Re: [R] KalmanLike: missing exogenous factor?
On Mon, 11 Oct 2004, Heng Sun wrote: From the help document on KalmanLike, KalmanRun, etc., I see the linear Gaussian state space model is a - T a + R e y = Z' a + eta following the book of Durbin and Koopman. In practice, it is useful to run Kalman filtering/smoothing/forecasting with exogenous factor: a - T a + L b + R e y = Z' a + M b + eta where b is some known vector (a function of time). Some other software like S-plus and Mathematica include the above exogenous factor. SsfPack by Koopman, etal. also has the factor built in the model to accommodate practical uses. So what is the rationale for R to leave off the exogenous factor? Is there a feasible way to convert the general model to the simple model in R? What is the rationale for your raising this? KalmanLike, KalmanRun, etc were written for R 1.5.0 as part of the ts package (see my article in R-news), and the ts applications (see the See Also section) do not need a so-called `exogenous factor' (which is not a `factor'). R does not pretend to have facilities for whatever subject area you mean (but do not say) by `in practice'. That's what addon packages are for (and some do touch on this area). We have no idea who [EMAIL PROTECTED]' is: it is courteous to use a signature and give your affiliation. -- 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
Re: [R] plotting directly to the printer
On Mon, 11 Oct 2004, T. Murlidharan Nair wrote: How do I send the plots directly to the printer ? On what operating system? On Windows, see the rw-FAQ Q2.9. On Linux/Unix, see ?dev.print and the print.it argument in ?postscript. Please do read the posting guide and supply enough information for us to answer your questions. -- 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] Statistical analysis of a large database
Deall all, We need to perform a statistical analysis of a large database (40,000 entries with approximately 500 fields in each entry) currently handled in Oracle. The data contains categorical variables only. At the current stage we suggest classification and clustering analysis. We are planning to perform the analysis in R and would be very grateful for any recommendations/suggestions/references regarding the packages/tools appropriate for this task. Thank you in advance for your attention, Vicky 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
Re: [R] plotting with multiple files
On Mon, 11 Oct 2004, T. Murlidharan Nair wrote: I need to generate plots from several multiple file and I am doing this I presume you are making plots from *data* in those files rather than R scripts, but you did not say. by reading it as a list using c(file1, file2.). Since I need to either print the plots or save it as an plot image while generating the plots in a loop, has anyone done this before. Yes. (I have no idea what you really want to know here, so am answering the question you actually asked. Please read the posting guide and its references and try to ask smarter questions.) Also is there a getchar() equivalent in R, I mean just to wait for an input from the console so that I can print before it plots the next plot. ?readline, par(ask=TRUE). However, whilst the console is waiting for input, *you* can print, but you cannot run R commands to print. -- 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
RE: [R] dot density maps
Dear Roger, Thank you very much for the reply. And now I have to a apologize for my late response. I was quite a while out of the office. This is exactly what I was looking for. I think it will be of great help to have this implemented in a package. Johannes -Original Message- From: Roger Bivand [mailto:[EMAIL PROTECTED] Sent: Thursday, September 30, 2004 8:42 PM To: Johannes SCHNITZLER Cc: [EMAIL PROTECTED] Subject: Re: [R] dot density maps On Wed, 22 Sep 2004, Johannes SCHNITZLER wrote: ... I'm looking for the possibility to draw points (randomly positioned or positioned according to a grid) into the polygons instead of filling the polygons with colors. I apologize for not replying earlier. I think that you can combine the polylist class from the maptools package with the csr function in the splancs package to position points randomly (the gridpts() finction can be used instead of csr() to give grid points): . . Could you indicate whether this function is any use, and if you would like it added to some suitable package? Roger Bivand __ [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 -- Roger Bivand Economic Geography Section, Department of Economics, Norwegian School of Economics and Business Administration, Breiviksveien 40, N-5045 Bergen, Norway. voice: +47 55 95 93 55; fax +47 55 95 93 93 e-mail: [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Statistical analysis of a large database
Hi, for your analysis use the package: ROracle Oracle database interface for R http://microarrays.unife.it/CRAN/src/contrib/Descriptions/ROracle.html see also: Diego Kuonen, Introduction au data mining avec R : vers la reconquête du `knowledge discovery in databases' par les statisticiens. Bulletin of the Swiss Statistical Society, 40:3-7, 2001. http://www.statoo.com/en/publications/2001.R.SSS.40/ Diego Kuonen and Reinhard Furrer, Data mining avec R dans un monde libre. Flash Informatique Spécial Été, pages 45-50, sep 2001. http://sawww.epfl.ch/SIC/SA/publications/FI01/fi-sp-1/sp-1-page45.html R Development Core Team, R Data Import/Export, versione 1.9.0, aprile 2004, pagg. 11-18 http://cran.r-project.org/doc/manuals/R-data.pdf Brian D. Ripley, Datamining: Large Databases and Methods, in Proceedings of useR! 2004 - The R User Conference, maggio 2004 http://www.ci.tuwien.ac.at/Conferences/useR-2004/Keynotes/Ripley.pdf Brian D. Ripley, Using Databases with R, R News, Gennaio 2001, pagg. 18-20 http://cran.r-project.org/doc/Rnews/Rnews_2001-1.pdf B. D. Ripley, R. M. Ripley, Applications of R Clients and Servers in Proceedings of the Distributed Statistical Computing 2001 Workshop, 2001, Vienna University of Technology. http://www.ci.tuwien.ac.at/Conferences/DSC-2001/Proceedings/Ripley.pdf Torsten Hothorn, David A. James, Brian D. Ripley, R/S Interfaces to Databases in Proceedings of the Distributed Statistical Computing 2001 Workshop, 2001,Vienna University of Technology. http://www.ci.tuwien.ac.at/Conferences/DSC-2001/Proceedings/HothornJamesRipley.pdf Luís Torgo, Data Mining with R. Learning by case studies, Maggio 2003 http://www.liacc.up.pt/~ltorgo/DataMiningWithR/ I hope I give you a little help. Best Vito You wrote: Deall all, We need to perform a statistical analysis of a large database (40,000 entries with approximately 500 fields in each entry) currently handled in Oracle. The data contains categorical variables only. At the current stage we suggest classification and clustering analysis. We are planning to perform the analysis in R and would be very grateful for any recommendations/suggestions/references regarding the packages/tools appropriate for this task. Thank you in advance for your attention, Vicky Landsman = Diventare costruttori di soluzioni The business of the statistician is to catalyze the scientific learning process. George E. P. Box 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://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Statistical analysis of a large database
I believe this is a reply to a posting -- since this is by no means the first time this has happened, please - use the Reply function of your mailer, or at least use Re: in the subject line and include the relevant part of the original posting, and - send the reply to the questioner, as well as possibly to the list. On Tue, 12 Oct 2004, Vito Ricci wrote: [...] Luís Torgo, Data Mining with R. Learning by case studies, Maggio 2003 http://www.liacc.up.pt/~ltorgo/DataMiningWithR/ Please note that that reference is not about large datasets, nor about `data mining' in the generally used sense. It has two studies, one incomplete, on linear regression (with 200 samples) and on time series. -- 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] lm#contrasts#one level in factor: bug or feature
(R.1.9.1; win2000) Since it's about the tenth time I had to write an if around this to catch the error ... Let's look at the line myfit-lm(res~groupvar,data=Data) Here res is of numeric type and groupvar is a factor. On first sight, it would be logical that if groupvar had only one (single) level we would get: Error in contrasts-(`*tmp*`, value = contr.treatment) : contrasts can be applied only to factors with 2 or more levels But then again, it's also inconsistent: Normally redundant variables on the right of the ~ (variables which are constant or linearly dependent on previous variables) don't lead to Error. They just get a NA as a coefficient estimate. A factor with a single level is of this type, it is just a constant. Obviously (to me) lm (or model.matrix.default) should not try to calculate contrasts on it and it's coefficient should be NA. Shouldn't it? Why is this not unimportant? Imagine you make a model of the above type for a data set with more than one level in groupvar. Then you select subsets and refit. It shouldn't die with error if you select a subset with only one level of groupvar. Chris. [[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] 64-bit R on Intel Xeon EM64T running fine
This is good news. As far as I know R has built for quite some time now on a number of 64 bit platforms (Linux on AMD Opteron/Athlon64, Solaris/Sparc) but I can't recall seeing a build on Intel with the 64 bit extensions. By the way, did you happen to run `make check' just for kicks? -roger Michael Seewald wrote: Dear mailing-list members, In the days of cheap RAM and microarray applications feasting on memory, 64-bit computers become more and more useful - to actually make use of memory beyond the magic 4GB border. I would like to report the success of running 64-bit R on an Intel Xeon EM64T machine under Linux. Just like on an AMD Opteron, R v2.0.0 compiles fine (and out of the box) and is happily allocating memory until RAM and swap reach their limit. Hardware: - HP xw6200 workstation - dual Intel Xeon 3.4GHz with hyper-threading enabled - 4GB RAM, 4GB swap System: either - Fedora Core 2 x86_64 bit Linux or - Red Hat Enterprise Linux Workstation 3.0 x86_64 bit R: - v2.0.0 Really, no problems at all during setup, a big thank you to the R developers making this possible! Best wishes, Michael __ [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] Diagnosing trouble with R-2.0, Fedora Core 2, and Rcmdf
Dear Paul, I'm currently out of town, limiting a bit my ability to answer these questions definitively, but I'll give it a shot (hoping that my memory isn't faulty): On Mon, 11 Oct 2004 13:08:39 -0500 Paul Johnson [EMAIL PROTECTED] wrote: Greetings, R-help! On 2 Fedora Core 2 Linux systems, i've completely erased the previous R and all packages and then installed R-2.0 and installed fresh packages. In using Rcmdr, I see some trouble and I wonder if other people see this and if it is due to the tcl/tk, or R, or Rcmdr. (If readers have not yet tried Rcmdr, I recommend it not just because of the GUI it provides, but also because Prof. Fox has created several very handy commands, like scatter3d, which make it possible to use advanced packages like rgl. Rcmdr provides several very handy wrapper commands that just work and users don't have to fuss over so many details! And, if you are new at R, you can use pull down menus and then Rcmdr displays the commands that correspond with those menu options. Very educational!) In no particular order, here are the issues I see with R-2.0 and Rcmdr 1. In the panel to select a dataset from a package (nenu: Data/Data in packages), I can type in the name of a dataset, but the GUI will not let me choose the name of a package from which to select a dataset. Nothing happens when I pick a package name. (With previous R/Rcmdr, I did not have this trouble). The problem is that the Rcmdr code doesn't quote the name of the package. I believe that I've fixed this problem, but haven't posted the fixed version either to my web site or to CRAN. 2. When using Rcmdr, some of the output from commands shows in the Rcmdr bottom window, but some is sent to the xterm in which R was started. For example, with a scatter3d() command, if I add the option model.summary=T, the model summary shows in the xterm. But some other models show results in the bottom pane of Rcmdr. And, it seems to me, there are some commands in which I've noticed some of the output going to the Rcmdr bottom buffer and some to the xterm. As I recall, I've fixed this problem as well, but perhaps not in a completely general way. Again, I haven't yet released the fixed version. 3. When I want to exit Rcmdr and choose the pull down exit from Commander and R, I get a Segmentation Fault that closes Rcmdr and R. This I haven't seen before. 4. 3d plots do not display on the screen until I click in the rgl window that displays the graph. This might be a video card/AGP driver problem, I suppose. These systems have the NVidia FX5200 cards and the NVidia proprietary X driver. If other users don't see the same on other kinds of systems, I guess that will tell me where the problem lies. I believe that I've noticed this problem on a Quantian system. 5. At random times, I get an error tcl/tk window popping up with a message about MASS and categorical variables. It says This reflects the first problem, and is produced by the way that the Rcmdr initially suppresses and then reports warnings. I'll check further when I return home. Sorry for the problems. John Error in data(package=parse(text=packageName)); 'MASS' must be character string or NULL It says that over and over, sometimes it has other package names, as in: Error in data(package=parse(text=packageName)); 'relimp' must be character string or NULL These seem to be harmless? -- Paul E. Johnson email: [EMAIL PROTECTED] Dept. of Political Sciencehttp://lark.cc.ku.edu/~pauljohn 1541 Lilac Lane, Rm 504 University of Kansas Office: (785) 864-9086 Lawrence, Kansas 66044-3177 FAX: (785) 864-5700 __ [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 John Fox Department of Sociology McMaster University Hamilton, Ontario, Canada http://socserv.mcmaster.ca/jfox/ __ [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] 64-bit R on Intel Xeon EM64T running fine
On Tue, 12 Oct 2004, Roger D. Peng wrote: This is good news. As far as I know R has built for quite some time now on a number of 64 bit platforms (Linux on AMD Opteron/Athlon64, Solaris/Sparc and Alpha and Irix and HP-UX and AIX as far as I understand. ) but I can't recall seeing a build on Intel with the 64 bit extensions. As I understand it, this is an Intel `clone' of AMD64, so the only news would be if there were any problems: in almost all cases the executable code is identical to that compiled for AMD64 and in the GNU classification it is also x86_64-unknown-linux-gnu. I would even expect a Goto BLAS to work, if not as well as on the specific Opteron it is tuned for. (Fast BLASes are an essential part of making the most of 64-bit processors on large problems.) Michael Seewald wrote: Dear mailing-list members, In the days of cheap RAM and microarray applications feasting on memory, 64-bit computers become more and more useful - to actually make use of memory beyond the magic 4GB border. I would like to report the success of running 64-bit R on an Intel Xeon EM64T machine under Linux. Just like on an AMD Opteron, R v2.0.0 compiles fine (and out of the box) and is happily allocating memory until RAM and swap reach their limit. Hardware: - HP xw6200 workstation - dual Intel Xeon 3.4GHz with hyper-threading enabled - 4GB RAM, 4GB swap System: either - Fedora Core 2 x86_64 bit Linux or - Red Hat Enterprise Linux Workstation 3.0 x86_64 bit R: - v2.0.0 Really, no problems at all during setup, a big thank you to the R developers making this possible! -- 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] tclk, tcltk
I have been having problems with these two 'libraries' since I installed 2.0.0. I have built a package with couple of functions so that I can load it at startup every time R is booted. The problem is that I have the following error every time I call the library Loading required package: tclk Error: package 'tclk' could not be loaded In addition: Warning message: There is no package called 'tclk' in: library(pkg, character.only = TRUE, logical = TRUE, lib.loc = lib.loc) However I can load tcltk library now (thanks to Dirk for helping me with this). I can load the functions I have in the library and everything works fine. I cannot find any information on this except the following thread which was not resolved. http://tolstoy.newcastle.edu.au/R/help/04/01/0410.html Any help or ideas are greatly appreciated. P.S. what is the difference between tclk and tcltk? and why calling the installed library is requiring tclk? I just made up a dummy package with a function fun-function(x) x^2 just to test and the same thing happens (Loading required package: tclk ...). This is on a Linux debian unstable with the kernel 2.4.20-bf2.4-xfs Thank you for all the help. jean __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Sweave and xtable
Dear List: I have some coded embedded within a LaTeX document where I subset a dataframe and use xtable to place it in my appendix. However, one of the tables is rather large and seems to extend beyond the page length. Is there a nice way to use Sweave to continue this table onto the next page? I could easily break this up in tex into two tables, but I imagine there is a better way of doing so. Thanks, Harold [[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] tclk, tcltk
What is package `tclk'? I think you have a typo somewhere in your code. At a guess you have Depends: tclk in a DESCRIPTION file for one of your packages. On Tue, 12 Oct 2004, Jean Eid wrote: I have been having problems with these two 'libraries' since I installed 2.0.0. I have built a package with couple of functions so that I can load it at startup every time R is booted. The problem is that I have the following error every time I call the library Loading required package: tclk Error: package 'tclk' could not be loaded In addition: Warning message: There is no package called 'tclk' in: library(pkg, character.only = TRUE, logical = TRUE, lib.loc = lib.loc) However I can load tcltk library now (thanks to Dirk for helping me with this). I can load the functions I have in the library and everything works fine. I cannot find any information on this except the following thread which was not resolved. http://tolstoy.newcastle.edu.au/R/help/04/01/0410.html It was resolved, and in any case it was about installing 64-bit Tcl/Tk on Solaris, something that was (and is) covered in the R Installation manual. Any help or ideas are greatly appreciated. P.S. what is the difference between tclk and tcltk? and why calling the installed library is requiring tclk? I just made up a dummy package with a function fun-function(x) x^2 just to test and the same thing happens (Loading required package: tclk ...). This is on a Linux debian unstable with the kernel 2.4.20-bf2.4-xfs -- 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
Re: [R] tclk, tcltk
Thanks to Brian. I was looking everywhere except the typo. I guess the previous build (1.9.1) ignored the dependece error. Thanks again, jean On Tue, 12 Oct 2004, Prof Brian Ripley wrote: What is package `tclk'? I think you have a typo somewhere in your code. At a guess you have Depends: tclk in a DESCRIPTION file for one of your packages. On Tue, 12 Oct 2004, Jean Eid wrote: I have been having problems with these two 'libraries' since I installed 2.0.0. I have built a package with couple of functions so that I can load it at startup every time R is booted. The problem is that I have the following error every time I call the library Loading required package: tclk Error: package 'tclk' could not be loaded In addition: Warning message: There is no package called 'tclk' in: library(pkg, character.only = TRUE, logical = TRUE, lib.loc = lib.loc) However I can load tcltk library now (thanks to Dirk for helping me with this). I can load the functions I have in the library and everything works fine. I cannot find any information on this except the following thread which was not resolved. http://tolstoy.newcastle.edu.au/R/help/04/01/0410.html It was resolved, and in any case it was about installing 64-bit Tcl/Tk on Solaris, something that was (and is) covered in the R Installation manual. Any help or ideas are greatly appreciated. P.S. what is the difference between tclk and tcltk? and why calling the installed library is requiring tclk? I just made up a dummy package with a function fun-function(x) x^2 just to test and the same thing happens (Loading required package: tclk ...). This is on a Linux debian unstable with the kernel 2.4.20-bf2.4-xfs -- 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
Re: [R] Sweave and xtable
Doran, Harold wrote: Dear List: I have some coded embedded within a LaTeX document where I subset a dataframe and use xtable to place it in my appendix. However, one of the tables is rather large and seems to extend beyond the page length. Is there a nice way to use Sweave to continue this table onto the next page? I could easily break this up in tex into two tables, but I imagine there is a better way of doing so. Thanks, Harold If xtable will not do that, you can use latex( ) in the Hmisc package with its longtable=TRUE argument. -- Frank E Harrell Jr Professor and Chair School of Medicine Department of Biostatistics Vanderbilt University __ [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] plot hclust - canberra dist + median linkage
Dan == Dan Bolser [EMAIL PROTECTED] on Mon, 11 Oct 2004 16:21:53 +0100 (BST) writes: Dan Gives strange results. Dan I get 'weird' dendrograms with canberra / binary distance metric and Dan median / centroid cluster methods. it doesn't depend on the metric: Both 'median' and 'centroid' methods are known to *not* guarantee ``monotone distance measures'', or equivalently to possibly lead to dendrograms with so called ``inversions''. We should add this to help page for hclust(). Probably for this reason, agnes() from the cluster package doesn't have these two methods [explicitly] -- though it now allows general parameter Lance-William formula methods which can also lead to inversions. Dan Is this just my data? evidently not. Though the problem does not appear for all data sets... Regards, Martin Maechler __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] A question in R
I started to learn the R language, but I didn't suceed to use an external file. Let say that I have an excel file called test1.xls in the directory C:/program files/R/rw2000/external_files that looks like that: name mark yair 80 yosi 70 ... In the appropriate directory I wrote this: x-read.delim(test1.xls) or this: x-read.csv(test1.xls) but I got: [1] X.. 0 rows (or 0-length row.names) way cannot I read the file? what is the appropriate command for reading an excel file? I looked at the site of R-project but I didn't find a suitable comand. __ [[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] A question in R
Yayira har wrote: I started to learn the R language, but I didn't suceed to use an external file. Let say that I have an excel file called test1.xls in the directory C:/program files/R/rw2000/external_files that looks like that: name mark yair 80 yosi 70 ... In the appropriate directory I wrote this: x-read.delim(test1.xls) or this: x-read.csv(test1.xls) but I got: [1] X.. 0 rows (or 0-length row.names) way cannot I read the file? what is the appropriate command for reading an excel file? I looked at the site of R-project but I didn't find a suitable comand. __ [[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 Hi Yayiya, R looks less than fondly on Excel files. The easiest solution for you will be to export your Excel file to a tab-delimited text format (Save - (.txt) Tab-delimited Text), and then use read.delim as you did. Does that make sense? Kevin __ [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] A question in R
Yayira har wrote: I started to learn the R language, but I didn't suceed to use an external file. Let say that I have an excel file called test1.xls in the directory C:/program files/R/rw2000/external_files that looks like that: name mark yair 80 yosi 70 ... In the appropriate directory I wrote this: x-read.delim(test1.xls) or this: x-read.csv(test1.xls) but I got: [1] X.. 0 rows (or 0-length row.names) way cannot I read the file? what is the appropriate command for reading an excel file? I looked at the site of R-project but I didn't find a suitable comand. Is the file an Excel file (probably from the extension) or can you open the file in notepad and read the text? If not, you need to save the file as csv or txt from within Excel and import the file into R as you tried above. You may also want to read the R Data Import/Export manual which is available from the r-project.org website. --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
Re: [R] A question in R
Take a look at the `R Data Import/Export' manual. Hint: if test1.xls is an Excel worksheet, it is not a tab-delimited file. On Tue, 12 Oct 2004, Yayira har wrote: I started to learn the R language, but I didn't suceed to use an external file. Let say that I have an excel file called test1.xls in the directory C:/program files/R/rw2000/external_files that looks like that: name mark yair 80 yosi 70 ... In the appropriate directory I wrote this: x-read.delim(test1.xls) or this: x-read.csv(test1.xls) but I got: [1] X.. 0 rows (or 0-length row.names) way cannot I read the file? what is the appropriate command for reading an excel file? I looked at the site of R-project but I didn't find a suitable comand. __ [[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 -- 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
Re: [R] A question in R
hi, check R-data import/export document at this link cran.r-project.org/doc/manuals/R-data.pdf let me give you an advice for further questions : it's easiest to find answer to your question searching directly with google -- -- Yves Magliulo [EMAIL PROTECTED] RD Engineer, CLIMPACT Tel. : +33 (0) 1 44 27 34 31 Fax. : +33 (0) 1 44 27 49 96 Universite Pierre et Marie Curie Boite 101 - Tour 45 - 5eme etage - Couloir 45/46 4 place Jussieu, 75252 Paris CEDEX 05, France Le mar 12/10/2004 à 17:56, Kevin Bartz a écrit : Yayira har wrote: I started to learn the R language, but I didn't suceed to use an external file. Let say that I have an excel file called test1.xls in the directory C:/program files/R/rw2000/external_files that looks like that: name mark yair 80 yosi 70 ... In the appropriate directory I wrote this: x-read.delim(test1.xls) or this: x-read.csv(test1.xls) but I got: [1] X.. 0 rows (or 0-length row.names) way cannot I read the file? what is the appropriate command for reading an excel file? I looked at the site of R-project but I didn't find a suitable comand. __ [[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 Hi Yayiya, R looks less than fondly on Excel files. The easiest solution for you will be to export your Excel file to a tab-delimited text format (Save - (.txt) Tab-delimited Text), and then use read.delim as you did. Does that make sense? Kevin __ [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] need help on GAM
Get some question about the function gam. Suppose I have a semiparametric model, Y~x1+x2+s(z1). Using gam, how could I get the estimates for the parametric part and nonparametric part respectively? And another question: we could find the coefficients for both parametric term and nonparametric term, what do these coefficients for the nonparametric term stand for, the coefficients for the base functions? Thank you! Song Song Liu School of Statistics 313 FordH 224 Church St. SE Minneapolis,MN 55455 __ [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] need help on GAM
Please read the posting guide and do give sufficient details for your questions to be answered. On Tue, 12 Oct 2004, Liu Song wrote: Get some question about the function gam. There are at least two functions gam() available for R, but none in standard R. Which package are you talking about? This is covered in the R FAQ, BTW. Suppose I have a semiparametric model, Y~x1+x2+s(z1). Using gam, how could I get the estimates for the parametric part and nonparametric part respectively? What do you mean by those terms? In the gam() in package gam, s(z1) is split into a linear and non-linear part, but all the terms *are* parametric (a smoothing spline is a parametric function of its argument). And another question: we could find the coefficients for both parametric term and nonparametric term, what do these coefficients for the nonparametric term stand for, the coefficients for the base functions? By definition, non-parametric terms do not have parameters aka coefficients. I suggest you ask for local advice on the statistical background to your questions, which you do need to understand before getting to using them in R. -- 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
Re: [R] need help on GAM
Prof Brian Ripley ripley at stats.ox.ac.uk writes: : And another question: we could find the coefficients for both : parametric term and nonparametric term, what do these coefficients : for the nonparametric term stand for, the coefficients for the base : functions? : : By definition, non-parametric terms do not have parameters aka : coefficients. This is definitely a candidate for the fortunes package. __ [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] locfit confidence intervals
hi, after m=locfit(y~x,..., family=binomial) plot(m,band=local) gives me a plot of locfit's result with a confidence interval around it. i would like to get the actual values that are being used to plot the lines in this figure. i tried using predict, but the standard error it returns when i supply the se=T argument, appears not to be the same as the CI produced by plot (it is much larger...in my one case). i have tried all the different type=.. options, but none of them matches the CI produced by plot. i only recently started using both R and local regression methods and am still feeling my way around, so even though I have gone through the FAQs and the online manuals and the archives of this group and the Loader book, I have been unable to make much progress. any help would be very appreciated !! thanks, suresh __ [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] need help on GAM
Get some question about the function gam. Suppose I have a semiparametric model, Y~x1+x2+s(z1). Using gam, how could I get the estimates for the parametric part and nonparametric part respectively? For a gam object (called, b,say) fitted using the mgcv `gam' then coef(b) will extract the coefficients for both the strictly parametric and smooth components of the model, labelled appropriately. summary(b) is also helpful. Often it is most useful to extract estimates of the smooths evaluated at a specified set of covariate values by using the gam `predict' method with type=terms. And another question: we could find the coefficients for both parametric term and nonparametric term, what do these coefficients for the nonparametric term stand for, the coefficients for the base functions? For mgcv gam objects the coefficients of the smooths are indeed the coefficients of the basis functions of the smooths. best, Simon _ Simon Wood [EMAIL PROTECTED]www.stats.gla.ac.uk/~simon/ Department of Statistics, University of Glasgow, Glasgow, G12 8QQ Direct telephone: (0)141 330 4530 Fax: (0)141 330 4814 __ [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] constrained optimization using nlm/optim?
I'm looking for an example of a simple R script that impliments a contrained nonlinear function using nlm or optim. I'm not exactly sure how to impliment the constraints within the objective function that is passed to nlm/optim. obj.func - function( p ) { x(p) - unconstrained obj function value if( constraint1 something ) { obj.func - x(p) } else { obj.func - some super huge number } } p - c(0.1,2.4, 5) nlm( obj.func, p, and a bunch of other arguments ) Any suggestions would be very helpful... Jeff. -- Jeff D. Hamann Forest Informatics, Inc. PO Box 1421 Corvallis, Oregon 97339-1421 phone 541-754-1428 fax 541-752-0288 [EMAIL PROTECTED] http://www.forestinformatics.com __ [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] locfit confidence intervals
Why not use predict(se=TRUE, band=local)? That's what plot.locfit does, via preplot.locfit. I'm afraid the only real way to answer questions like this is to read the sources. On Tue, 12 Oct 2004, Suresh Krishna wrote: hi, after m=locfit(y~x,..., family=binomial) plot(m,band=local) gives me a plot of locfit's result with a confidence interval around it. i would like to get the actual values that are being used to plot the lines in this figure. i tried using predict, but the standard error it returns when i supply the se=T argument, appears not to be the same as the CI produced by plot (it is much larger...in my one case). i have tried all the different type=.. options, but none of them matches the CI produced by plot. i only recently started using both R and local regression methods and am still feeling my way around, so even though I have gone through the FAQs and the online manuals and the archives of this group and the Loader book, I have been unable to make much progress. any help would be very appreciated !! thanks, suresh __ [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 -- 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] problem with lapack.so under Debian Sid
Dear list, I am sorry to bother you with this. I just upgraded yesterday to R 2.0.0 (using apt-get under Debian Sid), and now have problems running e.g., summary(lm(...)) the lm is calculated, but the summary statement gives me the following error: Error in La.chol2inv(x, size) : lapack routines cannot be loaded In addition: Warning message: unable to load shared library /usr/lob/R/modules/lapack.so: /usr/lib/Pentium4_SSE2_512KB/liblapack.so.3: undefined symbol: ieeeck_ I searched through the archives, but unfortunately I am just not LINUX-knowledgeable enough to understand the discussions concerning the algebra libraries. Following an older mail to this list, I did ldd /usr/lib/R/modules/lapack.so everything looks fine except probably for the first line, which reads libR.so = not found I would appreciate any hints on how to fix this problem! Thanks a lot. Christian -- Christian Fiebach, PhD Department of Psychology Helen Wills Institute of Neuroscience 132 Barker Hall, MC# 3190 University of California, Berkeley CA 94720-3190 Berkeley, USA fon: 510-642-2839 fax: 510-642-3192 web: http://fiebach.org email: mailto:[EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] graph question
I would like to produce a graph which plots a log scale variable on the y-axis but have the tick marks on the y-axis be the non log transformed values that are round like .5, 1, 2, 3, 4 etc. Has anyone done something like this in the past? How did you implement it in the code? Thanks, Dean __ [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] Course***R/S-plus Fundamentals and Programming Techniques In New York and Chicago
XLSolutions Corporation (www.xlsolutions-corp.com) is proud to announce 2-day R/S-plus Fundamentals and Programming Techniques. New York, NY --- November 18th-19th Chicago, IL -- November 11th-12th Reserve your seat now at the early bird rates! Payment due AFTER the class. Course Description: This two-day beginner to intermediate R/S-plus course focuses on a broad spectrum of topics, from reading raw data to a comparison of R and S. We will learn the essentials of data manipulation, graphical visualization and R/S-plus programming. We will explore statistical data analysis tools,including graphics with data sets. How to enhance your plots. We will perform basic statistics and fit linear regression models. Participants are encouraged to bring data for interactive sessions With the following outline: - An Overview of R and S - Data Manipulation and Graphics - Using Lattice Graphics - A Comparison of R and S-Plus - How can R Complement SAS? - Writing Functions - Avoiding Loops - Vectorization - Statistical Modeling - Project Management - Techniques for Effective use of R and S - Enhancing Plots - Using High-level Plotting Functions - Building and Distributing Packages (libraries) Email us for group discounts. Email Sue Turner: [EMAIL PROTECTED] Phone: 206-686-1578 Visit us: www.xlsolutions-corp.com/training.htm Please let us know if you and your colleagues are interested in this classto take advantage of group discount. Register now to secure your seat! Interested in R/Splus Advanced course? email us. Cheers, Elvis Miller, PhD Manager Training. XLSolutions Corporation 206 686 1578 www.xlsolutions-corp.com [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] constrained optimization using nlm/optim?
You want the penalty for the constraint to depend on how much the constraint is broken -- unlike your sample code. Page 336 of S Poetry has an example. Patrick Burns Burns Statistics [EMAIL PROTECTED] +44 (0)20 8525 0696 http://www.burns-stat.com (home of S Poetry and A Guide for the Unwilling S User) Jeff D. Hamann wrote: I'm looking for an example of a simple R script that impliments a contrained nonlinear function using nlm or optim. I'm not exactly sure how to impliment the constraints within the objective function that is passed to nlm/optim. obj.func - function( p ) { x(p) - unconstrained obj function value if( constraint1 something ) { obj.func - x(p) } else { obj.func - some super huge number } } p - c(0.1,2.4, 5) nlm( obj.func, p, and a bunch of other arguments ) Any suggestions would be very helpful... Jeff. __ [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] problem with lapack.so under Debian Sid
On Tue, 12 Oct 2004, Christian Fiebach wrote: Dear list, I am sorry to bother you with this. I just upgraded yesterday to R 2.0.0 (using apt-get under Debian Sid), and now have problems running e.g., summary(lm(...)) the lm is calculated, but the summary statement gives me the following error: Error in La.chol2inv(x, size) : lapack routines cannot be loaded In addition: Warning message: unable to load shared library /usr/lob/R/modules/lapack.so: /usr/lib/Pentium4_SSE2_512KB/liblapack.so.3: undefined symbol: ieeeck_ I searched through the archives, but unfortunately I am just not LINUX-knowledgeable enough to understand the discussions concerning the algebra libraries. Following an older mail to this list, I did ldd /usr/lib/R/modules/lapack.so everything looks fine except probably for the first line, which reads libR.so = not found Use R CMD ldd /usr/lib/R/modules/lapack.so to get an accurate picture. I think you need to ask this on a Debian list. The R developers do not recommend the use of an external Lapack library (and the Debian ones have had `patches' that were incorrect before now). (See the discussion in the R-admin manual.) In this case I guess that it has not been linked against the right set of external libraries for your particular machine. I would appreciate any hints on how to fix this problem! Should work if you build R from the sources with the default options. -- 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
Re: [R] constrained optimization using nlm/optim?
My guess is that approach will not work particularly well with nlm(), which expects the objective function to be smooth. You might have better luck using the Nelder-Mead method in optim(). -roger Jeff D. Hamann wrote: I'm looking for an example of a simple R script that impliments a contrained nonlinear function using nlm or optim. I'm not exactly sure how to impliment the constraints within the objective function that is passed to nlm/optim. obj.func - function( p ) { x(p) - unconstrained obj function value if( constraint1 something ) { obj.func - x(p) } else { obj.func - some super huge number } } p - c(0.1,2.4, 5) nlm( obj.func, p, and a bunch of other arguments ) Any suggestions would be very helpful... Jeff. __ [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] graph question
On Tue, 12 Oct 2004, Dean Sonneborn wrote: I would like to produce a graph which plots a log scale variable on the y-axis but have the tick marks on the y-axis be the non log transformed values that are round like .5, 1, 2, 3, 4 etc. Has anyone done something like this in the past? How did you implement it in the code? plot(c(0.5, 5), 1:2, log=x, xaxt=n) axis(1, at=c(0.5, 1:4)) Take a look at `An Introduction to R' and ?axis. BTW, please try to use a more precise subject line. Something like `Adding a custom axis to log-scale plots'? -- 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
Re: [R] 64-bit R on Intel Xeon EM64T running fine
Roger D. Peng wrote: This is good news. As far as I know R has built for quite some time now on a number of 64 bit platforms (Linux on AMD Opteron/Athlon64, Solaris/Sparc) Ill add SGI/IRIX 64 bit platform to the list. I've been running a 64 bit-compiled R on an SGI octane 2 for over a year now without a problem. R sessions often allocate around 8 GB of memory. but I can't recall seeing a build on Intel with the 64 bit extensions. By the way, did you happen to run `make check' just for kicks? -roger Michael Seewald wrote: Dear mailing-list members, In the days of cheap RAM and microarray applications feasting on memory, 64-bit computers become more and more useful - to actually make use of memory beyond the magic 4GB border. I would like to report the success of running 64-bit R on an Intel Xeon EM64T machine under Linux. Just like on an AMD Opteron, R v2.0.0 compiles fine (and out of the box) and is happily allocating memory until RAM and swap reach their limit. Hardware: - HP xw6200 workstation - dual Intel Xeon 3.4GHz with hyper-threading enabled - 4GB RAM, 4GB swap System: either - Fedora Core 2 x86_64 bit Linux or - Red Hat Enterprise Linux Workstation 3.0 x86_64 bit R: - v2.0.0 Really, no problems at all during setup, a big thank you to the R developers making this possible! Best wishes, Michael __ [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 -- Murad Nayal M.D. Ph.D. Department of Biochemistry and Molecular Biophysics College of Physicians and Surgeons of Columbia University 630 West 168th Street. New York, NY 10032 Tel: 212-305-6884 Fax: 212-305-6926 __ [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] bandwidths for bivariate density estimation
Hi, I am using the KernSmooth package to estimate nonparametrically bivariate density functions. However, it seems that the bandwidths (one for each co-ordinate direction) have to be selected manually. This does not apply for the univariate case, for which dpik (included in KernSmooth) uses up-to-date plug-in rules. Does anyone know about a package, or function, which estimates bandwidths for the BIVARIATE case, following, for instance, the plug-in methods suggested by Wand and Jones? I will appreciate your help a lot. Thanks, Emili References: Wand, M. P. and Jones, M. C. (1995). Kernel Smoothing. Chapman and Hall, London. [[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] bandwidths for bivariate density estimation
Hi, I am using the KernSmooth package to estimate nonparametrically bivariate density functions. However, it seems that the bandwidths (one for each co-ordinate direction) have to be selected manually. This does not apply for the univariate case, for which dpik (included in KernSmooth) uses up-to-date plug-in rules. Does anyone know about a package, or function, which estimates bandwidths for the BIVARIATE case, following, for instance, the plug-in methods suggested by Wand and Jones? I will appreciate your help a lot. Thanks, Emili References: Wand, M. P. and Jones, M. C. (1995). Kernel Smoothing. Chapman and Hall, London. [[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] plot hclust - canberra dist + median linkage
On Tue, 12 Oct 2004, Martin Maechler wrote: Dan == Dan Bolser [EMAIL PROTECTED] on Mon, 11 Oct 2004 16:21:53 +0100 (BST) writes: Dan Gives strange results. Dan I get 'weird' dendrograms with canberra / binary distance metric and Dan median / centroid cluster methods. it doesn't depend on the metric: Both 'median' and 'centroid' methods are known to *not* guarantee ``monotone distance measures'', or equivalently to possibly lead to dendrograms with so called ``inversions''. We should add this to help page for hclust(). Cool, I could do with more to read about these methods. It might be worth noteing that no bootstrap exists for the method (implemented in R that is). How can I quickly run a test on a sub-set of x? (i.e. quick bootstrap by hand)? Cheers, Dan. Probably for this reason, agnes() from the cluster package doesn't have these two methods [explicitly] -- though it now allows general parameter Lance-William formula methods which can also lead to inversions. Dan Is this just my data? evidently not. Though the problem does not appear for all data sets... Regards, Martin Maechler __ [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] problem with lapack.so under Debian Sid
On Tue, Oct 12, 2004 at 10:46:09AM -0700, Christian Fiebach wrote: Dear list, I am sorry to bother you with this. I just upgraded yesterday to R 2.0.0 (using apt-get under Debian Sid), and now have problems running e.g., summary(lm(...)) the lm is calculated, but the summary statement gives me the following error: Error in La.chol2inv(x, size) : lapack routines cannot be loaded In addition: Warning message: unable to load shared library /usr/lob/R/modules/lapack.so: /usr/lib/Pentium4_SSE2_512KB/liblapack.so.3: undefined symbol: ieeeck_ Is that the liblapack, i.e. do you actually have a Pentium IV ? It could be that this isn;t R specific, but related to the choice Debian gives with respect to lapack and blas libraries. You could try to downgrade to the refblas3 and lapack3 packages. We should probably continue this off-list. Dirk I searched through the archives, but unfortunately I am just not LINUX-knowledgeable enough to understand the discussions concerning the algebra libraries. Following an older mail to this list, I did ldd /usr/lib/R/modules/lapack.so everything looks fine except probably for the first line, which reads libR.so = not found I would appreciate any hints on how to fix this problem! Thanks a lot. Christian -- Christian Fiebach, PhD Department of Psychology Helen Wills Institute of Neuroscience 132 Barker Hall, MC# 3190 University of California, Berkeley CA 94720-3190 Berkeley, USA fon: 510-642-2839 fax: 510-642-3192 web: http://fiebach.org email: mailto:[EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Those are my principles, and if you don't like them... well, I have others. -- Groucho Marx __ [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] lm#contrasts#one level in factor: bug or feature
On Tue, 12 Oct 2004, Ritter, Christian C MCIL-CTANL/S wrote: (R.1.9.1; win2000) Since it's about the tenth time I had to write an if around this to catch the error ... Let's look at the line myfit-lm(res~groupvar,data=Data) Here res is of numeric type and groupvar is a factor. On first sight, it would be logical that if groupvar had only one (single) level we would get: Error in contrasts-(`*tmp*`, value = contr.treatment) : contrasts can be applied only to factors with 2 or more levels But then again, it's also inconsistent: Normally redundant variables on the right of the ~ (variables which are constant or linearly dependent on previous variables) don't lead to Error. They just get a NA as a coefficient estimate. A factor with a single level is of this type, it is just a constant. Obviously (to me) lm (or model.matrix.default) should not try to calculate contrasts on it and it's coefficient should be NA. Shouldn't it? No, a factor with k levels isn't just a set of k linearly dependent variables (otherwise a factor would always give an NA coefficient for one level). A factor specifies a set of contrasts (usually k-1 of them) that go into the model matrix. In that sense a factor with 1 level should specify zero columns of the model matrix, rather than giving an error. Why is this not unimportant? Imagine you make a model of the above type for a data set with more than one level in groupvar. Then you select subsets and refit. It shouldn't die with error if you select a subset with only one level of groupvar. Possibly, but your results are likely to be pretty meaningless. If the subset does not have *all* the levels of the factor then the contrasts will change and the variable will be coded differently, possibly giving coefficients with different meanings. In the case of contr.treatment, for example, if the lowest level of the factor is missing then all the other levels are recoded as contrasts to the next lowest level. -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
[R] covariate selection?
Hello, I am hoping someone can help me with the following multivariate issue: I have a model consisting of about 50 covariates. I would like to reduce this to about 5 covariate for the reduced model by combining cofactors that are strongly correlated. Is there a package or function that would help me with this in R? I appreciate any suggestions. Thanks, Ian __ [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] Why I can't retrieve GO identifier correctly?
Hello, R experts, I tried to retrieve all biological process GO terms at level 3 starting biological process as level 1 using the code as bellows: 1 library(GO) 2 library(GOstats) 3 level2-getGOChildren(GO:0008150)$GO:0008150$Children 4 for ( i in 1:length(level2)) { 5level3 - getGOChildren(level2[i])$level2[i]$Children 6for ( j in 1:length(level3)){ 7 level3term - getGOTerm(as.character(level3[j]))$BP$level3[j] 8 level3term 9} 10 } What is the difference between line 3 and line5? In the line 3 I retrieved the GO identifiers at level 2 successfullly but in the line 5 I got nothing. How to correct the line 5 to retrieve the GO terms at level 3 correctly? Thank you in advance! Josh __ [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] covariate selection?
Hello Ian, ?princomp If your covariates are scalars, and the following documents: http://www.jstatsoft.org/v07/i01/drdoc.pdf http://www.bioconductor.org/workshops/Milan/PDF/Lab12.pdf Best wishes. Saludos, Juan Carlos Martínez Ovando Banco de México Av. 5 de Mayo No. 18 Piso 5 Sección D Col. Centro 06059 México, D. F. Tel. +52 55 52.37.20.00 ext. 3594 Fax. +52 55 52.37.27.03 e-mail: [EMAIL PROTECTED] -Mensaje original- De: Ian Fiske [mailto:[EMAIL PROTECTED] Enviado el: Martes, 12 de Octubre de 2004 04:08 PM Para: [EMAIL PROTECTED] Asunto: [R] covariate selection? Hello, I am hoping someone can help me with the following multivariate issue: I have a model consisting of about 50 covariates. I would like to reduce this to about 5 covariate for the reduced model by combining cofactors that are strongly correlated. Is there a package or function that would help me with this in R? I appreciate any suggestions. Thanks, Ian __ [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] A question in R
Hi, On Tue, 12 Oct 2004, Kevin Bartz wrote: R looks less than fondly on Excel files. The easiest solution for you will be to export your Excel file to a tab-delimited text format (Save - (.txt) Tab-delimited Text), and then use read.delim as you did. Does that make sense? The alternative (tricky but more convinient in the long run) is to use the read.xls() function from gregmisc bundle. Cheers, Kevin Ko-Kang Kevin Wang PhD Student Centre for Mathematics and its Applications Building 27, Room 1004 Mathematical Sciences Institute (MSI) Australian National University Canberra, ACT 0200 Australia Homepage: http://wwwmaths.anu.edu.au/~wangk/ Ph (W): +61-2-6125-2431 Ph (H): +61-2-6125-7407 Ph (M): +61-40-451-8301 __ [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] Why I can't retrieve GO identifier correctly?
[EMAIL PROTECTED] writes: Hello, R experts, I tried to retrieve all biological process GO terms at level 3 starting biological process as level 1 using the code as bellows: 1 library(GO) 2 library(GOstats) 3 level2-getGOChildren(GO:0008150)$GO:0008150$Children 4 for ( i in 1:length(level2)) { 5level3 - getGOChildren(level2[i])$level2[i]$Children 6for ( j in 1:length(level3)){ 7 level3term - getGOTerm(as.character(level3[j]))$BP$level3[j] 8 level3term 9} 10 } What is the difference between line 3 and line5? In the line 3 I retrieved the GO identifiers at level 2 successfullly but in the line 5 I got nothing. How to correct the line 5 to retrieve the GO terms at level 3 correctly? Thank you in advance! Josh I think you'll have better luck with that question if you take it to the Bioconductor list. -- 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] covariate selection?
Thanks Juan. I thought that was what I was looking for, but really, I want to know which of the original covariates could best be used to take advantage of their colinearity without creating new variables. I think PCA creates new variables. SAS and SPSS can do what I'm talking about, but I would like to use R for this. Thanks, Ian Martínez Ovando Juan Carlos wrote: Hello Ian, ?princomp If your covariates are scalars, and the following documents: http://www.jstatsoft.org/v07/i01/drdoc.pdf http://www.bioconductor.org/workshops/Milan/PDF/Lab12.pdf Best wishes. Saludos, Juan Carlos Martínez Ovando Banco de México Av. 5 de Mayo No. 18 Piso 5 Sección D Col. Centro 06059 México, D. F. Tel. +52 55 52.37.20.00 ext. 3594 Fax. +52 55 52.37.27.03 e-mail: [EMAIL PROTECTED] -Mensaje original- De: Ian Fiske [mailto:[EMAIL PROTECTED] Enviado el: Martes, 12 de Octubre de 2004 04:08 PM Para: [EMAIL PROTECTED] Asunto: [R] covariate selection? Hello, I am hoping someone can help me with the following multivariate issue: I have a model consisting of about 50 covariates. I would like to reduce this to about 5 covariate for the reduced model by combining cofactors that are strongly correlated. Is there a package or function that would help me with this in R? I appreciate any suggestions. Thanks, Ian __ [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] problem with lapack.so under Debian Sid
On Tue, 12 Oct 2004, Christian Fiebach wrote: Dear list, I am sorry to bother you with this. I just upgraded yesterday to R 2.0.0 (using apt-get under Debian Sid), and now have problems running e.g., summary(lm(...)) the lm is calculated, but the summary statement gives me the following error: Error in La.chol2inv(x, size) : lapack routines cannot be loaded In addition: Warning message: unable to load shared library /usr/lob/R/modules/lapack.so: /usr/lib/Pentium4_SSE2_512KB/liblapack.so.3: undefined symbol: ieeeck_ I searched through the archives, but unfortunately I am just not LINUX-knowledgeable enough to understand the discussions concerning the algebra libraries. Following an older mail to this list, I did ldd /usr/lib/R/modules/lapack.so everything looks fine except probably for the first line, which reads libR.so = not found So does that mean that you have one of the Debian packages atlas3-base or atlas3-3dnow or atlas3-sse or atlas3-sse2 or lapack3 installed? You should have had at least one of those installed to be able to install the r-base-core Debian package. In the output from ldd /usr/lib/R/modules/lapack.so, does liblapack.so.3 get resolved to /usr/lib/liblapack.so.3 and, if so, what is that file? It is probably a link to /usr/lib/liblapack-3.so which, in turn, is a link to /etc/alternatives/liblapack-3.so. Check that the file linked as /etc/alternatives/liblapack-3.so actually exists. __ [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] covariate selection?
Have you considered stepwise regression, e.g., step or stepAIC in library(MASS)? The documentation for both contain examples. hope this helps. spencer graves Ian Fiske wrote: Hello, I am hoping someone can help me with the following multivariate issue: I have a model consisting of about 50 covariates. I would like to reduce this to about 5 covariate for the reduced model by combining cofactors that are strongly correlated. Is there a package or function that would help me with this in R? I appreciate any suggestions. Thanks, Ian __ [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 -- Spencer Graves, PhD, Senior Development Engineer O: (408)938-4420; mobile: (408)655-4567 __ [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] covariate selection?
Hello Ian, Sorry. I don't really understand your problem, which is of model selection. That's right? You could use some criteria based in likelihood. For instante Akaike (AIC) or Schwarz criteria (BIC), see: ?AIC ?mle.aic (The best model is determined minimizing AIC or BIC). I hope this help you. Greetings, Juan Carlos -Mensaje original- De: Ian Fiske [mailto:[EMAIL PROTECTED] Enviado el: Martes, 12 de Octubre de 2004 05:17 PM Para: Martínez Ovando Juan Carlos CC: [EMAIL PROTECTED] Asunto: Re: [R] covariate selection? Thanks Juan. I thought that was what I was looking for, but really, I want to know which of the original covariates could best be used to take advantage of their colinearity without creating new variables. I think PCA creates new variables. SAS and SPSS can do what I'm talking about, but I would like to use R for this. Thanks, Ian Martínez Ovando Juan Carlos wrote: Hello Ian, ?princomp If your covariates are scalars, and the following documents: http://www.jstatsoft.org/v07/i01/drdoc.pdf http://www.bioconductor.org/workshops/Milan/PDF/Lab12.pdf Best wishes. Saludos, Juan Carlos Martínez Ovando Banco de México Av. 5 de Mayo No. 18 Piso 5 Sección D Col. Centro 06059 México, D. F. Tel. +52 55 52.37.20.00 ext. 3594 Fax. +52 55 52.37.27.03 e-mail: [EMAIL PROTECTED] -Mensaje original- De: Ian Fiske [mailto:[EMAIL PROTECTED] Enviado el: Martes, 12 de Octubre de 2004 04:08 PM Para: [EMAIL PROTECTED] Asunto: [R] covariate selection? Hello, I am hoping someone can help me with the following multivariate issue: I have a model consisting of about 50 covariates. I would like to reduce this to about 5 covariate for the reduced model by combining cofactors that are strongly correlated. Is there a package or function that would help me with this in R? I appreciate any suggestions. Thanks, Ian __ [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] covariate selection?
Hi Ian Have you tried help.search(pca)? Christian -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Martínez Ovando Juan Carlos Sent: Tuesday, October 12, 2004 7:56 PM To: Ian Fiske Cc: [EMAIL PROTECTED] Subject: RE: [R] covariate selection? Hello Ian, Sorry. I don't really understand your problem, which is of model selection. That's right? You could use some criteria based in likelihood. For instante Akaike (AIC) or Schwarz criteria (BIC), see: ?AIC ?mle.aic (The best model is determined minimizing AIC or BIC). I hope this help you. Greetings, Juan Carlos -Mensaje original- De: Ian Fiske [mailto:[EMAIL PROTECTED] Enviado el: Martes, 12 de Octubre de 2004 05:17 PM Para: Martínez Ovando Juan Carlos CC: [EMAIL PROTECTED] Asunto: Re: [R] covariate selection? Thanks Juan. I thought that was what I was looking for, but really, I want to know which of the original covariates could best be used to take advantage of their colinearity without creating new variables. I think PCA creates new variables. SAS and SPSS can do what I'm talking about, but I would like to use R for this. Thanks, Ian Martínez Ovando Juan Carlos wrote: Hello Ian, ?princomp If your covariates are scalars, and the following documents: http://www.jstatsoft.org/v07/i01/drdoc.pdf http://www.bioconductor.org/workshops/Milan/PDF/Lab12.pdf Best wishes. Saludos, Juan Carlos Martínez Ovando Banco de México Av. 5 de Mayo No. 18 Piso 5 Sección D Col. Centro 06059 México, D. F. Tel. +52 55 52.37.20.00 ext. 3594 Fax. +52 55 52.37.27.03 e-mail: [EMAIL PROTECTED] -Mensaje original- De: Ian Fiske [mailto:[EMAIL PROTECTED] Enviado el: Martes, 12 de Octubre de 2004 04:08 PM Para: [EMAIL PROTECTED] Asunto: [R] covariate selection? Hello, I am hoping someone can help me with the following multivariate issue: I have a model consisting of about 50 covariates. I would like to reduce this to about 5 covariate for the reduced model by combining cofactors that are strongly correlated. Is there a package or function that would help me with this in R? I appreciate any suggestions. Thanks, Ian __ [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 __ [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] read 4-jan-02 as date
Thank you everyone. Indeed, I had read the data via read.csv and the date column was a factor. Everything works fine if I convert to character first. Regards, b. --- Sundar Dorai-Raj [EMAIL PROTECTED] wrote: bogdan romocea wrote: Dear R users, I have a column with dates (character) in a data frame: 12-Jan-01 11-Jan-01 10-Jan-01 9-Jan-01 8-Jan-01 5-Jan-01 and I need to convert them to (Julian) dates so that I can sort the whole data frame by date. I thought it would be very simple, but after checking the documentation and the list I still don't have something that works. 1. as.Date returns the error below. What am I doing wrong? As far as I can see the character strings are in standard format. d$Date - as.Date(d$Date, format=%d-%b-%y) Error in fromchar(x) : character string is not in a standard unambiguous format 2. as.date {Survival} produces this error, d$Date - as.date(d$Date, order = dmy) Error in as.date(d$Date, order = dmy) : Cannot coerce to date format 3. Assuming all else fails, is there a text function similar to SCAN in SAS? Given a string like 9-Jan-01 and - as separator, I'd like a function that can read the first, second and third values (9, Jan, 01), so that I can get Julian dates with mdy.date {survival}. Thanks in advance, b. If you're reading this from a file (via read.table, for example), then your date column is probably a factor. Convert to character first. x [1] 12-Jan-01 11-Jan-01 10-Jan-01 9-Jan-01 8-Jan-01 5-Jan-01 Levels: 10-Jan-01 11-Jan-01 12-Jan-01 5-Jan-01 8-Jan-01 9-Jan-01 Date(x, format=%d-%b-%y) Error in fromchar(x) : character string is not in a standard unambiguous format sort(as.Date(as.character(x), format=%d-%b-%y)) [1] 2001-01-05 2001-01-08 2001-01-09 2001-01-10 2001-01-11 [6] 2001-01-12 --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
RE: [R] 64-bit R on Intel Xeon EM64T running fine
This article might be of interest to some: http://www.linux-mag.com/2004-07/athlon_01.html Regarding 64-bit build of R, I can confirm Irix, Alpha and AIX, although some w/o readline/jpeg/png/X support, partially because of difficulties linking against 64-bit version of those libraries. Best, Andy From: Prof Brian Ripley On Tue, 12 Oct 2004, Roger D. Peng wrote: This is good news. As far as I know R has built for quite some time now on a number of 64 bit platforms (Linux on AMD Opteron/Athlon64, Solaris/Sparc and Alpha and Irix and HP-UX and AIX as far as I understand. ) but I can't recall seeing a build on Intel with the 64 bit extensions. As I understand it, this is an Intel `clone' of AMD64, so the only news would be if there were any problems: in almost all cases the executable code is identical to that compiled for AMD64 and in the GNU classification it is also x86_64-unknown-linux-gnu. I would even expect a Goto BLAS to work, if not as well as on the specific Opteron it is tuned for. (Fast BLASes are an essential part of making the most of 64-bit processors on large problems.) Michael Seewald wrote: Dear mailing-list members, In the days of cheap RAM and microarray applications feasting on memory, 64-bit computers become more and more useful - to actually make use of memory beyond the magic 4GB border. I would like to report the success of running 64-bit R on an Intel Xeon EM64T machine under Linux. Just like on an AMD Opteron, R v2.0.0 compiles fine (and out of the box) and is happily allocating memory until RAM and swap reach their limit. Hardware: - HP xw6200 workstation - dual Intel Xeon 3.4GHz with hyper-threading enabled - 4GB RAM, 4GB swap System: either - Fedora Core 2 x86_64 bit Linux or - Red Hat Enterprise Linux Workstation 3.0 x86_64 bit R: - v2.0.0 Really, no problems at all during setup, a big thank you to the R developers making this possible! -- 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 __ [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] Diagnosing trouble with R-2.0, Fedora Core 2, and Rcmdf
Dear Paul, I've now had a chance to check into these problems further: -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Paul Johnson Sent: Monday, October 11, 2004 1:09 PM To: r help Subject: [R] Diagnosing trouble with R-2.0, Fedora Core 2, and Rcmdf Greetings, R-help! On 2 Fedora Core 2 Linux systems, i've completely erased the previous R and all packages and then installed R-2.0 and installed fresh packages. In using Rcmdr, I see some trouble and I wonder if other people see this and if it is due to the tcl/tk, or R, or Rcmdr. (If readers have not yet tried Rcmdr, I recommend it not just because of the GUI it provides, but also because Prof. Fox has created several very handy commands, like scatter3d, which make it possible to use advanced packages like rgl. Rcmdr provides several very handy wrapper commands that just work and users don't have to fuss over so many details! And, if you are new at R, you can use pull down menus and then Rcmdr displays the commands that correspond with those menu options. Very educational!) In no particular order, here are the issues I see with R-2.0 and Rcmdr 1. In the panel to select a dataset from a package (nenu: Data/Data in packages), I can type in the name of a dataset, but the GUI will not let me choose the name of a package from which to select a dataset. Nothing happens when I pick a package name. (With previous R/Rcmdr, I did not have this trouble). I had in fact fixed this in the development version of the package (Rcmdr 0.9-12). I've now posted that to my web site, at http://socserv.socsci.mcmaster.ca/jfox/Misc/Rcmdr/index.html. I'd like to do some more testing before sending this version to CRAN. 2. When using Rcmdr, some of the output from commands shows in the Rcmdr bottom window, but some is sent to the xterm in which R was started. For example, with a scatter3d() command, if I add the option model.summary=T, the model summary shows in the xterm. But some other models show results in the bottom pane of Rcmdr. And, it seems to me, there are some commands in which I've noticed some of the output going to the Rcmdr bottom buffer and some to the xterm. I did in fact try to deal with this problem. The solution that I employed didn't work for scatter3d() with model.summary=TRUE, since the printed summaries were produced as side effects by scatter3d(). That's a bad practice, and I've fixed it; now scatter3d() returns a list of summaries, and the Rcmdr can cope with that. (As you know, it's not my intention that model.summary=TRUE be set from the Rcmdr GUI, but it's better to have it work if you enter this from the script editor.) You imply, however, that there are other cases of this problem. It would help to know what they are -- assuming that they're not fixed in the development version now on my web site. 3. When I want to exit Rcmdr and choose the pull down exit from Commander and R, I get a Segmentation Fault that closes Rcmdr and R. I haven't been able to duplicate this problem, to this point testing only under Windows XP. Does this occur only when an rgl graphics device is open? (The Rcmdr should close the rgl device before it exits.) 4. 3d plots do not display on the screen until I click in the rgl window that displays the graph. This might be a video card/AGP driver problem, I suppose. These systems have the NVidia FX5200 cards and the NVidia proprietary X driver. If other users don't see the same on other kinds of systems, I guess that will tell me where the problem lies. As I said, I believe that I've seen this problem before under Quantian. It would help to know whether it's peculiar to the Rcmdr, scatter3d(), or rgl. 5. At random times, I get an error tcl/tk window popping up with a message about MASS and categorical variables. It says Error in data(package=parse(text=packageName)); 'MASS' must be character string or NULL It says that over and over, sometimes it has other package names, as in: Error in data(package=parse(text=packageName)); 'relimp' must be character string or NULL These seem to be harmless? This should go away now that problem 1 is fixed. Regards, John -- Paul E. Johnson email: [EMAIL PROTECTED] Dept. of Political Sciencehttp://lark.cc.ku.edu/~pauljohn 1541 Lilac Lane, Rm 504 University of Kansas Office: (785) 864-9086 Lawrence, Kansas 66044-3177 FAX: (785) 864-5700 __ [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] covariate selection?
Ian Fiske wrote: Hello, I am hoping someone can help me with the following multivariate issue: I have a model consisting of about 50 covariates. I would like to reduce this to about 5 covariate for the reduced model by combining cofactors that are strongly correlated. Is there a package or function that would help me with this in R? I appreciate any suggestions. Thanks, Ian __ [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 have a look at package leaps, and also consider ridge regression. -- Kjetil Halvorsen. Peace is the most effective weapon of mass construction. -- Mahdi Elmandjra __ [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] covariate selection?
I like Kjetil's suggestion of a shrinkage estimator. Perhaps this would be a good time to experiment with Trevor Hastie's 'lars' package. If you have a lot of correlated inputs I might suggest using Andy Liaw's randomforest package. I have found this technique to be very valuable in this setting. The partial dependency plots are a good way to explore the functional relationships of the variables. --Matt -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Kjetil Brinchmann Halvorsen Sent: Tuesday, October 12, 2004 17:16 PM To: Ian Fiske Cc: [EMAIL PROTECTED] Subject: Re: [R] covariate selection? Ian Fiske wrote: Hello, I am hoping someone can help me with the following multivariate issue: I have a model consisting of about 50 covariates. I would like to reduce this to about 5 covariate for the reduced model by combining cofactors that are strongly correlated. Is there a package or function that would help me with this in R? I appreciate any suggestions. Thanks, Ian __ [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 have a look at package leaps, and also consider ridge regression. -- Kjetil Halvorsen. Peace is the most effective weapon of mass construction. -- Mahdi Elmandjra __ [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] tclk, tcltk
From NEWS of R-2.0.0: o library() now checks the version dependence (if any) of required packages mentioned in the Depends: field of the DESCRIPTION file. This is not done in R-1.9.1 and prior, I believe (or it wouldn't be in NEWS). Andy From: Jean Eid Thanks to Brian. I was looking everywhere except the typo. I guess the previous build (1.9.1) ignored the dependece error. Thanks again, jean On Tue, 12 Oct 2004, Prof Brian Ripley wrote: What is package `tclk'? I think you have a typo somewhere in your code. At a guess you have Depends: tclk in a DESCRIPTION file for one of your packages. On Tue, 12 Oct 2004, Jean Eid wrote: I have been having problems with these two 'libraries' since I installed 2.0.0. I have built a package with couple of functions so that I can load it at startup every time R is booted. The problem is that I have the following error every time I call the library Loading required package: tclk Error: package 'tclk' could not be loaded In addition: Warning message: There is no package called 'tclk' in: library(pkg, character.only = TRUE, logical = TRUE, lib.loc = lib.loc) However I can load tcltk library now (thanks to Dirk for helping me with this). I can load the functions I have in the library and everything works fine. I cannot find any information on this except the following thread which was not resolved. http://tolstoy.newcastle.edu.au/R/help/04/01/0410.html It was resolved, and in any case it was about installing 64-bit Tcl/Tk on Solaris, something that was (and is) covered in the R Installation manual. Any help or ideas are greatly appreciated. P.S. what is the difference between tclk and tcltk? and why calling the installed library is requiring tclk? I just made up a dummy package with a function fun-function(x) x^2 just to test and the same thing happens (Loading required package: tclk ...). This is on a Linux debian unstable with the kernel 2.4.20-bf2.4-xfs -- 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 __ [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] KalmanLike: missing exogenous factor?
Prof Ripley, Thanks for explanation. I now understand where KalmanLike fits. I should not use exogenous factor. It should be called exogenous variable or inputs or known effects. My study on how trading sizes impact on stock prices has trading sizes as this exogenous variable. As you said, this should belong to some package. I did internet searches and found something similar but not covering my case. The restrictive access to web makes subscription from my work place not convenient. Sorry. Heng Sun Senior Quantitative Analyst Depository Trust and Clearing Corporation New York, USA 10041 Tel: 212-755-5754 --- Prof Brian Ripley [EMAIL PROTECTED] wrote: On Mon, 11 Oct 2004, Heng Sun wrote: From the help document on KalmanLike, KalmanRun, etc., I see the linear Gaussian state space model is a - T a + R e y = Z' a + eta following the book of Durbin and Koopman. In practice, it is useful to run Kalman filtering/smoothing/forecasting with exogenous factor: a - T a + L b + R e y = Z' a + M b + eta where b is some known vector (a function of time). Some other software like S-plus and Mathematica include the above exogenous factor. SsfPack by Koopman, etal. also has the factor built in the model to accommodate practical uses. So what is the rationale for R to leave off the exogenous factor? Is there a feasible way to convert the general model to the simple model in R? What is the rationale for your raising this? KalmanLike, KalmanRun, etc were written for R 1.5.0 as part of the ts package (see my article in R-news), and the ts applications (see the See Also section) do not need a so-called `exogenous factor' (which is not a `factor'). R does not pretend to have facilities for whatever subject area you mean (but do not say) by `in practice'. That's what addon packages are for (and some do touch on this area). We have no idea who [EMAIL PROTECTED]' is: it is courteous to use a signature and give your affiliation. -- 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
RE: [R] graph question
Dean -- I believe just setting log = y in your plot command should do this. For example: plot(runif(100, 1, 100), runif(100, 1, 100), log = xy) gives me tick marks at 2, 5, 10, 20, 50, and 100. (YMMV because of the random numbers.) Hope this helps, Matt Wiener -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dean Sonneborn Sent: Tuesday, October 12, 2004 2:02 PM To: [EMAIL PROTECTED] Subject: [R] graph question I would like to produce a graph which plots a log scale variable on the y-axis but have the tick marks on the y-axis be the non log transformed values that are round like .5, 1, 2, 3, 4 etc. Has anyone done something like this in the past? How did you implement it in the code? Thanks, Dean __ [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] diagonal matrix construction
Hi, I have worked long and hard and looked in the manuals and am having a hard time constructing a diagonal matrix. I can get the diagonals out of a matrix but can't construct the matrix with just the diagonals. I have been on the web site and manuals and I think that it says to use: dsj - diag (three = 1, nrow, ncol = 7) three is the name of my matrix and dsj is where I'm trying to put it but it comes back: unused argument(s) (three ...) every time that I try it. Am I doing something wrong in trying to make this matrix? Thank you very much for your time, Chris Ryan Christopher W. Ryan Black Bear Project Leader West Virginia Division of Natural Resources Capitol Complex, Bldg 3, Rm 824 Charleston, WV 25305 [[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] data(eurodist) and PCA ??
If I perform PCA on the 'eurodist' data, should I get an accurate geographic layout of the cities with biplot? (barring inversions, i.e. their is no way to define north.. but you get the idea...) I have a complex distance matrix, and I am thinking about how to cluster it and how to visualize the quality of the resulting clusters. If I could 'see' the clusters in space I could understand how / what the cluster algorithms were doing. Can I use PCA over the distance matrix to to do that? Sorry for the dumb questions. Dan. __ [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] diagonal matrix construction
Chris Ryan chrisryan at wvdnr.gov writes: : : Hi, : : I have worked long and hard and looked in the manuals and am having a hard time constructing a diagonal : matrix. I can get the diagonals out of a matrix but can't construct the matrix with just the diagonals. I : have been on the web site and manuals and I think that it says to use: : : dsj - diag (three = 1, nrow, ncol = 7) three is the name of my matrix and dsj is where I'm trying to put it but it : comes back: : : unused argument(s) (three ...) : every time that I try it. : : Am I doing something wrong in trying to make this matrix? : : Thank you very much for your time, Chris Ryan : Looking at ?diag we see that diag takes 3 arguments: 'x', 'nrow' and 'ncol'. 'three' is not one of them so it is correctly telling you that you specified an unused argument. diag(nrow = 7) # will create a 7x7 identity matrix. diag(1, 7) # same using positional rather than named arguments diag(, 7) # same since x defaults to 1 Look at the examples at the bottom of the ?diag page and try them out by running example(diag) at the R prompt or by just typing them in to get more insight. Also you might consider rereading the Introduction to R manual where it is covered on page 22. __ [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] diagonal matrix construction
Did you try working the examples in the diag documentation, one of which is as follows: diag(10,3,4) # guess what? [,1] [,2] [,3] [,4] [1,] 10000 [2,]0 1000 [3,]00 100 (R 1.9.1 for Windows). The R syntax expects three in that context to be an argument of the function diag. If you want a 7x7 identity matrix, try diag(7). If you want something else, please study the examples in the documentation for diag. If that is not adequate, let us know. hope this helps. spencer graves Chris Ryan wrote: Hi, I have worked long and hard and looked in the manuals and am having a hard time constructing a diagonal matrix. I can get the diagonals out of a matrix but can't construct the matrix with just the diagonals. I have been on the web site and manuals and I think that it says to use: dsj - diag (three = 1, nrow, ncol = 7) three is the name of my matrix and dsj is where I'm trying to put it but it comes back: unused argument(s) (three ...) every time that I try it. Am I doing something wrong in trying to make this matrix? Thank you very much for your time, Chris Ryan Christopher W. Ryan Black Bear Project Leader West Virginia Division of Natural Resources Capitol Complex, Bldg 3, Rm 824 Charleston, WV 25305 [[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 -- Spencer Graves, PhD, Senior Development Engineer O: (408)938-4420; mobile: (408)655-4567 __ [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] debugging non-visible functions
Hi, I would like to step-through a non-visible function. but apparently I don't know enough about namespaces to get that to work: methods(predict) ... deleted lines ... [27] predict.rpart* predict.smooth.spline* [31] predict.survreg.penal* Non-visible functions are asterisked debug(predict.rpart) Error: Object predict.rpart not found getAnywhere(predict.rpart) A single object matching 'predict.rpart' was found It was found in the following places registered S3 method for predict from namespace rpart namespace:rpart with value function (object, newdata = list(), type = c(vector, prob, class, matrix), ...) { ... deleted code ... } environment: namespace:rpart debug(predict.rpart,pos=package:rpart) Error: Object predict.rpart not found how can I 'debug' non-visible functions, like predict.rpart? many thanks Murad Nayal __ [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] debugging non-visible functions
Murad Nayal mn216 at columbia.edu writes: : : Hi, : : I would like to step-through a non-visible function. but apparently I : don't know enough about namespaces to get that to work: : : methods(predict) : ... deleted lines ... : [27] predict.rpart* predict.smooth.spline* : [31] predict.survreg.penal* : : Non-visible functions are asterisked : : : debug(predict.rpart) : Error: Object predict.rpart not found : : : getAnywhere(predict.rpart) : A single object matching 'predict.rpart' was found : It was found in the following places : registered S3 method for predict from namespace rpart : namespace:rpart : with value : : function (object, newdata = list(), type = c(vector, prob, : class, matrix), ...) : { : ... deleted code ... : } : environment: namespace:rpart : : : debug(predict.rpart,pos=package:rpart) : Error: Object predict.rpart not found : : : how can I 'debug' non-visible functions, like predict.rpart? : From the ?debug page we see that that only one argument can be given to debug, the function name, so there is no two argument form as per your example above. What you can do is that you can debug it like this: require(rpart) debug(rpart:::predict.rpart) Also note that even if its not visible it may still have a help page, which in this case it does: ?predict.rpart __ [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