try: zz - ifelse(z 2, z, 0) zz [1] 0 0 3 if that's what you want. the null - assignment is to keep the console silent in the other hand, it curious that null has reasonable values null [1] FALSE FALSE TRUE Thanks in advance Ulisses Debian GNU/Linux: a dream come true - Computers are useless. They can only give answers.Pablo Picasso Humans are slow, innaccurate, and brilliant. Computers are fast, acurrate, and dumb. Together they are unbeatable --- Visita http://www.valux.org/ para saber acerca de la--- --- AsociaciĆ³n Valenciana de Usuarios de Linux --- __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Roger Bivand Econonic Geography Section, Department of Economics, Norwegian School of Economics and Business Administration, Breiviksveien 40, N-5045 Bergen, Norway, voice: +47-55959355, fax: +47-55959393; [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] conditional assignment
From: Simon Cullen On Mon, 26 Jan 2004 20:15:51 +0100, [EMAIL PROTECTED] wrote: I want to conditionally operate on certain elements of a matrix, let me explain it with a simple vector example z- c(1, 2, 3) zz - c(0,0,0) null - (z 2) ( zz - z) zz [1] 1 2 3 why zz is not (0, 0, 3) ? snip in the other hand, it curious that null has reasonable values null [1] FALSE FALSE TRUE What you have done there is create a boolean vector, null, of the same length as z (and zz). For instance: (z 2) (zz - z) =(F F T) (T T T) (as assignment - presumably - returns T) Don't think so. zz - z has the value z; i.e., c(1, 2, 3). When evaluated as logicals, non-zero values are treated as true (as in C), I believe. For example: z - rep(0, 3) ifelse(zz - z, 1, 0) [1] 0 0 0 (zz - z) == TRUE [1] FALSE FALSE FALSE However, what tripped me is the fact that even though non-zero is logically `true', it's not necessarily equal to TRUE (which is numerically equal to 1): z - 0:2 (zz - z) == TRUE [1] FALSE TRUE FALSE ifelse(zz - z, 1, 0) [1] 0 1 1 if(3) TRUE else FALSE [1] TRUE Andy =(F F T). What will work is: z - c(1, 2, 3) index - z2 zz - z * index -- SC Simon Cullen Room 3030 Dept. Of Economics Trinity College Dublin Ph. (608)3477 Email [EMAIL PROTECTED] -- Notice: This e-mail message, together with any attachments,...{{dropped}} __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] conditional assignment
On Mon, 26 Jan 2004, Simon Cullen wrote: On Mon, 26 Jan 2004 20:15:51 +0100, [EMAIL PROTECTED] wrote: I want to conditionally operate on certain elements of a matrix, let me explain it with a simple vector example z- c(1, 2, 3) zz - c(0,0,0) null - (z 2) ( zz - z) zz [1] 1 2 3 why zz is not (0, 0, 3) ? snip in the other hand, it curious that null has reasonable values null [1] FALSE FALSE TRUE What you have done there is create a boolean vector, null, of the same length as z (and zz). For instance: (z 2) (zz - z) =(F F T) (T T T) (as assignment - presumably - returns T) assignment returns the new value of zz, which as it is all non-zero coerces to the logical vector T T T =(F F T). What will work is: z - c(1, 2, 3) index - z2 zz - z * index Rather better I think is zz - ifelse(z 2, z, zz) or even ind - z 2 zz[ind] - z[ind] -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] HTML help pages
Devin Johnson wrote: I am having trouble getting the HTML help pages to work. When I try the search engine I get an error on page response and nothing happens. When I try to click on the listed topics nothing happens. I am using both Mozilla and IE6 on XP and the same thing happens on each. Is there something I'm missing? I have (almost exactly) the same problem. If I start an R session, do help.start(), and then say, e.g. ?plot the help comes up in the Mozilla window as it should. But if I attempt a search using the search engine in the window provided by help.start(), nothing happens. I don't any error from clicking on a topic, just a total lack of response. I'm using Mozilla and R 1.8.1 under Linux: ===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+=== _ platform i686-pc-linux-gnu arch i686 os linux-gnu system i686, linux-gnu status major1 minor8.1 year 2003 month11 day 21 language R ===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+=== I've no idea what the solution is. This is contributed just to point out that the problem is not unique. cheers, Rolf Turner [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] HTML help pages
On Mon, 2004-01-26 at 15:58, Rolf Turner wrote: Devin Johnson wrote: I am having trouble getting the HTML help pages to work. When I try the search engine I get an error on page response and nothing happens. When I try to click on the listed topics nothing happens. I am using both Mozilla and IE6 on XP and the same thing happens on each. Is there something I'm missing? I have (almost exactly) the same problem. If I start an R session, do help.start(), and then say, e.g. ?plot the help comes up in the Mozilla window as it should. But if I attempt a search using the search engine in the window provided by help.start(), nothing happens. I don't any error from clicking on a topic, just a total lack of response. I'm using Mozilla and R 1.8.1 under Linux: ===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+=== _ platform i686-pc-linux-gnu arch i686 os linux-gnu system i686, linux-gnu status major1 minor8.1 year 2003 month11 day 21 language R ===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+===+=== I've no idea what the solution is. This is contributed just to point out that the problem is not unique. In general, one key thing to check is when you click on the link for Search Engine Keywords after the help.start() page comes up, look at the status line in Mozilla in the lower left hand corner. It should indicate something to the effect of Applet SearchEngine Started. If you do not see this message, then the Java applet has not been initialized and the queries from the search page will not work. It takes both Java and JavaScript to be functioning properly for the search applet to work. If you do not see that message, it is a good indication that there is a problem with your Java installation, which might include Java and/or browser version conflict issues. Be sure to check the browser release notes and/or Java installation notes as appropriate. It has been a while since I have used IE (or even Windows for that matter), but I am guessing that there should be a similar indication there. I am running R 1.8.1 Patched along with Firebird 0.7 and Mozilla 1.6 under Fedora Core 1 and have had no problems using the help.search() mechanism with either browser. FWIW, after the previously discussed issues with the installation of Java and Mozilla, I stopped putting the oft discussed symlink to Java in the Mozilla version specific folders (/usr/lib/mozilla-version/plugins) and now put it in ~/.mozilla/plugins. While this makes the Java installation user specific, I am the only user of my laptop. This does however remove the need to redo the symlink everytime I upgrade Mozilla and it also works for Firebird. The same goes for Flash, Helix and other plugins. HTH, Marc Schwartz __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] write.table file=file.txt help
Hi all, I have a R script that creates several input files for an analysis program. It loops through the matrix read into R and picks out submatrices and then creates a separate output file for each submatrix. The loop works great, but I am having trouble getting all the separate output files written. The line I have is: write.table(ch1d, file=C:/WINDOWS/Desktop/SNPs/haplo.txt, row.names=F, col.names=F, append=F, quote=F) Which works just fine if I just wanted to create a single file from the loop. However, I need to somehow get it to change the name of the output file (haplo.txt) each time it goes through the loop so it doesn't overwrite each time. In perl, I'd create $n=1 and increment up each loop, and call the file something like haplo.txt.$n I tried to do something like that but R doesn't recognize the variable that would be $n in perl (because it's part of the quoted name of the output file). Adding it after the ending just gave me an error, as I thought it would. I also tried to use system(copy ...) to change the name of the file in dos, but my knowledge of dos is abysmal, so I was unable to do it. Any ideas on how to go about doing this would be most appreciated! Thanks in advance, KK Nicodemus __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] write.table file=file.txt help
Look at ?paste for (j in 1:10) { write.table(j, file=paste(haplo.txt, j, sep=.), row.names=F, col.names=F, append=F, quote=F) } BTW, there have been many similar posts like this in the past. They are easily found using the search function at http://cran.r-project.org/search.html HTH, Andy __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] write.table file=file.txt help
Thanks to Andy Bunn and Patrick Connolly for their help! Kristin Nicodemus __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Function R
Hello, Please, I need a R function to optimize a function (maximizer ou minimizer) under restriction. What function do this? Thank you. Marcos __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] write.table file=file.txt help
Kristin Kay Nicodemus wrote: Hi all, I have a R script that creates several input files for an analysis program. It loops through the matrix read into R and picks out submatrices and then creates a separate output file for each submatrix. The loop works great, but I am having trouble getting all the separate output files written. The line I have is: write.table(ch1d, file=C:/WINDOWS/Desktop/SNPs/haplo.txt, row.names=F, col.names=F, append=F, quote=F) Which works just fine if I just wanted to create a single file from the loop. However, I need to somehow get it to change the name of the output file (haplo.txt) each time it goes through the loop so it doesn't overwrite each time. In perl, I'd create $n=1 and increment up each loop, and call the file something like haplo.txt.$n I tried to do something like that but R doesn't recognize the variable that would be $n in perl (because it's part of the quoted name of the output file). Adding it after the ending just gave me an error, as I thought it would. I also tried to use system(copy ...) to change the name of the file in dos, but my knowledge of dos is abysmal, so I was unable to do it. Any ideas on how to go about doing this would be most appreciated! Thanks in advance, KK Nicodemus Use paste(). for(i in 1:n) { file - paste(C:/WINDOWS/Desktop/SNPs/haplo, i, txt, sep = .) cat(Writing data to, file, \n) write.table(ch1d, file=file, row.names=F, col.names=F, append=F, quote=F) } -sundar __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] The results of your email commands
The results of your email command are provided below. Attached is your original message. - Results: Ignoring non-text/plain MIME parts - Done. ---BeginMessage--- The message contains Unicode characters and has been sent as a binary attachment. ---End Message--- __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Reply : hi
The recipient's mailbox is full. receiver: [EMAIL PROTECTED] Received: from [165.132.77.225] ([EMAIL PROTECTED]) by mh02.hanafos.com (Terrace MailWatcher) with ESMTP id 2004012713:21:46:219514.20887.5524 for [EMAIL PROTECTED]; Tue, 27 Jan 2004 13:21:46 +0900 (KST) From: [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: hi Date: Tue, 27 Jan 2004 13:21:07 +0900 MIME-Version: 1.0 Content-Type: multipart/mixed; boundary==_NextPart_000_0010_B206E143.4B87CD4D X-Priority: 3 X-MSMail-Priority: Normal X-TERRACE-SPAMMARK:(SR:6.43) __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html