Re: [R] on.exit() inside local()
Vadim Ogranovich vograno at evafunds.com writes: : : Hi, : : Since I routinely open files in a loop I've developed a habit of using : on.exit() to close them. Since on.exit() needs to be called within a : function I use eval() as a surrogate. For example: : : for (fileName in c(a, b)) eval({ : con - file(fileName); : on.exit(close(con)) : }) : : and con will be closed no matter what. : : However it stopped working once I wrapped the loop in local(): : local( : + for (foo in seq(2)) eval({ : + on.exit(cat(foo, \n)) : + }) : + ) : Error in cat(foo, \n) : Object foo not found : : : W/o local()it works just fine :for (foo in seq(2)) eval({ : + on.exit(cat(foo, \n)) : + }) : 1 : 2 : : The reason I wanted the local() is to keep 'foo' from interfering with : the existing environments, but somehow this breaks the thing. : At this point I am stuck. Could someone please tell what's going on? The on.exit code is executing in an environment whose parent is namespace:base and so cannot access the environment created by local. Use evalq, instead of eval, which has the effect of running the on.exit code in the environment created by local: local({ for (i in c(a, b)) evalq( on.exit(cat(i, \n)) ) }) or use an inner local, which has the effect of creating a new environment for each iteration of the loop in which the on.exit code runs: local({ for (i in c(a, b)) local( on.exit(cat(i, \n)) ) }) __ [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 import specific column(s) using read.table?
Gabor Grothendieck ggrothendieck at myway.com writes: : : F Duan f.duan at yale.edu writes: : : I have a very big tab-delim txt file with header and I only want to import : several columns into R. I checked the options for read.table and only : : Try using scan with the what=list(...) and flush=TRUE arguments. : For example, if your data looks like this: : : 1 2 3 4 : 5 6 7 8 : 9 10 11 12 : 13 14 15 16 : : then you could read columns 2 and 4 into a list with: : oops. That should be 1 and 3. :scan(myfile, what = list(0, NULL, 0), flush = TRUE) : : or read in and convert to a data frame via: : :do.call(cbind, scan(myfile, what = list(0, NULL, 0), flush = TRUE)) __ [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] date axes and formats in levelplot
Hi all (and particularly Deepayan), A while back Deepayan helped me with the query in the text below (thanks again). Specifically it was about changing the way that dates plotted on the axes of lattice plots. While this works using xyplot, everything comes apart when I use levelplot. The axis labels on the date axis are shown as the integer representation of the date (number of seconds since the origin I assume). I guess that the POSIX dates are getting coerced into numeric objects somewhere along the way and that there is no easy fix for this. I would be really grateful if there is a work around that would allow me to plot recognizable dates. Any suggestions? Cheers Toby -Original Message- From: Deepayan Sarkar [mailto:[EMAIL PROTECTED] Sent: Monday, July 05, 2004 12:16 PM To: [EMAIL PROTECTED] Cc: Patterson, Toby (Marine, Hobart) Subject: Re: [R] date Axes and formats in lattice plots On Sunday 04 July 2004 21:02, [EMAIL PROTECTED] wrote: All, I have some data of animal movements that I'm plotting using xyplot() from lattice. I want to have the date (class POSIXct object) on the Y-axis and the animals longitude on X-axis. Eg. xyplot(date ~ longitude, groups = animal, data = my.data) with data like: animal ptt year month daylondate 125 03P0014 13273 2003 7 10 150.38 2003-07-10 14:03:48 126 03P0192 20890 2003 7 10 151.13 2003-07-10 14:00:47 127 03P0197 30466 2003 7 10 150.74 2003-07-10 14:02:21 ...etc It all works fine except for the format of the dates that gets displayed. I am not sure what I need to change within the lattice frame work to get a specific date format (eg. %Y-%-m-%d). Does anyone have any tips or, even better, some example code that they could pass on? For R 1.9.0 and above, you should be able to do this with xyplot(date ~ longitude, groups = animal, data = my.data, scales = list(y = list(format = %Y-%-m-%d))) Deepayan __ [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
AW: AW: [R] built-in Sweave-like documentation in R-2.x
Is selecting and 'C-c C-r'-ing the 3 chunks separately that bad? Yes. The UI should take care of it for him. right. Others may have better suggestions. A bit more work on the chunk evaluation approach within Emacs is one; it almost does what is needed, but not quite. why almost, but not quite? ...without these almost, but not quite I would rather confirm your statemnt :) -- Valery __ [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] Using R boxplot function in Excel
Hi, I tried to create boxplot in Excel using Rcom (ver. 1.0) and it works correctly. Explain better A,B,C: are three samples? groups? do you want a boxplot using A,B,C for grouping? Maybe there are corrupted files in Rcom installation, re-install Rcom. Best Vito Hi, I have downloaded the R-Com and I was able to run Interactive Graphics Demo 2 in excel. However, I couldn't create my own boxplot. Whenever I tried to run any code, it always say Error in loading DLL, even =rput(A1,A2:A20). Any idea about what's going wrong? A detailed explaination about how to use R-Excel tool would be greatly appreciated. Thanks a lot in advance! PS: I would like you to use the following data as an example. AB C 12.5186182 7.394714354 6.58360308 11.37597453 16.66820087 3.900166247 7.059103407 9.696804606 3.738396698 13.80587153 21.95622475 5.365668029 7.933769009 9.572635842 4.195704277 14.80409653 12.39208079 6.883236109 8.974253685 12.02387754 5.842696863 7.6083609 18.08369863 4.75223318 10.01654143 10.61151753 4.940416728 10.22753966 7.59634933 5.150066626 9.638591817 17.68393592 5.427933173 12.9405328 17.35731932 5.079704705 7.758718564 14.28801913 5.319497531 9.873025445 16.89445473 5.044402668 8.023517946 16.28102329 5.637006679 7.214663381 24.19544618 5.083052782 11.82039457 5.482319845 5.26250973 8.432808752 14.50188112 7.040906111 10.41255589 8.92899781 3.335806595 14.0030136 18.31841647 3.26446583 9.75501396 18.97398026 6.075650289 11.25837687 16.9443803 5.077193363 13.51650669 22.33716661 2.850945874 = Diventare costruttori di soluzioni Visitate il portale http://www.modugno.it/ e in particolare la sezione su Palese http://www.modugno.it/archivio/cat_palese.shtml __ [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] Using R boxplot function in Excel
Hi, there is a mailing list about R Com, if you still have problems write to that list. See: http://mailman.csd.univie.ac.at/mailman/listinfo/rcom-l Best Vito Hi, I have downloaded the R-Com and I was able to run Interactive Graphics Demo 2 in excel. However, I couldn't create my own boxplot. Whenever I tried to run any code, it always say Error in loading DLL, even =rput(A1,A2:A20). Any idea about what's going wrong? A detailed explaination about how to use R-Excel tool would be greatly appreciated. Thanks a lot in advance! PS: I would like you to use the following data as an example. AB C 12.5186182 7.394714354 6.58360308 11.37597453 16.66820087 3.900166247 7.059103407 9.696804606 3.738396698 13.80587153 21.95622475 5.365668029 7.933769009 9.572635842 4.195704277 14.80409653 12.39208079 6.883236109 8.974253685 12.02387754 5.842696863 7.6083609 18.08369863 4.75223318 10.01654143 10.61151753 4.940416728 10.22753966 7.59634933 5.150066626 9.638591817 17.68393592 5.427933173 12.9405328 17.35731932 5.079704705 7.758718564 14.28801913 5.319497531 9.873025445 16.89445473 5.044402668 8.023517946 16.28102329 5.637006679 7.214663381 24.19544618 5.083052782 11.82039457 5.482319845 5.26250973 8.432808752 14.50188112 7.040906111 10.41255589 8.92899781 3.335806595 14.0030136 18.31841647 3.26446583 9.75501396 18.97398026 6.075650289 11.25837687 16.9443803 5.077193363 13.51650669 22.33716661 2.850945874 = Diventare costruttori di soluzioni Visitate il portale http://www.modugno.it/ e in particolare la sezione su Palese http://www.modugno.it/archivio/cat_palese.shtml __ [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
AW: [R] built-in Sweave-like documentation in R-2.x
hi tony, What exactly do you mean by this? 1. generation of Sweave-style docs from R programs or interaction? neither (if i correctly interpret your question). 2. tools for doing docs and analysis at the same time? Emacs Speaks Statistics has supported this with R since last century (1997 or so). as you have seen, i use emacs and even since last century :) 3. the vignettes of Bioconductor? not sure. 4. a text book in line with the above? nope. I think just smarter C-c C-r would be kind of trade-off here. hm, maybe there are some other voices here similar to mine? It would be easier to discuss the subj. -- Valery. __ [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] RSPerl on Redhat 9 (i386 box) and R-1.9.1
Hi, Sorry for the open-ended nature of the question but was *anyone* able to successfully install RSPerl(version 0.5-7) on Redhat 9, i386 box for the latest R version (1.9.1)? I tried far too many things to make it work but am unable to get it working for any case (R within Perl or Perl within R... my main interest is R within perl). If anyone is interested, I can document all my (unsuccessful) efforts that I tried to make it work. Cheers Manoj __ [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] linear constraint optim with bounds/reparametrization
On 8/9/04 4:52 PM, Thomas Lumley [EMAIL PROTECTED] wrote: On Mon, 9 Aug 2004, Kahra Hannu wrote: 1) constrOptim does not work in this case because it only fits inequality constraints, ie A%*%theta = c --- I was struggling with the same problem a few weeks ago in the portfolio optimization context. You can impose equality constraints by using inequality constraints = and = simultaneously. See the example bellow. Ick. You do not want to use constrOptim for equality constraints. constrOptim is a log-barrier interior-point method, meaning that it adds a multiple of log(A%*%theta-c) to the objective function. This is a really bad idea as a way of faking equality constraints. Use Lagrange multipliers and optim. Is there a package that does all that for me? Or is there example code that does something similar? ingmar __ [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] two-way ANOVA
R-help This is more a statistic thing than an R question. I have length measurements of organisms which I want to use for an two-way ANOVA( fixed factors) The problem is that I have different number of replicates for each combination of factors. What are the strengths and weakness of these approach when I applly aov function in R. The help file states : 'aov' is designed for balanced designs, and the results can be hard to interpret without balance. Thank you Luis Ridao Cruz Fiskirannsóknarstovan Nóatún 1 P.O. Box 3051 FR-110 Tórshavn Faroe Islands Phone: +298 353900 Phone(direct): +298 353912 Mobile: +298 580800 Fax: +298 353901 E-mail: [EMAIL PROTECTED] Web:www.frs.fo __ [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] linear constraint optim with bounds/reparametrization
If A%*%thetac, then log(c-A%*%theta) returns NA. if A%*%thetac, log(A%*%theta-c) returns NA. Only when A%*%theta==c do you get a number from log(A%*%theta-c), and that's (-Inf). However, for an equality constraint, I've had good luck by with an objective function that adds something like the following to my objective function: constraintViolationPenalty*(A%*%theta-c)^2, where constraintViolationPenalty is passed via ... in a call to optim. If I want only (A%*%theta=c), then I might write this as follows: constraintViolationPenalty*(A%*%thetac)*(A%*%theta-c)^2 This term is everywhere differentiable and is 0 when the constraint is satisfied. I may first run optim with a modest value for constraintViolationPenalty then restart it with the output of the initial run as starting values and with a larger value for constraintViolationPenalty. hope this helps. spencer graves Ingmar Visser wrote: On 8/9/04 4:52 PM, Thomas Lumley [EMAIL PROTECTED] wrote: On Mon, 9 Aug 2004, Kahra Hannu wrote: 1) constrOptim does not work in this case because it only fits inequality constraints, ie A%*%theta = c --- I was struggling with the same problem a few weeks ago in the portfolio optimization context. You can impose equality constraints by using inequality constraints = and = simultaneously. See the example bellow. Ick. You do not want to use constrOptim for equality constraints. constrOptim is a log-barrier interior-point method, meaning that it adds a multiple of log(A%*%theta-c) to the objective function. This is a really bad idea as a way of faking equality constraints. Use Lagrange multipliers and optim. Is there a package that does all that for me? Or is there example code that does something similar? ingmar __ [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] Help with Normal Range Estimation for repated measures
I would be grateful if members of the list could point me in the direction of any code (preferably in R) that will allow me to estimate 95th percentiles from a set of repeated measurements. For example, we are interested in a clinical measurement where we have 3 measures for 14 subjects and 2 measurements on 24 subjects and single measurement on 36 subjects. We want to combine these to form a Normal range by using something that takes account that some of the measures are repeats. Something non-parametric would be ideal like a weighted empirical distribution function. In other words we don't simply want to use 84 single values from the 84 subjects but use all the data (but we are aware this needs to be corrected for). Any help, however small, with this problem will be gratefully received. --- Dr. David Crabb School of Science, The Nottingham Trent University, Clifton Campus, Nottingham. NG11 8NS Tel: 0115 848 3275 Fax: 0115 848 6690 __ [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] Easy acf and pacf for irregular time series in R
R: Is there an easy way to get the acf and pacf for an irregular times series? That is, the acf and pacf with lag lengths that are in units of time, not observation number. There are several solutions available depending on the particular problem, some of them statistically cleaner than others: For example eliminate non-business days (NA's) from the series and compute the acf and pacf (e.g. with na.remove from tseries). For example interpolate to get a regular series and compute acf and pacf (e.g. with approx.irts from tseries). For example use a methodology which can treat NA's (e.g. use Kalman filtering from ts (R-1.8.1) now ??) and compute the acf and pacf from the estimated model... best Adrian Thanks, Jason Higbee Research Associate Federal Reserve Bank of St. Louis The views expressed in this email are the author's and not necessarily those of the Federal Reserve Bank of St. Louis or the Federal Reserve System [[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] R packages install problems linux - X not found (WhiteBoxEL 3)
-Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Marc Schwartz Sent: 09 August 2004 15:13 To: Dr Mike Waters Cc: R-Help Subject: RE: [R] R packages install problems linux - X not found (WhiteBoxEL 3) On Mon, 2004-08-09 at 08:13, Dr Mike Waters wrote: snip Marc, Sorry for the confusion yesterday - in my defence, it was very hot and humid here in Hampshire (31 Celsius at 15:00hrs and still 25 at 20:00hrs). What had happened was that I had done a clean install of WB Linux, including the XFree86 and other developer packages. However, the on-line updating system updated the XFree86 packages to a newer sub version. It seems that it didn't do this correctly for the XFree86 developer package, which was missing vital files. However it showed up in the rpm database as being installed (i.e. rpm -qa | grep XFree showed it thus). I downloaded another rpm for this manually and I only forced the upgrade because it was the same version as already 'installed' (as far as the rpm database was concerned). I assumed that all dependencies were sorted out through the install in the first place. OK, that helps. I still have a lingering concern that, given the facts above, there may be other integrity issues in the RPM database, if not elsewhere. From reading the WB web site FAQ's (http://www.whiteboxlinux.org/faq.html) , it appears that they are using up2date/yum for system updates. Depending upon the version in use, there have been issues especially with up2date (hangs, incomplete updates, etc.) which could result in other problems. I use yum via the console here (under FC2), though I note that a GUI version of yum has been created, including replacing the RHN/up2date system tray alert icon. A thought relative to this specifically: If there is or may be an integrity problem related to the rpm database, you should review the information here: http://www.rpm.org/hintskinks/repairdb/ which provides instructions on repairing the database. Note the important caveats regarding backups, etc. The two key steps there are to remove any residual lock files using (as root): rm -f /var/lib/rpm/__* and then rebuilding the rpm database using (also as root): rpm -vv --rebuilddb I think that there needs to be some level of comfort that this basic foundation for the system is intact and correct. I only mentioned RH9 to show that I had some familiarity with the RedHat policy of separating out the 'includes' etc into a separate developer package. Once all this had been sorted out, I was then left with a compilation error which pointed to a missing dependency or similar, which was not due to missing developer packages, but, as you and Prof Ripley correctly point out, from the R installation itself. Having grown fat and lazy on using R under the MS Windows environment, I was struggling to identify the precise nature of this remaining problem. As regards the R installation, I did this from the RH9 binary for version 1.9.1, as I did not think that the Fedora Core 2 binary would be appropriate here. Perhaps I should now compile from the source instead? I would not use the FC2 RPM, since FC2 has many underlying changes not the least of which includes the use of the 2.6 kernel series and the change from XFree86 to x.org. Both changes resulted in significant havoc during the FC2 testing phases and there was at least one issue here with R due to the change in X. According to the WB FAQs: If you cannot find a package built specifically for RHEL3 or WBEL3 you can try a package for RH9 since many of the packages in RHEL3 are the exact same packages as appeared in RH9. Thus, it would seem reasonable to use the RH9 RPM that Martyn has created. An alternative would certainly be to compile R from the source tarball. In either case, I would remove the current installation of R and after achieving a level of comfort that your RPM database is OK, reinstall R using one of the above methods. Pay close attention to any output during the installation process, noting any error or warning messages that may occur. If you go the RPM route, be sure that the MD5SUM of the RPM file matches the value that Martyn has listed on CRAN to ensure that the file has been downloaded in an intact fashion. These are my thoughts at this point. You need to get to a point where the underlying system is stable and intact, then get R to the same state before attempting to install new packages. HTH, Marc From unpacking the tarball and running ./configure in the R source directory, I obtain the fact that crti.o is needed by ld.so and was not found. This file is not present on the system. This file, along with crtn.o is usually installed by the gnu libc packages, I believe. However, I know that not all *nix
Re: AW: [R] built-in Sweave-like documentation in R-2.x
Khamenia, Valery [EMAIL PROTECTED] writes: I think just smarter C-c C-r would be kind of trade-off here. hm, maybe there are some other voices here similar to mine? It would be easier to discuss the subj. Within ESS, you've got the ess-thread-eval (similar to ess-chunk-eval), so the guts for cross-chunk evals are there, so the next part would be as you say, making C-c C-r SWeave-aware. An simpler alternative to code would be to allow one to eval-chunk-and-step, stepping through chunks, similar to C-c C-n for stepping through lines. Would this solve the basic problem? 3 x (2 or 3 keystrokes) for 3 chunks. I can't imagine an evaluation which would cross chunks but use only part of chunks (this suggests bad programming design to me), but perhaps you (or others) have an example of when this functionality would be useful? (i.e. actual regions to eval which cross code-chunk boundaries but contain 1 or 2 incomplete code-chunks?). best, -tony -- Anthony Rossini Research Associate Professor [EMAIL PROTECTED]http://www.analytics.washington.edu/ Biomedical and Health Informatics University of Washington Biostatistics, SCHARP/HVTN Fred Hutchinson Cancer Research Center UW (Tu/Th/F): 206-616-7630 FAX=206-543-3461 | Voicemail is unreliable FHCRC (M/W): 206-667-7025 FAX=206-667-4812 | use Email CONFIDENTIALITY NOTICE: This e-mail message and any attachme...{{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
[R] Re: date axes and formats in levelplot
On Tuesday 10 August 2004 01:34, [EMAIL PROTECTED] wrote: Hi all (and particularly Deepayan), A while back Deepayan helped me with the query in the text below (thanks again). Specifically it was about changing the way that dates plotted on the axes of lattice plots. While this works using xyplot, everything comes apart when I use levelplot. The axis labels on the date axis are shown as the integer representation of the date (number of seconds since the origin I assume). I guess that the POSIX dates are getting coerced into numeric objects somewhere along the way and that there is no easy fix for this. You are right. At first glance, it appears that I have been negligent in properly updating the default prepanel function for levelplot to handle DateTime objects. For now, you could use xyplot's default instead: levelplot(z ~ x * y, prepanel = lattice:::prepanel.default.xyplot) Add a 'scales = list(axs = i)' to get a better looking result (with the bordering rectangles partially clipped). Deepayan __ [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] R packages install problems linux - X not found (WhiteBoxEL 3)
On Tue, 2004-08-10 at 08:15, Dr Mike Waters wrote: snip From unpacking the tarball and running ./configure in the R source directory, I obtain the fact that crti.o is needed by ld.so and was not found. This file is not present on the system. This file, along with crtn.o is usually installed by the gnu libc packages, I believe. However, I know that not all *nix distributions include these files among their packages. From a web search, I have not been able to ascertain whether this lack of a crti.o is due to there not being one in the distribution, or to another incomplete package install. So, I did a completely fresh installation of WhiteBox, followed by R built from source, checked that it ran and then installed the R packages. Only then did I run up2date. At least crti.o and crtn.o are still there this time, along with the XFree86 includes. A bit of a cautionary tale, all in all. Thanks for all the help and support. Regards M Mike, From my FC2 system: $ rpm -qf /usr/lib/crti.o glibc-devel-2.3.3-27 $ rpm -qf /usr/lib/crtn.o glibc-devel-2.3.3-27 So, you are correct relative to the source of these two files. A follow up question might be, did you include the devel packages during your initial install? If not, that would explain the lack of these files. if you did, then it would add another data point to support the notion that your system was, to some level, compromised and a clean install was probably needed, rather than just trying to re-create the RPM database. Glad that you are up and running at this point. Given Martyn's follow up messages, it looks like there may be an issue with the RH9 RPM, so for the time being using the source tarball would be appropriate. Best regards, Marc __ [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] Approaches to using RUnit
Hi Seth, first of all note that it was a deliberate decision to leave it up to the RUnit user to load all the functions and packages to be tested because loading and sourcing is always very site-specific. RUnit just assumes that all functionality to be tested is already present in the R session. If you don't organize your code into packages but source individual R files your approach to source the code at the beginning of a test file looks the right thing to do. We mainly use packages and the code we use to test packages A and B, say, looks like this: library(A) library(B) testsuite.A - defineTestSuite(A, location_of_package_A/tests) testsuite.B - defineTestSuite(B, location_of_package_B/tests) testresult - runTestSuite(list(testsuite.A, testsuite.B)) printHTMLProtocol(testresult, location_of_testProtocol) We use the tests subdirectory of a package to store our RUnit tests even though this is not really according to R conventions. The nice thing is that this code can be executed in batch mode from a shell script. This script is executed nightly (and before starting R checks out and installs the packages from CVS). In this way, we know the test status of our code every morning. Hope this helps, Klaus -- Klaus Juenemann Software Engineer/ Biostatistician Epigenomics AGKleine Praesidentenstr. 110178 Berlin, Germany phone:+49-30-24345-393 fax:+49-30-24345-555 http://www.epigenomics.com [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
[R] Re: Enduring LME confusion or Psychologists and Mixed-Effects
Hello, Suppose I have a typical psychological experiment that is a within-subjects design with multiple crossed variables and a continuous response variable. Subjects are considered a random effect. So I could model aov1 - aov(resp~fact1*fact2+Error(subj/(fact1*fact2)) However, this only holds for orthogonal designs with equal numbers of observation and no missing values. These assumptions are easily violated so I seek refuge in fitting a mixed-effects model with the nlme library. I suppose that you have, for each subject, enough observations to compute his/her average response for each combination of factor1 and factor2, no? If this is the case, you can perform the analysis with the above formula on the data obtained by 'aggregate(resp,list(subj,fact1,fact2),mean)'. This is an analysis with only *within-subject* factors and there *cannot* be a problem of unequal number of observation when you have only within-subject factors (supposing you have at least one observations for each subject in each condition). I believe the problem with unequal number of observations only occurs when you have at least two crossed *between-subject* (group) variables. Let's imagine you have two binary group factors (A and B) yielding four subgroups of subjects, and for some reason, you do have the same number of observations in each subgroup, Then there are several ways of defining the main effects of A and B. In many cases, the most reasonable definition of the main effect of A is to take the average of A in B1 and in B2 (thus ignoring the number of observations, or weithting equally the four subgroups). To test the null hypothesis of no difference in A when all groups are equally weighted, one common approach in psychology is to pretend that the number of observation is each group is equal to the harmonic mean of the number of observations in each subgroups. The sums of square thud obtained can be compared with the error sum of square in the standard anova to form an F-test. This is called the unweighted approach. This can easily be done 'by hand' in R, but there is another approach: You get equivalent statistics as in the unweighted anova when you use so called 'type III' sums of square (I read this in Howell, 1987 'Statistical methods in psychology', and in John Fox book 'An R and S-plus companion to appied regression, p. 140). It is possible to get type III sums of square using John Fox 'car' library. library(car) contrasts(A)=contr.sum contrasts(B)=contr.sum Anova(aov(resp~A*B),type='III') You can compute the equally weighted cell means defining the effect of A with, say: with(aggregate(resp,list(a=a,b=b),mean),tapply(x,a,mean)) I have seen some people advise against using 'type III' sums of square but I do not know their rationale. The important thing, it seems to me, is to know which null hypothesis is tested in a given test. If indeed the type III sums of square test the effect on equally weighted means, they seem okay to me (when this is indeed the hypothesis I want to test). Sorry for not answering any of your questions about the use of 'lme' (I hope others will do), but I feel that 'lme' is not needed in the context of unequal cell frequencies. (I am happy to be corrected if I am wrong). It seems to me that 'lme' is useful when some assumptions of standard anova are violate (e.g. with repeated measurements when the assumption of sphericity is false), or when you have several random factors. Christophe Pallier http://www.pallier.org __ [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] persp, array and colors
Dear R-users, I'd like to plot a three-dimensional surface and at the meantime I'm using an array. I would like to have the values of my first matrix in the heights of the plot and the colors of the single facet taking into account the second matrix. I hope that the next code will help all of you to understand better my issue, Thanks in advance, Giancarlo ## creating my array m1 - runif(30,min=1, max=10) m2 - c(rep(0,12), rep(1,5), rep(0,3), rep(1,30-12-5-3)) mm - array(c(m1,m2), dim=c(6,5,2)) ## colors colo - c(red, blue) ## axis x - 1:6 y - 1:5 z - mm[,,1] z1 - mm[,,2] ## surface with heights and colors ## related to the first matrix (no good) persp(x, y, z, theta = 30, phi = 30, expand = 0.5, col = colo, ltheta = 120, ticktype = detailed, xlab = X, ylab = Y, zlab = values ) ## surface with heights and colors ## related to the second matrix (no good as well) persp(x, y, z1, theta = 30, phi = 30, expand = 0.5, col = colo, ltheta = 120, ticktype = detailed, xlab = X, ylab = Y, zlab = values ) + This mail has been sent through the MPI for Demographic Rese...{{dropped}} __ [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] Question about mle function
Dear all, I'd like to find the mle esttimates using the mle function mle(negloglik, start = list(), fixed=list(), method=...). I am using the L-BGFS-B method and I don't supply the gradient function. Is there a way to print the gradients found at the solution value? I am using R-1.9.1 on Windows and on Unix. Thank you in advance, Victoria Landsman. [[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] summary output for inverse Gaussian GLM
I'm getting inconsistent output about the link function from the summary() command when fitting an inverse Gaussian GLM: summary(glm(ig.formula, family=inverse.gaussian(link = log), + data=mydata, start=start.vals))$call glm(formula = ig.formula, family = inverse.gaussian(link = log), data = mydata, start = start.vals) summary(glm(ig.formula, family=inverse.gaussian(link = log), + data=mydata, start=start.vals))$family Family: inverse.gaussian Link function: 1/mu^2 Has anyone else run into this problem? I'm running v1.9.1 in Windows. E.J. __ [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] summary output for inverse Gaussian GLM
On Tue, 10 Aug 2004, Edward Dick wrote: I'm getting inconsistent output about the link function from the summary() command when fitting an inverse Gaussian GLM: Yes, it's a bug in the name, not in the result, fortunately. The inverse.gaussian family object always has the name of the link set to 1/m^2, regardless of the actual link function. -thomas __ [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] How to import specific column(s) using read.table?
Thanks a lot. Your way works perfect. And one more tiny question related to your codes: My data file has many columns to be omitted (suppose the first 20 ones), but I found scan(myfile, what=list(rep(NULL, 20), rep(0, 5)) doesn't work. I had to to type NULL 20 times and 0 five times in the list(...). But anyway, it works and saves a lot of memory for me. Thank you again. Frank Quoting Gabor Grothendieck [EMAIL PROTECTED]: Gabor Grothendieck ggrothendieck at myway.com writes: : : F Duan f.duan at yale.edu writes: : : I have a very big tab-delim txt file with header and I only want to import : several columns into R. I checked the options for read.table and only : : Try using scan with the what=list(...) and flush=TRUE arguments. : For example, if your data looks like this: : : 1 2 3 4 : 5 6 7 8 : 9 10 11 12 : 13 14 15 16 : : then you could read columns 2 and 4 into a list with: : oops. That should be 1 and 3. :scan(myfile, what = list(0, NULL, 0), flush = TRUE) : : or read in and convert to a data frame via: : :do.call(cbind, scan(myfile, what = list(0, NULL, 0), flush = TRUE)) __ [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://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] How to import specific column(s) using read.table?
Use as.list. Andy From: F Duan Thanks a lot. Your way works perfect. And one more tiny question related to your codes: My data file has many columns to be omitted (suppose the first 20 ones), but I found scan(myfile, what=list(rep(NULL, 20), rep(0, 5)) doesn't work. I had to to type NULL 20 times and 0 five times in the list(...). But anyway, it works and saves a lot of memory for me. Thank you again. Frank Quoting Gabor Grothendieck [EMAIL PROTECTED]: Gabor Grothendieck ggrothendieck at myway.com writes: : : F Duan f.duan at yale.edu writes: : : I have a very big tab-delim txt file with header and I only want to import : several columns into R. I checked the options for Âread.table and only : : Try using scan with the what=list(...) and flush=TRUE arguments. : For example, if your data looks like this: : : 1 2 3 4 : 5 6 7 8 : 9 10 11 12 : 13 14 15 16 : : then you could read columns 2 and 4 into a list with: : oops. That should be 1 and 3. :scan(myfile, what = list(0, NULL, 0), flush = TRUE) : : or read in and convert to a data frame via: : :do.call(cbind, scan(myfile, what = list(0, NULL, 0), flush = TRUE)) __ [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://stat.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://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Error message in function mars() in package mda
Hi, I am using function mars() in package mda to find knots in a whole bunch of predictor variables. I hope to be able to replicate all or some of the basis functions that the MARS software from Salford Systems creates. When I ran mars() on a small dataset, I was able to get the knots. However, when I tried running mars() on a larger dataset (145 predictor variables), for a different project, I get the following error message: fit1 - mars(disney2[,-146], disney2[,146]) Error in mars(disney2[, -146], disney2[, 146]) : NA/NaN/Inf in foreign function call (arg 5) In addition: Warning messages: 1: NAs introduced by coercion 2: NAs introduced by coercion Does arg 5 refer to the 5th column in my dataset? This seems to be a data problem, is this correct? Are there any other functions in R that will give me the knots for a set of predictor variables? Any help is greatly appreciated. Thanks, Jude __ [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] How to import specific column(s) using read.table?
At Tuesday 01:55 PM 8/10/2004, F Duan wrote: Thanks a lot. Your way works perfect. And one more tiny question related to your codes: My data file has many columns to be omitted (suppose the first 20 ones), but I found scan(myfile, what=list(rep(NULL, 20), rep(0, 5)) doesn't work. I had to to type NULL 20 times and 0 five times in the list(...). That's because rep(NULL, 20) returns a single NULL -- it's not obvious what else it could sensibly return. What you need to do is replicate 20 times a list containing NULL (and a list containing NULL is quite a different object to NULL). E.g.: rep(NULL, 20) NULL c(rep(list(NULL), 3), rep(list(0), 2)) [[1]]: NULL [[2]]: NULL [[3]]: NULL [[4]]: [1] 0 [[5]]: [1] 0 Tony Plate But anyway, it works and saves a lot of memory for me. Thank you again. Frank Quoting Gabor Grothendieck [EMAIL PROTECTED]: Gabor Grothendieck ggrothendieck at myway.com writes: : : F Duan f.duan at yale.edu writes: : : I have a very big tab-delim txt file with header and I only want to import : several columns into R. I checked the options for read.table and only : : Try using scan with the what=list(...) and flush=TRUE arguments. : For example, if your data looks like this: : : 1 2 3 4 : 5 6 7 8 : 9 10 11 12 : 13 14 15 16 : : then you could read columns 2 and 4 into a list with: : oops. That should be 1 and 3. :scan(myfile, what = list(0, NULL, 0), flush = TRUE) : : or read in and convert to a data frame via: : :do.call(cbind, scan(myfile, what = list(0, NULL, 0), flush = TRUE)) __ [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://stat.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://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Question about mle function
Victoria Landsman [EMAIL PROTECTED] writes: Dear all, I'd like to find the mle esttimates using the mle function mle(negloglik, start = list(), fixed=list(), method=...). I am using the L-BGFS-B method and I don't supply the gradient function. Is there a way to print the gradients found at the solution value? No. The details slot in an mle object is simply the return value from optim(), and that doesn't provide the gradient. Might be an idea to change that, though, since the gradient is obviously not zero where the box constraints are active. -- 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] Question about mle function
On 10 Aug 2004, Peter Dalgaard wrote: Victoria Landsman [EMAIL PROTECTED] writes: Dear all, I'd like to find the mle esttimates using the mle function mle(negloglik, start = list(), fixed=list(), method=...). I am using the L-BGFS-B method and I don't supply the gradient function. Is there a way to print the gradients found at the solution value? No. The details slot in an mle object is simply the return value from optim(), and that doesn't provide the gradient. Might be an idea to change that, though, since the gradient is obviously not zero where the box constraints are active. Not so easy, as optim does not return it and indeed the gradient is not known at R level (it is computed in the C code, for L-BFGS-B rather deep inside at that). It may not even have been computed at the final solution. We would need something similar to do_optimHess, or even to make use of that code as it does evaluate the gradient at the solution. -- 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
[R] hist and normal curve
Hi R People: I have a data set and I want to use the hist command and produce a histogram. Then I want to superimpose a normal curve over the histogram. Is there a simple way to do this, please? R version 1.9.1 Windows. thanks in advance, Sincerely, Laura Holt mailto: [EMAIL PROTECTED] hthttp://messenger.msn.click-url.com/go/onm00200471ave/direct/01/ __ [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] Polar decomposition of a rectangular matrix
Dear R users, Is anyone aware of an R implementation of a matrix polar decomposition? X = US. I can get it from SVD but I understand that this is inefficient when dealing with large matrices. Cheers Simon Gatehouse --- School of Biological, Earth Environmental Sciences University of New South Wales Kensington Sydney NSW ph 61 2 9385 8720 mb 61 0407 130 635 email [EMAIL PROTECTED] (Hellman Schofield Pty.Ltd. [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] tel 61 2 9858 3863) - [[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
Re: [R] hist and normal curve
Laura Holt wrote: Hi R People: I have a data set and I want to use the hist command and produce a histogram. Then I want to superimpose a normal curve over the histogram. Is there a simple way to do this, please? R version 1.9.1 Windows. thanks in advance, Sincerely, Laura Holt mailto: [EMAIL PROTECTED] This has been answered numerous times. See, for example, http://finzi.psych.upenn.edu/R/Rhelp01/archive/5029.html I found this using superimpose histogram normal density on the R Site Search page. --sundar __ [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] intersect two files
Hi all; Im working with two datasets in R, say data1 and data2. Both datasets are composed of several rows and columns (dataframe) and some of the rows are identical in both datasets. Im wondering if there is any way to remove from one set, say data1, the rows that are identical in the other set, say data2, using R? Thanks for any hint in advance 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] intersect two files
You have not given enough info. Do the data sets have the same columns? If not, you need to tell us more about how you can tell whether one row of a data frame is `identical' to some row of another. Assuming the columns are the same between the two, the basic idea is to combine all columns into a single vector for each, then check which elements of one is in the other. Something like (code untested!): id1 - do.call(paste, c(data1, sep=:) id2 - do.call(paste, c(data2, sep=:) ## Rows of data1 that are in data2: r1 - which(id1 %in% id2) ## Remove: data1.reduced - data1[-r1,] Andy From: Christian Mora Hi all; Im working with two datasets in R, say data1 and data2. Both datasets are composed of several rows and columns (dataframe) and some of the rows are identical in both datasets. Im wondering if there is any way to remove from one set, say data1, the rows that are identical in the other set, say data2, using R? Thanks for any hint in advance 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 __ [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] intersect two files
In short, merge with all=FALSE followed by removal of redundant columns might do the trick. If rownames serve as common key, use the argument by=0. See http://tolstoy.newcastle.edu.au/R/help/04/07/1250.html and many other hits on http://maths.newcastle.edu.au/~rking/R/ On Tue, 2004-08-10 at 23:44, Christian Mora wrote: Hi all; Im working with two datasets in R, say data1 and data2. Both datasets are composed of several rows and columns (dataframe) and some of the rows are identical in both datasets. Im wondering if there is any way to remove from one set, say data1, the rows that are identical in the other set, say data2, using R? Thanks for any hint in advance 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 __ [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] str and Surv objects
Dear R People: I used the Surv function to produce the following object: a - Surv(1:4,2:5,c(0,1,1,0)) a [1] (1,2+] (2,3 ] (3,4 ] (4,5+] str(a) Error in [.Surv(object, 1:ile) : subscript out of bounds Why does str(a) give an error, please? Or did I do something wrong? Thanks in advance. R Version 1.9.1 Windows Sincerely, Laura Holt mailto: [EMAIL PROTECTED] hthttp://messenger.msn.click-url.com/go/onm00200471ave/direct/01/ __ [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] str and Surv objects
Buried in str.default is a line: iv.len - round(2.5 * v.len) which seems to be the culprit. v.len is 4, which seems correct for the data given. Why multiply by 2.5? Andy From: Laura Holt Dear R People: I used the Surv function to produce the following object: a - Surv(1:4,2:5,c(0,1,1,0)) a [1] (1,2+] (2,3 ] (3,4 ] (4,5+] str(a) Error in [.Surv(object, 1:ile) : subscript out of bounds Why does str(a) give an error, please? Or did I do something wrong? Thanks in advance. R Version 1.9.1 Windows Sincerely, Laura Holt mailto: [EMAIL PROTECTED] hthttp://messenger.msn.click-url.com/go/onm00200471ave/direct/01/ __ [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 __ [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] Observation error in arima
Hi, Does anyone know how to include observation errors in the arima of R. I read the manual and tried the example codes, but did not find the solution. From the outputs of the components model, it seems to me that the default setting of the arima does not include the observational error in the fitting. Am I right on this one? Thanks in advance. Sincerely, Guiming Wang Natural Resource Ecology Lab Colorado State University Fort Collins, CO 80523 [[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
RE: [R] R packages install problems linux - X not found(WhiteBoxEL 3)
Marc, Yes - the glibc-devel package was shown to be installed using rpm -qa. It is also one of the packages upgraded by up2date from the original version supplied with the WhiteBox distribution. I concluded that there were probably more such improperly/incompletely upgraded packages and cut my losses. Everything seems to be fine second time around. I must have been unlucky. Regards Mike -Original Message- From: Marc Schwartz [mailto:[EMAIL PROTECTED] Sent: 10 August 2004 15:30 To: Dr Mike Waters Cc: R-Help Subject: RE: [R] R packages install problems linux - X not found(WhiteBoxEL 3) On Tue, 2004-08-10 at 08:15, Dr Mike Waters wrote: snip From unpacking the tarball and running ./configure in the R source directory, I obtain the fact that crti.o is needed by ld.so and was not found. This file is not present on the system. This file, along with crtn.o is usually installed by the gnu libc packages, I believe. However, I know that not all *nix distributions include these files among their packages. From a web search, I have not been able to ascertain whether this lack of a crti.o is due to there not being one in the distribution, or to another incomplete package install. So, I did a completely fresh installation of WhiteBox, followed by R built from source, checked that it ran and then installed the R packages. Only then did I run up2date. At least crti.o and crtn.o are still there this time, along with the XFree86 includes. A bit of a cautionary tale, all in all. Thanks for all the help and support. Regards M Mike, From my FC2 system: $ rpm -qf /usr/lib/crti.o glibc-devel-2.3.3-27 $ rpm -qf /usr/lib/crtn.o glibc-devel-2.3.3-27 So, you are correct relative to the source of these two files. A follow up question might be, did you include the devel packages during your initial install? If not, that would explain the lack of these files. if you did, then it would add another data point to support the notion that your system was, to some level, compromised and a clean install was probably needed, rather than just trying to re-create the RPM database. Glad that you are up and running at this point. Given Martyn's follow up messages, it looks like there may be an issue with the RH9 RPM, so for the time being using the source tarball would be appropriate. Best regards, Marc __ [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