Re: [R] How to append to a data.frame?
Dear Prof. Ripley, Thank you very much for your fast and helpful reply! Is there a canonical way of doing this without a kludge or is the below adequate? d-as.data.frame(matrix(nrow=1000)) With many thanks again for your help, Yours sincerely, David Kreil. Dr David Philip Kreil (`-''-/).___..--''`-._ Research Fellow`6_ 6 ) `-. ( ).`-.__.`) University of Cambridge(_Y_.)' ._ ) `._ `. ``-..-' ++44 1223 764107, fax 333992 _..`--'_..-_/ /--'_.' ,' www.inference.phy.cam.ac.uk/dpk20 (il),-'' (li),' ((!.-' __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] How to append to a data.frame?
On Tue, 9 Dec 2003, David Kreil wrote: Dear Prof. Ripley, Thank you very much for your fast and helpful reply! Is there a canonical way of doing this without a kludge or is the below adequate? d-as.data.frame(matrix(nrow=1000)) You need to set up the columns as you need them to be eventually. With many thanks again for your help, Yours sincerely, David Kreil. Dr David Philip Kreil (`-''-/).___..--''`-._ Research Fellow`6_ 6 ) `-. ( ).`-.__.`) University of Cambridge(_Y_.)' ._ ) `._ `. ``-..-' ++44 1223 764107, fax 333992 _..`--'_..-_/ /--'_.' ,' www.inference.phy.cam.ac.uk/dpk20 (il),-'' (li),' ((!.-' -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] Font problem
Some plots fail due to a problem with the X11 fonts. I get a message that X11 font at size 22 could not be loaded. The demo() graphics routine for instance dies during the third chart. The graphics demo calls font.main=1 and that seems to be where the error is. I believe this is due to a configuration problem on my system, however I can't find where in the environment font.main looks for the font to use. I am running SuSE 9.0 and use the KDE desktop. However, I have also replicated this in GNOME and WindowMaker. Varying the fonts used by the console does notb effect the result. Thanks, John Dougherty __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] How to append to a data.frame?
Yes, once I've named the first column, I can add further ones by saying d[c(x,y,z)]=NA or such. I was just wondering whether that was the way to do it or whether there was a more elegant approach. Preallocation was the critical clue I needed. Thanks again for your help, David. Dr David Philip Kreil (`-''-/).___..--''`-._ Research Fellow`6_ 6 ) `-. ( ).`-.__.`) University of Cambridge(_Y_.)' ._ ) `._ `. ``-..-' ++44 1223 764107, fax 333992 _..`--'_..-_/ /--'_.' ,' www.inference.phy.cam.ac.uk/dpk20 (il),-'' (li),' ((!.-' __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] axes that meet
Hank Stevens wrote: R v. 1.7.1, Windows 2000. A particular journal wants me to provide scatter plots with no box, but with axes that meet in the lower left corner. It seems as though there must be an easy way of doing this, but my reading the help on plot.default, axis, and box have not provided any clues. I would be most appreciative of any feedback. Thank you, Hank Stevens Dr. Martin Henry H. Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/botany/bot/henry.html http://www.muohio.edu/ecology __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help See argument bty in ?par, e.g.: par(bty=l) plot(1:10) Uwe Ligges __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Interfacing R and Python in MS Windows
Héctor Villafuerte D. wrote: Hi all, I need the power of R from within some of my Python programs... I use debian linux (woody) at home and windows XP at work (the latter is where I need to get things done!) This are my packages: R 1.8.0 Python 2.3 RSPython 0.5-3 This is what I've done: (1) Since the Windows Binary of RSPython is compiled against Python 2.2 I downloaded the tarball (2) Followed the instructions in INSTALL.win (with pexports and everything) (3) In the RGUI Install package(s) from local zip files... If you have downloaded the tarball of RSPython, you have to install the package using Rcmd INSTALL, and you cannot use the RGUI which is designed to install binary packages. Uwe Ligges (4) NO errors reported during this process (5) When I try to Load package in R it show this error: local({pkg - select.list(sort(.packages(all.available = TRUE))) + if(nchar(pkg)) library(pkg, character.only=TRUE)}) Error in testRversion(descfields) : This package has not been installed properly See the Note in ?library (6) In Python import RS Traceback (most recent call last): File stdin, line 1, in ? ImportError: No module named RS Please help me to get this excelent tools going on in Windows. Thanks in advance, Hector __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] How to append to a data.frame?
On Tue, 9 Dec 2003, David Kreil wrote: Yes, once I've named the first column, I can add further ones by saying d[c(x,y,z)]=NA or such. I was just wondering whether that was the way to do it or whether there was a more elegant approach. Preallocation was the critical clue I needed. Use an initial data.frame call naming all the columns. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] How to append to a data.frame?
Ok, how can I both allocate storage and specify column names in a data.frame call, please? Apologies if I'm being slow here. With many thanks again, David. Yes, once I've named the first column, I can add further ones by saying d[c(x,y,z)]=NA or such. I was just wondering whether that was the way to do it or whether there was a more elegant approach. Preallocation was the critical clue I needed. Use an initial data.frame call naming all the columns. -- 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 Dr David Philip Kreil (`-''-/).___..--''`-._ Research Fellow`6_ 6 ) `-. ( ).`-.__.`) University of Cambridge(_Y_.)' ._ ) `._ `. ``-..-' ++44 1223 764107, fax 333992 _..`--'_..-_/ /--'_.' ,' www.inference.phy.cam.ac.uk/dpk20 (il),-'' (li),' ((!.-' __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] p-value from chisq.test working strangely on 1.8.1
Hello everybody, I'm seeing some strange behavior on R 1.8.1 on Intel/Linux compiled with gcc 3.2.2. The p-value calculated from the chisq.test function is incorrect for some input values: chisq.test(matrix(c(0, 1, 1, 12555), 2, 2), simulate.p.value=TRUE) Pearson's Chi-squared test with simulated p-value (based on 2000 replicates) data: matrix(c(0, 1, 1, 12555), 2, 2) X-squared = 1e-04, df = NA, p-value = 1 chisq.test(matrix(c(0, 1, 1, 12556), 2, 2), simulate.p.value=TRUE) [...] data: matrix(c(0, 1, 1, 12556), 2, 2) X-squared = 1e-04, df = NA, p-value = 2.2e-16 chisq.test(matrix(c(0, 1, 1, 12557), 2, 2), simulate.p.value=TRUE) [...] data: matrix(c(0, 1, 1, 12557), 2, 2) X-squared = 1e-04, df = NA, p-value = 1 this does not happen with R-1.8.1 and gcc 2.95.4 on Debian stable: R chisq.test(matrix(c(0, 1, 1, 12555), 2, 2), simulate.p.value=TRUE)$p.value [1] 1 R chisq.test(matrix(c(0, 1, 1, 12556), 2, 2), simulate.p.value=TRUE)$p.value [1] 1 R chisq.test(matrix(c(0, 1, 1, 12557), 2, 2), simulate.p.value=TRUE)$p.value [1] 1 neither with R-1.9.0 (unstable). Is this reproducible without using `set.seed' on your system? Best, Torsten In these three calls to chisq.test, I'm varying the input matrix by only 1 observation, but the p-value changes by 16 orders of magnitude. This is reproducible on my system. Please let me know if any other information would be useful. chisq.test works properly for these inputs on Mac OS X 10.3.1 with R 1.8.0. I don't know if the problem is with Linux or 1.8.1. This bug looks very similar to bug 4718, which was reported in R 1.8.0 and fixed in R 1.8.1. They may be related. http://r-bugs.biostat.ku.dk/cgi-bin/R/Analyses-fixed?id=4718; user=guest;selectid=4718 Jeff __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] The spdep package
Hi, Here is a listw object z corresponding to the matrix W. I understand n, nn, S0, S1 and S2 in the weights constants summary. Is it simply so that n1 = n-1, n2 = n-2 and n3 = n-3? If this is true where they are needed? Just wondering Osmo W [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [1,]010100000 [2,]101010000 [3,]010001000 [4,]100010100 [5,]010101010 [6,]001010001 [7,]000100010 [8,]000010101 [9,]000001010 z=mat2listw(W);z Characteristics of weights list object: Neighbour list object: Number of regions: 9 Number of nonzero links: 24 Percentage nonzero weights: 29.62963 Average number of links: 2.67 Weights style: M Weights constants summary: n n1 n2 n3 nn S0 S1 S2 M 9 8 7 6 81 24 48 272 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] R: the spdep package
Hi, Here is a listw object z corresponding to the matrix W. I understand n, nn, S0, S1 and S2 in the weights constants summary. Is it simply so that n1 = n-1, n2 = n-2 and n3 = n-3? If this is true where they are needed? Just wondering Osmo W [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [1,] 0 1 0 1 0 0 0 0 0 [2,] 1 0 1 0 1 0 0 0 0 [3,] 0 1 0 0 0 1 0 0 0 [4,] 1 0 0 0 1 0 1 0 0 [5,] 0 1 0 1 0 1 0 1 0 [6,] 0 0 1 0 1 0 0 0 1 [7,] 0 0 0 1 0 0 0 1 0 [8,] 0 0 0 0 1 0 1 0 1 [9,] 0 0 0 0 0 1 0 1 0 z=mat2listw(W);z Characteristics of weights list object: Neighbour list object: Number of regions: 9 Number of nonzero links: 24 Percentage nonzero weights: 29.62963 Average number of links: 2.67 Weights style: M Weights constants summary: n n1 n2 n3 nn S0 S1 S2 M 9 8 7 6 81 24 48 272 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] Alpha
Hello, I just install red-hat 7.2 for alpha station. I downloaded R-1.8.1-1.alpha.rpm, but unable to install it. I need the libblas.so.3 library. I did have a look around and did not find where to get that library for linux alpha version. -- regards Emmanuel POIZOT Cnam/Intechmer Digue de Collignon 50110 Tourlaville Tél : (33)(0)2 33 88 73 42 Fax : (33)(0)2 33 88 73 39 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Alpha
Try to install ATLAS (improved version of BLAS libraries): sources form http://www.netlib.org/atlas/index.html#software Best regards A.S. Alessandro Semeria Models and Simulations Laboratory Montecatini Environmental Research Center (Edison Group), Via Ciro Menotti 48, 48023 Marina di Ravenna (RA), Italy Tel. +39 544 536811 Fax. +39 544 538663 E-mail: [EMAIL PROTECTED] Poizot Emmanuel [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent by: cc: [EMAIL PROTECTED] Subject: [R] Alpha ath.ethz.ch 09-12-2003 10.07 Please respond to poizot Hello, I just install red-hat 7.2 for alpha station. I downloaded R-1.8.1-1.alpha.rpm, but unable to install it. I need the libblas.so.3 library. I did have a look around and did not find where to get that library for linux alpha version. -- regards Emmanuel POIZOT Cnam/Intechmer Digue de Collignon 50110 Tourlaville Tél : (33)(0)2 33 88 73 42 Fax : (33)(0)2 33 88 73 39 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] axes that meet
Hank, I think the graphical parameter you are looking for is bty, as in par(bty=l) Details on the ?par help topic. Hope that helps. Bill Bill Pikounis, Ph.D. Biometrics Research Department Merck Research Laboratories PO Box 2000, MailDrop RY33-300 126 E. Lincoln Avenue Rahway, New Jersey 07065-0900 USA Phone: 732 594 3913 Fax: 732 594 1565 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Hank Stevens Sent: Tuesday, December 09, 2003 1:18 AM To: [EMAIL PROTECTED] Subject: [R] axes that meet R v. 1.7.1, Windows 2000. A particular journal wants me to provide scatter plots with no box, but with axes that meet in the lower left corner. It seems as though there must be an easy way of doing this, but my reading the help on plot.default, axis, and box have not provided any clues. I would be most appreciative of any feedback. Thank you, Hank Stevens Dr. Martin Henry H. Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/botany/bot/henry.html http://www.muohio.edu/ecology __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] How to append to a data.frame?
David Kreil wrote: Ok, how can I both allocate storage and specify column names in a data.frame call, please? Apologies if I'm being slow here. With many thanks again, David. Something like (UNTESTED code follows) templateColumn - rep(NA,1000) # for 1000 rows foo - data.frame( x = templateColumn, y = templateColumn, z = templateColumn ) # or however many columns you need bar - foo for(ii in your.iterative.sequence) { if(ii 1000) { bar - rbind(bar,foo) } bar[ii,] - your.function() } Cheers Jason -- Indigo Industrial Controls Ltd. http://www.indigoindustrial.co.nz 64-21-343-545 [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] How to append to a data.frame?
David Kreil [EMAIL PROTECTED] writes: Ok, how can I both allocate storage and specify column names in a data.frame call, please? Apologies if I'm being slow here. It gets a little tricky. I'd try something along the lines of data.frame(age=as.numeric(NA),sex=factor(NA,levels=c(m,f)))[rep(1,20),] or data.frame(age=0,sex=factor(m,levels=c(m,f)))[rep(NA,20),] and of course the brute force way is data.frame(age=rep(as.numeric(NA),20), sex=factor(rep(NA,20),levels=c(m,f)) ) Also, (a) there's no idea in ensuring that you're filling with NA if they are all going to be changed anyway, and (b) recycling works so that you only need to specify the length of one variable, so data.frame(age=numeric(20), sex=factor(,levels=c(m,f)) ) works too. Extending a data frame can be as simple as mydata - mydata[1:newlength,] (plus fixup of row names later on). With many thanks again, David. Yes, once I've named the first column, I can add further ones by saying d[c(x,y,z)]=NA or such. I was just wondering whether that was the way to do it or whether there was a more elegant approach. Preallocation was the critical clue I needed. Use an initial data.frame call naming all the columns. -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Font problem
John Dougherty [EMAIL PROTECTED] writes: Some plots fail due to a problem with the X11 fonts. I get a message that X11 font at size 22 could not be loaded. The demo() graphics routine for instance dies during the third chart. The graphics demo calls font.main=1 and that seems to be where the error is. I believe this is due to a configuration problem on my system, however I can't find where in the environment font.main looks for the font to use. I am running SuSE 9.0 and use the KDE desktop. However, I have also replicated this in GNOME and WindowMaker. Varying the fonts used by the console does notb effect the result. I think that in principle the bug is in R, but as far as I remember, the workaround is to ensure that you either have scalable PostScript fonts or have non-scalable versions in both 100 and 75 dpi. This in turn is assured by configuring the font server. On RedHat (dunno about SuSE),/etc/X11/fs/config needs to have something like catalogue = /usr/X11R6/lib/X11/fonts/misc:unscaled, /usr/X11R6/lib/X11/fonts/75dpi:unscaled, /usr/X11R6/lib/X11/fonts/100dpi:unscaled, /usr/X11R6/lib/X11/fonts/misc, /usr/X11R6/lib/X11/fonts/Type1, ... or lose the :unscaled, but that tends to look horrible. -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] trouble with predict.l1ce
clayton == clayton springer [EMAIL PROTECTED] on Mon, 8 Dec 2003 13:48:11 -0500 writes: clayton Dear R-help, I am having trouble with the predict clayton function in lasso2. For example: data(Iowa) l1c.I - l1ce(Yield ~ ., Iowa, bound = 10, absolute.t=TRUE) predict (l1c.I) # this works is fine predict (l1c.I,Iowa) clayton Error in eval(exper,envir, enclos) : couldn't find clayton function Yield clayton And I have similar trouble whenever I use the clayton newdata argument in prediction. yes. This is something that needs to be added to the lasso2 package. Volunteers are sought ... Martin Maechler [EMAIL PROTECTED] http://stat.ethz.ch/~maechler/ Seminar fuer Statistik, ETH-Zentrum LEO C16Leonhardstr. 27 ETH (Federal Inst. Technology) 8092 Zurich SWITZERLAND phone: x-41-1-632-3408 fax: ...-1228 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R]: global and local variables
Hi all I have a problem pertaining to local and global variables. Say I have a function defined as follows: a-function(x) {yx^2} i.e a(2) [1] 4 function b is now defined to take the value of y and do some manipulation with it. As it stands I dont know how to store the variable y such that other functions are able to reference its value. I know that I can simply put the operations found in b simply into a but this is not what I want. I would like to have stand alone functions say a, b and c which could be run independently as well as have a function called say control that can run a, b and c. i.e. control- function( x) { a(x) b(x) c(x) } I hope that you guys understand what I'm trying to do. Cheers Allan __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R]: global and local variables
allan clark wrote: Hi all I have a problem pertaining to local and global variables. Say I have a function defined as follows: a-function(x) {yx^2} i.e a(2) [1] 4 The function a specified above won't return 4! function b is now defined to take the value of y and do some manipulation with it. As it stands I dont know how to store the variable y such that other functions are able to reference its value. I know that I can simply put the operations found in b simply into a but this is not what I want. I would like to have stand alone functions say a, b and c which could be run independently as well as have a function called say control that can run a, b and c. i.e. control- function( x) { a(x) b(x) c(x) } I hope that you guys understand what I'm trying to do. You are trying to read An Introduction to R??? If not, please try! What you are really going to do: using assigments and return() statements as in: a - function(x) return(x^2) foo - function(x) { y - a(x) z - b(x) return(list(y=y, z=z)) } foo(.) Uwe Ligges Cheers Allan __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] levelplot parameters
Hi, I have a levelplot with one panel. I just can't find out how I can manipulate the size of the axis lables. e.g. scales.cex doesn't work, the usual par-parameters either. Any hint? joerg __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] axes that meet
It might also mean that he is looking for par(xaxs = i, yaxs = i) Pikounis, Bill [EMAIL PROTECTED] Sent by: [EMAIL PROTECTED] 12/09/2003 05:23 AM To: 'Hank Stevens' [EMAIL PROTECTED], [EMAIL PROTECTED] cc: Subject:RE: [R] axes that meet Hank, I think the graphical parameter you are looking for is bty, as in par(bty=l) Details on the ?par help topic. Hope that helps. Bill Bill Pikounis, Ph.D. Biometrics Research Department Merck Research Laboratories PO Box 2000, MailDrop RY33-300 126 E. Lincoln Avenue Rahway, New Jersey 07065-0900 USA Phone: 732 594 3913 Fax: 732 594 1565 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Hank Stevens Sent: Tuesday, December 09, 2003 1:18 AM To: [EMAIL PROTECTED] Subject: [R] axes that meet R v. 1.7.1, Windows 2000. A particular journal wants me to provide scatter plots with no box, but with axes that meet in the lower left corner. It seems as though there must be an easy way of doing this, but my reading the help on plot.default, axis, and box have not provided any clues. I would be most appreciative of any feedback. Thank you, Hank Stevens Dr. Martin Henry H. Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/botany/bot/henry.html http://www.muohio.edu/ecology __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R]: global and local variables
Hi Thanx for those who responded to my problem. In my previous email I tried to ask a general question and probably never explained myself correctly. I wanted to prevent sending this long email. My apologies. This is my actual problem. I have a regression problem. I am writing some R code in order to calculate some collinearity diagnostics. The diagnostics all rely on a function named preprocess. I've written the different diagnostics as separate functions so that they may be evaluated separately if required. The two functions are named mci and vif. (I will be writing some others later) mci calculates the mixed condition index as well as the condition indices of a given X matrix while vif calculates the variance inflation factors of the X matrix. Another function named colldiag has been written. This function will calculate all of the collinearity diagnostics by simply calling the separate functions defined previously. I've attached the code of the different functions as well as a data file (say a2) below. The functions mci and vif work perfectly. i.e. mci(a2) [1] DATA MATRIX CENTERED AND SCALED [1] CENTERED AND SCALED MATRIX = $data [1] MEANS OF XDATA = $means [1] STDS OF XDATA = $stds [1] THE CONDITION NUMBER AND THE CONDITION INDICES $CN [1] 27.34412 $CI [1] 1.00 1.615690 27.344123 $MCI Principal.Component Singular.Values Condition.Index 1 1 1.47206801.00 2 2 0.91110781.615690 3 3 0.0538349 27.344123 vif(a2) [1] DATA MATRIX CENTERED AND SCALED [1] CENTERED AND SCALED MATRIX = $data [1] MEANS OF XDATA = $means [1] STDS OF XDATA = $stds [1] THE VARIANCE INFLATION FACTORS $vif x1 x2 x3 169.3542 175.6667 1.6875 The output from colldiag is as follows: colldiag(a2) [1] DATA MATRIX CENTERED AND SCALED [1] CENTERED AND SCALED MATRIX = $data [1] MEANS OF XDATA = $means [1] STDS OF XDATA = $stds [1] THE CONDITION NUMBER AND THE CONDITION INDICES $CN [1] 27.34412 $CI [1] 1.00 1.615690 27.344123 $MCI Principal.Component Singular.Values Condition.Index 1 1 1.47206801.00 2 2 0.91110781.615690 3 3 0.0538349 27.344123 [1] DATA MATRIX CENTERED AND SCALED [1] CENTERED AND SCALED MATRIX = $data [1] MEANS OF XDATA = $means [1] STDS OF XDATA = $stds [1] THE VARIANCE INFLATION FACTORS $vif x1 x2 x3 169.3542 175.6667 1.6875 Once you check the colldiag code below you will see that it calls mci and vif. In both of these functions they call preprocess. This is unnecessary. How can I write the code such that R only calls preprocess once? ONCE AGAIN I APOLOGIZE FOR THE LENGTH OF THIS EMAIL!!! Cheers Allan The data file: x1 x2 x3 1 20 -4 5 2 21 -4 4 3 22 -3 3 4 23 -2 2 5 24 -1 1 6 25 0 2 7 26 1 3 8 27 2 4 9 28 3 5 10 29 4 6 11 20 -4 5 12 21 -4 4 13 22 -3 3 14 23 -2 2 15 24 -1 1 16 25 0 2 17 26 1 3 18 27 2 4 19 28 3 5 20 29 4 6 preprocess-function (xdata,center=1,scale=1) { if(center==1 scale==1) { means-apply(xdata,2,mean) stds-apply(xdata,2, function(x) sqrt(var(x))) scalefactor-((nrow(xdata)-1)^.5)*stds data.centsca-sweep(sweep(xdata,2,means,-),2,scalefactor,/) print(DATA MATRIX CENTERED AND SCALED) print(CENTERED AND SCALED MATRIX = $data) print(MEANS OF XDATA = $means) print(STDS OF XDATA = $stds) list(data=data.centsca,means=means,stds=stds,prep=1) } else if(center==1 scale==0) { means-apply(xdata,2,mean) data.cen-sweep(xdata,2,means,-) print(DATA MATRIX CENTERED) list(data=data.cen,means=means,prep=1) } else if(center==0 scale==1) { stds-apply(xdata,2, function(x) sqrt(var(x))) scalefactor-((nrow(xdata)-1)^.5)*stds data.sca-sweep(xdata,2,scalefactor,/) print(DATA MATRIX SCALED) list(data=data.sca,stds=stds,prep=1) } else { print(YOU HAVE TO SPECIFY WHETHER YOU WANT TO SCALE OR CENTER THE MATRIX) print(THE preprocess FUNCTION HAS THREE ARGUMENTS. i.e. preprocess(xdata,center,scale)) print(xdata IS THE MATRIX TO BE TRANSFORMED) print(TO CENTER SPECIFY center=1) print(TO SCALE SPECIFY scale=1) } # A matrix is standardised as follows: # X*(i,j) = ( X(i,j)- XBAR(j) ) / ( sqrt(n-1)* STD(j) ) } mci-function (xdata) { a-preprocess(xdata) b-svd(a$data) cn-(b$d)[1]/(b$d)[ncol(a$data)] ci-(b$d)[1]/(b$d)[1:ncol(a$data)] #paste(THE CONDITION NUMBER = ,cn) #the following produces a table in order to output the mci values
Re: [R] p-value from chisq.test working strangely on 1.8.1
On Tue, 2003-12-09 at 02:23, Torsten Hothorn wrote: Hello everybody, I'm seeing some strange behavior on R 1.8.1 on Intel/Linux compiled with gcc 3.2.2. The p-value calculated from the chisq.test function is incorrect for some input values: chisq.test(matrix(c(0, 1, 1, 12555), 2, 2), simulate.p.value=TRUE) Pearson's Chi-squared test with simulated p-value (based on 2000 replicates) data: matrix(c(0, 1, 1, 12555), 2, 2) X-squared = 1e-04, df = NA, p-value = 1 chisq.test(matrix(c(0, 1, 1, 12556), 2, 2), simulate.p.value=TRUE) [...] data: matrix(c(0, 1, 1, 12556), 2, 2) X-squared = 1e-04, df = NA, p-value = 2.2e-16 chisq.test(matrix(c(0, 1, 1, 12557), 2, 2), simulate.p.value=TRUE) [...] data: matrix(c(0, 1, 1, 12557), 2, 2) X-squared = 1e-04, df = NA, p-value = 1 this does not happen with R-1.8.1 and gcc 2.95.4 on Debian stable: R chisq.test(matrix(c(0, 1, 1, 12555), 2, 2), simulate.p.value=TRUE)$p.value [1] 1 R chisq.test(matrix(c(0, 1, 1, 12556), 2, 2), simulate.p.value=TRUE)$p.value [1] 1 R chisq.test(matrix(c(0, 1, 1, 12557), 2, 2), simulate.p.value=TRUE)$p.value [1] 1 neither with R-1.9.0 (unstable). Is this reproducible without using `set.seed' on your system? Best, Torsten snip Confirmed on Fedora Core 1 with R Version 1.8.1 Patched (2003-12-07) compiled using gcc (GCC) 3.3.2 20031107 (Red Hat Linux 3.3.2-2). chisq.test(matrix(c(0, 1, 1, 12555), 2, 2), simulate.p.value=TRUE) ... X-squared = 1e-04, df = NA, p-value = 1 chisq.test(matrix(c(0, 1, 1, 12556), 2, 2), simulate.p.value=TRUE) X-squared = 1e-04, df = NA, p-value = 2.2e-16 ... chisq.test(matrix(c(0, 1, 1, 12557), 2, 2), simulate.p.value=TRUE) ... X-squared = 1e-04, df = NA, p-value = 1 HTH, Marc Schwartz __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] Importing TIFF files into a R matrix
Hi I am facing a problem where I would like to import a TIFF image (of spots on a nylon filter) into R (into a matrix for example). When plotting the matrix using fx. scatterplot3d I would then be able to see how the pixel-intensities are distributed in spot-areas on the filter - which would be very helpful. Does anynone know of a way to do this? Best regards, Peter Hagedorn ... Peter Hagedorn Risø National Laboratory Plant Research Department Building PRD-330 P.O. Box 49 Frederiksborgvej 399 DK-4000 Roskilde Denmark Phone +45 4677 4293 Fax +45 4677 4109 e-mail [EMAIL PROTECTED] web http://www.risoe.dk/pbk/staff_uk/phah.htm __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Font problem
On Tue, 2003-12-09 at 08:07, Peter Dalgaard wrote: Marc Schwartz [EMAIL PROTECTED] writes: snip Actually, I think you can leave the :unscaled on. Seems to work for me. Could be. That is, I believe, the default on a new install. It was on FC1 for me (I have not had font problems so far) and I recall the same default on RH 8.0 and RH 9. Though I do have recollections of font problems on 8.0, solved by adding the two additional lines without the ':unscaled', which is why this issue sticks in my mind. Then again, it may just be old age and the snow this morning... It may be as simple as being sure that both the 75 dpi and 100 dpi fonts are loaded as the SUSE query earlier this year seemed to indicate that the 100 dpi fonts were not installed initially. I believe that a restart of X may be required to make the change, but a restart of the X font server may suffice using: /sbin/service xfs restart Restarting xfs is certainly necessary (and, I believe, not implied by restarting X) but beware that it may freeze currently running X applications (that used to be the case anyway), so don't do it with important stuff running on your desktop. Good point :-) Marc __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] PROC MIXED vs. lme()
I'm trying to learn how to do a repeated measures ANOVA in R using lme(). A data set that comes from the book Design and Analysis has the following structure: Measurements (DV) were taken on 8 subjects (SUB) with two experimental levels (GROUP) at four times (TRIAL). In SAS, I use the code: PROC MIXED DATA=[data set below]; CLASS sub group trial; MODEL dv = group trial group*trial; REPEATED trial / SUBJECT=sub TYPE=CS; run; which gives the results: Tests of Fixed Effects SourceNDF DDF Type III F Pr F GROUP 1 62.51 0.1645 TRIAL 318 22.34 0.0001 GROUP*TRIAL 3180.58 0.6380 In R, I'm trying the code: results.cs - lme(DV ~ factor(GROUP)*factor(TRIAL), data=[data set below], random= ~factor(TRIAL)|SUB, correlation=corCompSymm() ) anova(results.cs) which gives the results: numDF denDF F-value p-value (Intercept) 118 3383.953 .0001 factor(GROUP) 1 64.887 0.0691 factor(TRIAL) 318 239.102 .0001 factor(GROUP):factor(TRIAL) 3181.283 0.3103 Why are these results different? I'm a newbie to R, have the book Mixed Effects Models in S and S-Plus, but can't seem to get this analysis to work. Any suggestions? Thanks! Manuel Data: SUB GROUP DV TRIAL 1 1 3 1 1 1 4 2 1 1 7 3 1 1 3 4 2 1 6 1 2 1 8 2 2 1 12 3 2 1 9 4 3 1 7 1 3 1 13 2 3 1 11 3 3 1 11 4 4 1 0 1 4 1 3 2 4 1 6 3 4 1 6 4 5 2 5 1 5 2 6 2 5 2 11 3 5 2 7 4 6 2 10 1 6 2 12 2 6 2 18 3 6 2 15 4 7 2 10 1 7 2 15 2 7 2 15 3 7 2 14 4 8 2 5 1 8 2 7 2 8 2 11 3 8 2 9 4 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] levelplot parameters
See the scales argument in ?xyplot, which has: scales: list determining how the x- and y-axes (tick marks and labels) are drawn. The list contains parameters in name=value form, and may also contain two other lists called 'x' and 'y' of the same form (described below). Components of 'x' and 'y' affect the respective axes only, while those in 'scales' affect both. (When parameters are specified in both lists, the values in 'x' or 'y' are used.) The components are : [...] cex: factor to control character sizes for axis labels. Can be a vector of length 2, to control left/bottom and right/top separately. HTH, Andy -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Joerg Schaber Sent: Tuesday, December 09, 2003 7:12 AM To: [EMAIL PROTECTED] Subject: [R] levelplot parameters Hi, I have a levelplot with one panel. I just can't find out how I can manipulate the size of the axis lables. e.g. scales.cex doesn't work, the usual par-parameters either. Any hint? joerg __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] Importing TIFF files into a R matrix
Hi, I do not know of any free TIFF readers for R, so I suggest that you use an external TIFF-to-Portable Pixmap coverter and then use the pixmap package available on CRAN. I recommend ImageMagick's convert program available for Unix, Linux, Windows, Windows/Cygwin etc at http://www.imagemagick.org/. (If you have one already installed, be careful not to work with an old version; I ran into a problem convert 16-bits TIFF with a 2 years old convert and it made it only into 8-bit images without warnings. That should not be a problem now.) The Portable Pixmap format includes i) RGB images (PPM), gray scale images (PGM) and monochrome images (PBM). In your case (I assume) you're working with 16-bits grayscale TIFF images so you should convert to PGM. If you have your PATH setup correctly an example would then be: library(pixmap) system(convert foo.tiff foo.pgm) img - read.pnm(foo.pgm) and then work from there. Hope this helps... Henrik Bengtsson Dept. of Mathematical Statistics @ Centre for Mathematical Sciences Lund Institute of Technology/Lund University, Sweden (Sweden +1h UTC, Melbourne +11 UTC, Calif. -8h UTC) +46 708 909208 (cell), +46 46 320 820 (home), +1 (508) 464 6644 (global fax), +46 46 2229611 (off), +46 46 2224623 (dept. fax) h b @ m a t h s . l t h . s e, http://www.maths.lth.se/~hb/ -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: den 9 december 2003 15:16 To: [EMAIL PROTECTED] Subject: [R] Importing TIFF files into a R matrix Hi I am facing a problem where I would like to import a TIFF image (of spots on a nylon filter) into R (into a matrix for example). When plotting the matrix using fx. scatterplot3d I would then be able to see how the pixel-intensities are distributed in spot-areas on the filter - which would be very helpful. Does anynone know of a way to do this? Best regards, Peter Hagedorn ... Peter Hagedorn Risø National Laboratory Plant Research Department Building PRD-330 P.O. Box 49 Frederiksborgvej 399 DK-4000 Roskilde Denmark Phone +45 4677 4293 Fax +45 4677 4109 e-mail [EMAIL PROTECTED] web http://www.risoe.dk/pbk/staff_uk/phah.htm __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailma n/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Font problem
Marc Schwartz [EMAIL PROTECTED] writes: It may be as simple as being sure that both the 75 dpi and 100 dpi fonts are loaded as the SUSE query earlier this year seemed to indicate that the 100 dpi fonts were not installed initially. I actually wrote the bug so I'm supposed to know what it is about It is just that it was so difficult to write that I have been reluctant to try and fix it. The exact issue is that the X11 driver jumps through a few hoops to help you get the real Adobe-designed fonts rather that some ugly rescaled ones. These exist in pixel sizes 8,10,11,12,14,17,18,20,24,25,34 (of which 10/11 and 24/25 are actually identical). If you use unscaled fonts, the logic inside RLoadFont will try to give you one of the above, by choosing the one closest to the one you specified. If you request a 22 pixel font, the system will load the 20 pixel font. The catch is that e.g. the 20 pixel font only exists in 100 dpi blueberry:~/xlsfonts -fn '-adobe-helvetica-medium-r-*-*-20-*-*-*-*-*-*-*' -adobe-helvetica-medium-r-normal--20-140-100-100-p-100-iso10646-1 -adobe-helvetica-medium-r-normal--20-140-100-100-p-100-iso8859-1 (actually, on my machine, only the 12 pixel version exists in both 100 and 75 dpi). I.e. if you use unscaled fonts, you need to have both dpi sets installed. With scaled fonts you don't run into this issue, but some screen fonts may (will!) be ugly. Of course the default SuSE install gives you only one dpi set, unscaled. The bug is that there is a gap in the font-substitution logic so that if you have only one set of fonts, then RLoadFont may end up returning NULL, rather than the fixed font or a suitable fallback size. -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Interfacing R and Python in MS Windows
Uwe Ligges wrote: Héctor Villafuerte D. wrote: Hi all, I need the power of R from within some of my Python programs... I use debian linux (woody) at home and windows XP at work (the latter is where I need to get things done!) This are my packages: R 1.8.0 Python 2.3 RSPython 0.5-3 This is what I've done: (1) Since the Windows Binary of RSPython is compiled against Python 2.2 I downloaded the tarball (2) Followed the instructions in INSTALL.win (with pexports and everything) (3) In the RGUI Install package(s) from local zip files... If you have downloaded the tarball of RSPython, you have to install the package using Rcmd INSTALL, and you cannot use the RGUI which is designed to install binary packages. Uwe Ligges (4) NO errors reported during this process (5) When I try to Load package in R it show this error: local({pkg - select.list(sort(.packages(all.available = TRUE))) + if(nchar(pkg)) library(pkg, character.only=TRUE)}) Error in testRversion(descfields) : This package has not been installed properly See the Note in ?library (6) In Python import RS Traceback (most recent call last): File stdin, line 1, in ? ImportError: No module named RS Please help me to get this excelent tools going on in Windows. Thanks in advance, Hector Thanks for your help guys. Tim: I had already seen RPy; but its Windows binary is compiled against Python 2.2 (and I have 2.3) so it didn't work. Uwe: (1) I installed Active Perl (it seems to be needed by Rcmd INSTALL) (2) I then created a tar.gz with the modifications found in INSTALL.WIN (3) Here's what I got: E:\to_doRcmd INSTALL -c e:/to_do/RSPython.tar.gz -- Making package RSPython ** WARNING: this package has a configure script It probably needs manual configuration ** installing inst files A package must contain a DESCRIPTION file make[1]: *** [frontmatter] Error 27 make: *** [pkg-RSPython] Error 2 *** Installation of RSPython failed *** make --no-print-directory DLLNM= RHOME=C:/PROGRA~1/R/rw1080 BUILD=MINGW \ -C E:/to_do/R.INSTALL/RSPython PKG=RSPython -f C:/PROGRA~1/R/rw1080/src/gnuwin32/MakePkg clean make: *** [pkgclean-RSPython] Error 255 What's a description file? Any suggestions compiling this? Thanks again in advance, Hector __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] axes that meet
Thank you to all who replied. par(bty) and par(xaxs, yaxs) gave me all I need (and more!) for specifiying axes in the fashion required. Thanks, Hank At 07:28 AM 12/9/2003, [EMAIL PROTECTED] wrote: It might also mean that he is looking for par(xaxs = i, yaxs = i) Pikounis, Bill [EMAIL PROTECTED] Sent by: [EMAIL PROTECTED] 12/09/2003 05:23 AM To: 'Hank Stevens' [EMAIL PROTECTED], [EMAIL PROTECTED] cc: Subject:RE: [R] axes that meet Hank, I think the graphical parameter you are looking for is bty, as in par(bty=l) Details on the ?par help topic. Hope that helps. Bill Bill Pikounis, Ph.D. Biometrics Research Department Merck Research Laboratories PO Box 2000, MailDrop RY33-300 126 E. Lincoln Avenue Rahway, New Jersey 07065-0900 USA Phone: 732 594 3913 Fax: 732 594 1565 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Hank Stevens Sent: Tuesday, December 09, 2003 1:18 AM To: [EMAIL PROTECTED] Subject: [R] axes that meet R v. 1.7.1, Windows 2000. A particular journal wants me to provide scatter plots with no box, but with axes that meet in the lower left corner. It seems as though there must be an easy way of doing this, but my reading the help on plot.default, axis, and box have not provided any clues. I would be most appreciative of any feedback. Thank you, Hank Stevens Dr. Martin Henry H. Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/botany/bot/henry.html http://www.muohio.edu/ecology __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. Dr. Martin Henry H. Stevens, Assistant Professor 338 Pearson Hall Botany Department Miami University Oxford, OH 45056 Office: (513) 529-4206 Lab: (513) 529-4262 FAX: (513) 529-4243 http://www.cas.muohio.edu/botany/bot/henry.html http://www.muohio.edu/ecology __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Design functions after Multiple Imputation
On Mon, 08 Dec 2003 21:23:40 + Umberto Maggiore [EMAIL PROTECTED] wrote: I am a new user of R for Windows, enthusiast about the many functions of the Design and Hmisc libraries. I combined the results of a Cox regression model after multiple imputation(of missing values in some covariates). Now I got my vector of coefficients (and of standard errors). My question is: How could I use directly that vector to run programs such as 'nomogram', 'calibrate', 'validate.cph' which, in contrast, call for the saved results form 'cph' ? I did not use 'aregImpute' for multiple imputation. However, even if I did it, 'fit.mult.impute' seems not to allow specifying the option 'surv=TRUE' (essential to get a nomogram) or 'x=TRUE, y=TRUE' (which are essential for 'calibrate' and 'validate.cph'. Therefore, I dont't see how I could get a nomogram or run other Design functions after 'aregImpute'. thank you so much in advance Umberto Good questions Umberto. The Design package handles multiple imputation and model validation, but currently not at the same time. But model descriptions such as those provided by summary.Design, nomogram.Design, contrast.Design are fully operational after the variance-covariance matrix is corrected for multiple imputation by fit.mult.impute. There is one caveat. When the nomogram not only has predicted relative hazards but absolute survival estimates (probabilities, quantiles, restricted mean life), fit.mult.impute gets the baseline survival estimates from surv=T from the first imputation. We need to extend that to average baseline survival estimates over all multiple imputations. You should have had no difficulty using nomogram after fit.mult.impute (after aregImpute); only fully trust the relative hazard estimates from the resulting nomogram. I have put out a note on the IMPUTE e-mail list (see http://hesweb1.med.virginia.edu/biostat/rms for subscription information) about how to develop a reasonable algorithm for simultaneous multiple imputation and bootstrap model validation/calibration. So far no takers. Frank --- Frank E Harrell JrProfessor and ChairSchool of Medicine Department of BiostatisticsVanderbilt University __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] p-value from chisq.test working strangely on 1.8.1
It happens with suse 9.0 as well. ... chisq.test(matrix(c(0, 1, 1, 12556), 2, 2), simulate.p.value=TRUE) Pearson's Chi-squared test with simulated p-value (based on 2000 replicates) data: matrix(c(0, 1, 1, 12556), 2, 2) X-squared = 1e-04, df = NA, p-value = 5e-04 . JWDougherty __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] FW: Symposium COMPSTAT 2004
Dear all, I just received the following message, and I think it might be of interest to the R-list. Cordially, Roland - all people interested in COMPSTAT 2004 Symposium Prague December 8, 2004 Dear colleague, thank you very much for your interest in COMPSTAT 2004 Symposium. This is to remind you that the deadline for delivering your manuscript in electronic form to the hands of the editors is approaching very fast. Due to a series of question to us concerning the selection of papers for proceedings, we repeat (especially for those who participated to the previous COMPSTAT's) that the choice will be done based on complete papers and not on abstracts as in the past. Submission of Papers for inclusion in the Proceedings Volume (subject to selection by SPC) is February 2, 2004 !!! For details see: http://www.compstat2004.cuni.cz Please, be in time. Thanks a lot in advance Looking forward seeing you next year for COMPSTAT 2004 in Prague We remain sincerely yours LOC + editors Please, forgive us if you have received this e-mail more than ones due to a duplication in different databases. PS Most frequently asked questions. 1) I HAVE PARTICIPATED TO SEVERAL PREVIOUS COMPSTAT SYMPOSIA. WHAT IS THE MOST IMPORTANT ORGANIZATIONAL CHANGE? During the previous COMPSTAT's the selection of papers and their division as short and standard contributed papers has been based on the submitted two page long abstract. This scheme has been abandoned for COMPSTAT 2004. Instead, the participants are asked to submit the full papers (up to 8 pages) not just an abstract. The SPC will make a selection of the best of these for inclusion in the published Proceedings Volume. These will be reviewed and subject to revision before publication. The other papers that are deemed to be acceptable will be, after reviewing, included on a CD that will form the integral part of the Proceedings Volume. The time given for the presentation may reflect the status of the paper as assessed by the SPC - the allocation will depend on the time pressures in putting together the final programme. In the case that it will not be possible to schedule all oral contributions, SPC has the right to only offer a poster presentation to some of them. 2) WHY SHOULD I SUBMIT ALSO THE ABSTRACT IF MY PAPER WILL BE AVAILABLE DURING THE SYMPOSIUM? The abstracts should inform the participants about the content of his/her lecture. All abstract will be included in the Abstracts Volume produced by the LOC, distributed during the symposium and mounted before the symposium on its web page. Notice also that not all people have with them always a notebook to check abstracts on the CD. Both SPC and LOC hope it will be more easy for the participants to orient oneself and to chose the session where to go. Latest day for the authors by which abstracts of submitted contributions and invited papers must be delivered to the LOC in Prague in electronic format is May 1, 2004!!! 3) WHAT IS THE LATEST DAY FOR DELIVERING MANUSCRIPTS? Latest day for the authors by which complete manuscripts of submitted contributions and invited papers must be delivered to the LOC in Prague in electronic format is February 2, 2004!!! + This mail has been sent through the MPI for Demographic Research. Should you receive a mail that is apparently from a MPI user without this text displayed, then the address has most likely been faked. If you are uncertain about the validity of this message, please check the mail header or ask your system administrator for assistance. __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] problem with pls(x, y, ..., ncomp = 16): Error in inherits(x, data.frame) : subscript out of bounds
When I try to use ncomp parameter in pls procedure I get following error: library(pls.pcr) m - pls(x, y, validation = CV, niter = 68, ncomp = 16) Error in inherits(x, data.frame) : subscript out of bounds Without ncomp parameter everything seems to work OK dim(x) [1] 68 116 dim(y) [1] 68 1 m - pls(x, y, validation = CV, niter = 68) length(m$ncomp) [1] 67 Ryszard __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] p-value from chisq.test working strangely on 1.8.1
On Dec 9, 2003, at 9:37 AM, Peter Dalgaard wrote: Marc Schwartz [EMAIL PROTECTED] writes: Confirmed on Fedora Core 1 with R Version 1.8.1 Patched (2003-12-07) compiled using gcc (GCC) 3.3.2 20031107 (Red Hat Linux 3.3.2-2). chisq.test(matrix(c(0, 1, 1, 12555), 2, 2), simulate.p.value=TRUE) ... X-squared = 1e-04, df = NA, p-value = 1 chisq.test(matrix(c(0, 1, 1, 12556), 2, 2), simulate.p.value=TRUE) X-squared = 1e-04, df = NA, p-value = 2.2e-16 ... chisq.test(matrix(c(0, 1, 1, 12557), 2, 2), simulate.p.value=TRUE) ... X-squared = 1e-04, df = NA, p-value = 1 Ditto on RH8 with Martyn's RPM of 1.8.0 (yeah, I know...) and ditto with a reasonably current r-devel (gcc 3.2) Anyways, it is yet another fudge-factor issue: If you debug to the point in chisq.test where it calculates PVAL - sum(tmp$results = STATISTIC)/B you'll find that Browse[1] any(diff(tmp$result)) [1] FALSE Browse[1] tmp$result[1] [1] 7.96432e-05 Browse[1] STATISTIC [1] 7.96432e-05 Browse[1] tmp$result[1] - STATISTIC [1] -1.355253e-20 so PVAL becomes zero and yaddayaddayadda The obvious fix would seem to be PVAL - sum(tmp$results = (1-1e-10)*STATISTIC)/B Yes, this is also the behavior that I am seeing. I do not know about numerical programming to comment on the fix, but it would solve the problem in my case. Thanks to everyone who has looked into this! Jeff __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] Matrix to dates
z - matrix( c(1960,1960,1961,1,9,6), 3, 2 ) 1. Using chron: require(chron) chron( paste( z[,2], 1, z[,1], sep=/ ) ) 2. Using POSIXct: ISOdate( z[,1], z[,2], 1 ) # relative to GMT time zone or ISOdate( z[,1], z[,2], 1, tz= ) # relative to current time zone --- Date: Mon, 8 Dec 2003 15:48:20 -0600 From: Erin Hodgess [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: [R] Matrix to dates Let's try again! I have a matrix in which the first column is a four digit year, and the second column is a 2 digit month. How do I convert the matrix to a date function, please? Thanks, Erin Version 1.8.0 mailto: [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] R: the spdep package
On Tue, 9 Dec 2003, Osmo Kolehmainen wrote: Hi, Here is a listw object z corresponding to the matrix W. I understand n, nn, S0, S1 and S2 in the weights constants summary. Is it simply so that n1 = n-1, n2 = n-2 and n3 = n-3? If this is true where they are needed? Yes, used in calculating the variance of the statistics (Moran, Geary) elsewhere, are returned by spweights.constants(). Will be hidden in summary.listw() in next release. Roger Just wondering Osmo W [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [1,] 0 1 0 1 0 0 0 0 0 [2,] 1 0 1 0 1 0 0 0 0 [3,] 0 1 0 0 0 1 0 0 0 [4,] 1 0 0 0 1 0 1 0 0 [5,] 0 1 0 1 0 1 0 1 0 [6,] 0 0 1 0 1 0 0 0 1 [7,] 0 0 0 1 0 0 0 1 0 [8,] 0 0 0 0 1 0 1 0 1 [9,] 0 0 0 0 0 1 0 1 0 z=mat2listw(W);z Characteristics of weights list object: Neighbour list object: Number of regions: 9 Number of nonzero links: 24 Percentage nonzero weights: 29.62963 Average number of links: 2.67 Weights style: M Weights constants summary: n n1 n2 n3 nn S0 S1 S2 M 9 8 7 6 81 24 48 272 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help -- 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://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Windows Memory Issues
I would also like some clarification about R memory management. Like Doug, I didn't find anything about consecutive calls to gc() to free more memory. We run into memory limit problems every now and then and a better understanding of R's memory management would go a long way. I am interested in learning more and was wondering if there is any specific R documentation that explains R's memory usage? Or maybe some good links about memory and garbage collection. Thanks. Benjamin Stabler Transportation Planning Analysis Unit Oregon Department of Transportation 555 13th Street NE, Suite 2 Salem, OR 97301 Ph: 503-986-4104 --- Message: 21 Date: Mon, 8 Dec 2003 09:51:12 -0800 (PST) From: Douglas Grove [EMAIL PROTECTED] Subject: Re: [R] Windows Memory Issues To: Prof Brian Ripley [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] Content-Type: TEXT/PLAIN; charset=US-ASCII On Sat, 6 Dec 2003, Prof Brian Ripley wrote: I think you misunderstand how R uses memory. gc() does not free up all the memory used for the objects it frees, and repeated calls will free more. Don't speculate about how memory management works: do your homework! Are you saying that consecutive calls to gc() will free more memory than a single call, or am I misunderstanding? Reading ?gc and ?Memory I don't see anything about this mentioned. Where should I be looking to find more comprehensive info on R's memory management?? I'm not writing any packages, just would like to have a better handle on efficiently using memory as it is usually the limiting factor with R. FYI, I'm running R1.8.1 and RedHat9 on a P4 with 2GB of RAM in case there is any platform specific info that may be applicable. Thanks, Doug Grove Statistical Research Associate Fred Hutchinson Cancer Research Center In any case, you are using an outdated version of R, and your first course of action should be to compile up R-devel and try that, as there has been improvements to memory management under Windows. You could also try compiling using the native malloc (and that *is* described in the INSTALL file) as that has different compromises. On Sat, 6 Dec 2003, Richard Pugh wrote: Hi all, I am currently building an application based on R 1.7.1 (+ compiled C/C++ code + MySql + VB). I am building this application to work on 2 different platforms (Windows XP Professional (500mb memory) and Windows NT 4.0 with service pack 6 (1gb memory)). This is a very memory intensive application performing sophisticated operations on large matrices (typically 5000x1500 matrices). I have run into some issues regarding the way R handles its memory, especially on NT. In particular, R does not seem able to recollect some of the memory used following the creation and manipulation of large data objects. For example, I have a function which receives a (large) numeric matrix, matches against more data (maybe imported from MySql) and returns a large list structure for further analysis. A typical call may look like this . myInputData - matrix(sample(1:100, 750, T), nrow=5000) myPortfolio - createPortfolio(myInputData) It seems I can only repeat this code process 2/3 times before I have to restart R (to get the memory back). I use the same object names (myInputData and myPortfolio) each time, so I am not create more large objects .. I think the problems I have are illustrated with the following example from a small R session . # Memory usage for Rui process = 19,800 testData - matrix(rnorm(1000), 1000) # Create big matrix # Memory usage for Rgui process = 254,550k rm(testData) # Memory usage for Rgui process = 254,550k gc() used (Mb) gc trigger (Mb) Ncells 369277 9.9 667722 17.9 Vcells 87650 0.7 24286664 185.3 # Memory usage for Rgui process = 20,200k In the above code, R cannot recollect all memory used, so the memory usage increases from 19.8k to 20.2. However, the following example is more typical of the environments I use . # Memory 128,100k myTestData - matrix(rnorm(1000), 1000) # Memory 357,272k rm(myTestData) # Memory 357,272k gc() used (Mb) gc trigger (Mb) Ncells 478197 12.8 818163 21.9 Vcells 9309525 71.1 31670210 241.7 # Memory 279,152k Here, the memory usage increases from 128.1k to 279.1k Could anyone point out what I could do to rectify this (if anything), or generally what strategy I could take to improve this? Many thanks, Rich. Mango Solutions Tel : (01628) 418134 Mob : (07967) 808091 [[alternative HTML version deleted]] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied
Re: [R] How to append to a data.frame?
Dear Peter Dalgaard, Thank you for these examples, they are very neat! I really like the data.frame(x=as.numeric(NA),y=factor(NA))[rep(NA,1000),] trick. With best regards, David. Dr David Philip Kreil (`-''-/).___..--''`-._ Research Fellow`6_ 6 ) `-. ( ).`-.__.`) University of Cambridge(_Y_.)' ._ ) `._ `. ``-..-' ++44 1223 764107, fax 333992 _..`--'_..-_/ /--'_.' ,' www.inference.phy.cam.ac.uk/dpk20 (il),-'' (li),' ((!.-' __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Windows Memory Issues
On Tue, 9 Dec 2003 [EMAIL PROTECTED] wrote: I would also like some clarification about R memory management. Like Doug, I didn't find anything about consecutive calls to gc() to free more memory. It was a statement about Windows, and about freeing memory *to Windows*. Douglas Grove apparently had misread both the subject line and the sentence. We run into memory limit problems every now and then and a better understanding of R's memory management would go a long way. I am interested in learning more and was wondering if there is any specific R documentation that explains R's memory usage? Or maybe some good links about memory and garbage collection. Thanks. There are lots of comments in the source files. And as I already said (but has been excised below), this is not relevant to the next version of R anyway. BTW, the message below has been selectively edited, so please consult the original. Message: 21 Date: Mon, 8 Dec 2003 09:51:12 -0800 (PST) From: Douglas Grove [EMAIL PROTECTED] Subject: Re: [R] Windows Memory Issues To: Prof Brian Ripley [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] Content-Type: TEXT/PLAIN; charset=US-ASCII On Sat, 6 Dec 2003, Prof Brian Ripley wrote: I think you misunderstand how R uses memory. gc() does not free up all the memory used for the objects it frees, and repeated calls will free more. Don't speculate about how memory management works: do your homework! Are you saying that consecutive calls to gc() will free more memory than a single call, or am I misunderstanding? Reading ?gc and ?Memory I don't see anything about this mentioned. Where should I be looking to find more comprehensive info on R's memory management?? I'm not writing any packages, just would like to have a better handle on efficiently using memory as it is usually the limiting factor with R. FYI, I'm running R1.8.1 and RedHat9 on a P4 with 2GB of RAM in case there is any platform specific info that may be applicable. Thanks, Doug Grove Statistical Research Associate Fred Hutchinson Cancer Research Center In any case, you are using an outdated version of R, and your first course of action should be to compile up R-devel and try that, as there has been improvements to memory management under Windows. You could also try compiling using the native malloc (and that *is* described in the INSTALL file) as that has different compromises. On Sat, 6 Dec 2003, Richard Pugh wrote: Hi all, I am currently building an application based on R 1.7.1 (+ compiled C/C++ code + MySql + VB). I am building this application to work on 2 different platforms (Windows XP Professional (500mb memory) and Windows NT 4.0 with service pack 6 (1gb memory)). This is a very memory intensive application performing sophisticated operations on large matrices (typically 5000x1500 matrices). I have run into some issues regarding the way R handles its memory, especially on NT. In particular, R does not seem able to recollect some of the memory used following the creation and manipulation of large data objects. For example, I have a function which receives a (large) numeric matrix, matches against more data (maybe imported from MySql) and returns a large list structure for further analysis. A typical call may look like this . myInputData - matrix(sample(1:100, 750, T), nrow=5000) myPortfolio - createPortfolio(myInputData) It seems I can only repeat this code process 2/3 times before I have to restart R (to get the memory back). I use the same object names (myInputData and myPortfolio) each time, so I am not create more large objects .. I think the problems I have are illustrated with the following example from a small R session . # Memory usage for Rui process = 19,800 testData - matrix(rnorm(1000), 1000) # Create big matrix # Memory usage for Rgui process = 254,550k rm(testData) # Memory usage for Rgui process = 254,550k gc() used (Mb) gc trigger (Mb) Ncells 369277 9.9 667722 17.9 Vcells 87650 0.7 24286664 185.3 # Memory usage for Rgui process = 20,200k In the above code, R cannot recollect all memory used, so the memory usage increases from 19.8k to 20.2. However, the following example is more typical of the environments I use . # Memory 128,100k myTestData - matrix(rnorm(1000), 1000) # Memory 357,272k rm(myTestData) # Memory 357,272k gc() used (Mb) gc trigger (Mb) Ncells 478197 12.8 818163 21.9 Vcells 9309525 71.1 31670210 241.7 # Memory 279,152k Here, the memory usage increases from 128.1k to 279.1k Could anyone point out what I could do to rectify this (if anything), or generally what strategy I could take to improve this? Many thanks, Rich.
RE: [R] Windows Memory Issues
[snipped] Or maybe some good links about memory and garbage collection. As is mentioned time-to-time on this list when the above subject comes up, Windows memory is a complicated topic. One open-source utility I have found helpful to monitor memory when I work under XP is called RAMpage, authored by John Fitzgibbon, and is available at http://www.jfitz.com/software/RAMpage/ In its FAQ / Help, it touches on a lot of general memory and resource issues, which I found helpful to learn about. http://www.jfitz.com/software/RAMpage/RAMpage_FAQS.html (Though the author clearly warns that its usefulness for freeing memory may not be anymore than cosmetic on NT / 2000 / XP systems.) Hope that helps. Bill Bill Pikounis, Ph.D. Biometrics Research Department Merck Research Laboratories PO Box 2000, MailDrop RY33-300 126 E. Lincoln Avenue Rahway, New Jersey 07065-0900 USA Phone: 732 594 3913 Fax: 732 594 1565 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Tuesday, December 09, 2003 12:09 PM To: [EMAIL PROTECTED] Subject: Re: [R] Windows Memory Issues I would also like some clarification about R memory management. Like Doug, I didn't find anything about consecutive calls to gc() to free more memory. We run into memory limit problems every now and then and a better understanding of R's memory management would go a long way. I am interested in learning more and was wondering if there is any specific R documentation that explains R's memory usage? Or maybe some good links about memory and garbage collection. Thanks. Benjamin Stabler Transportation Planning Analysis Unit Oregon Department of Transportation 555 13th Street NE, Suite 2 Salem, OR 97301 Ph: 503-986-4104 --- Message: 21 Date: Mon, 8 Dec 2003 09:51:12 -0800 (PST) From: Douglas Grove [EMAIL PROTECTED] Subject: Re: [R] Windows Memory Issues To: Prof Brian Ripley [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] Content-Type: TEXT/PLAIN; charset=US-ASCII On Sat, 6 Dec 2003, Prof Brian Ripley wrote: I think you misunderstand how R uses memory. gc() does not free up all the memory used for the objects it frees, and repeated calls will free more. Don't speculate about how memory management works: do your homework! Are you saying that consecutive calls to gc() will free more memory than a single call, or am I misunderstanding? Reading ?gc and ?Memory I don't see anything about this mentioned. Where should I be looking to find more comprehensive info on R's memory management?? I'm not writing any packages, just would like to have a better handle on efficiently using memory as it is usually the limiting factor with R. FYI, I'm running R1.8.1 and RedHat9 on a P4 with 2GB of RAM in case there is any platform specific info that may be applicable. Thanks, Doug Grove Statistical Research Associate Fred Hutchinson Cancer Research Center In any case, you are using an outdated version of R, and your first course of action should be to compile up R-devel and try that, as there has been improvements to memory management under Windows. You could also try compiling using the native malloc (and that *is* described in the INSTALL file) as that has different compromises. On Sat, 6 Dec 2003, Richard Pugh wrote: Hi all, I am currently building an application based on R 1.7.1 (+ compiled C/C++ code + MySql + VB). I am building this application to work on 2 different platforms (Windows XP Professional (500mb memory) and Windows NT 4.0 with service pack 6 (1gb memory)). This is a very memory intensive application performing sophisticated operations on large matrices (typically 5000x1500 matrices). I have run into some issues regarding the way R handles its memory, especially on NT. In particular, R does not seem able to recollect some of the memory used following the creation and manipulation of large data objects. For example, I have a function which receives a (large) numeric matrix, matches against more data (maybe imported from MySql) and returns a large list structure for further analysis. A typical call may look like this . myInputData - matrix(sample(1:100, 750, T), nrow=5000) myPortfolio - createPortfolio(myInputData) It seems I can only repeat this code process 2/3 times before I have to restart R (to get the memory back). I use the same object names (myInputData and myPortfolio) each time, so I am not create more large objects .. I think the problems I have are illustrated with the following example from a small R session . # Memory usage for Rui process = 19,800 testData - matrix(rnorm(1000), 1000) #
[R] Contour plots
I am drawing several contour plots in one page. I use image, and get several contours. But I don't know how to control the color in more than one plots. For example, same color corresponds to different dependent values in different plots. For example, yellow means z=100(the highest value) in plot1, but yellow means z=40(the highest value) in plot2--this is not what I want. In fact,I want the same color corresponds the same depedent value in each plot. Say, in all plots, yellow means z=100,red means z=40, etc. I was wondering if someone can tell me which function I can use. Because it seems to me that image function doesn't have such option to control that. Thank you. - Qinghua Song Department of Statistics UW-Madison office phone:262-8181 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Contour plots
I am drawing several contour plots in one page. I use image, and get several contours. But I don't know how to control the color in more than one plots. For example, same color corresponds to different dependent values in different plots. For example, yellow means z=100(the highest value) in plot1, but yellow means z=40(the highest value) in plot2--this is not what I want. In fact,I want the same color corresponds the same depedent value in each plot. Say, in all plots, yellow means z=100,red means z=40, etc. I was wondering if someone can tell me which function I can use. Because it seems to me that image function doesn't have such option to control that. Thank you. I think you need to use the zlim argument in image(). For example image(x,y,z,zlim=c(0,100)) will make sure that all the plots are drawn using the same values for z (between 0 and 100 here). Bill __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] R Interface handholding
Hello, I need a bit of handholding with R, specifically, with writing packages for it. I'm a systems programmer, and am, on the request of several users of our software, working on generating R interfaces. For starters, I've written the following R function (which compiles): SEXP myincr(SEXP Rinput) { // Returns input integer incremented by one int input; SEXP returner; PROTECT(Rinput = AS_NUMERIC(Rinput)); input = * INTEGER(Rinput); input++; PROTECT(returner = NEW_INTEGER(input)); Rprintf(Hey there\n); return returner; } I've made this into a package, by dropping it into a stub directory along with something called init.c: #include areone.h #include R_ext/Rdynload.h #include Rinternals.h R_NativePrimitiveArgType myincr_t[1] = {INTSXP}; static const R_CMethodDef cMethods[] = { {myincr, (DL_FUNC) myincr, 1, myincr_t} }; void R_init_myincr(DllInfo* dll) { R_registerRoutines(dll, cMethods, NULL, NULL, NULL); } R is happy to install this for me, but after doing a library(myincr), the function doesn't seem to be available, so I presume I'm missing something. Does R normally call, at library load, R_init_$MODULENAME() ? My other question is.. our software produces data structures (we call datsets) which resemble limited database tables, and I'd like some advice on exposing them to R -- columns, in our scheme, either hold doubles or strings, the columns have names, and we need the ability to export these into appropriate R structures as well as populate them from R. I notice that the R DBI, at least as according to its documentation, uses the database to hold the data (presumably in temporary tables) and returns parts of it as requested, via R functions. I could use a static global pointer (in C) into a storage space of datsets, and write bridge functions exporting them as R arrays, or I could attempt to find an appropriate native format and export to it.. any advice? It might be worth mentioning that all of the code to do this will eventually be auto-generated -- we already have code to do this for C that doesn't need to expose people linking our code to all of our custom structures and stuff. -- Pat Gunn Research/Systems Programmer, Auton Group, CMU __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] Windows Memory Issues
Thanks for the reply. So are you saying that multiple calls to gc() frees up memory to Windows and then other processes can use that newly freed memory? So multiple calls to gc() does not actually make more memory available to new R objects that I might create. The reason I ask is because I want to know how to use all the available memory that I can to store object in R. ?gc says that garbage collection is run without user intervention so there is really nothing I can do to improve memory under Windows except increase the --max-mem-size at startup (which is limited to 1.7GB under the current version of R for Windows and will be greater ~3GB for R 1.9). Thanks again. Ben Stabler -Original Message- From: Prof Brian Ripley [mailto:[EMAIL PROTECTED] Sent: Tuesday, December 09, 2003 9:29 AM To: STABLER Benjamin Cc: [EMAIL PROTECTED] Subject: Re: [R] Windows Memory Issues On Tue, 9 Dec 2003 [EMAIL PROTECTED] wrote: I would also like some clarification about R memory management. Like Doug, I didn't find anything about consecutive calls to gc() to free more memory. It was a statement about Windows, and about freeing memory *to Windows*. Douglas Grove apparently had misread both the subject line and the sentence. We run into memory limit problems every now and then and a better understanding of R's memory management would go a long way. I am interested in learning more and was wondering if there is any specific R documentation that explains R's memory usage? Or maybe some good links about memory and garbage collection. Thanks. There are lots of comments in the source files. And as I already said (but has been excised below), this is not relevant to the next version of R anyway. BTW, the message below has been selectively edited, so please consult the original. Message: 21 Date: Mon, 8 Dec 2003 09:51:12 -0800 (PST) From: Douglas Grove [EMAIL PROTECTED] Subject: Re: [R] Windows Memory Issues To: Prof Brian Ripley [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] Content-Type: TEXT/PLAIN; charset=US-ASCII On Sat, 6 Dec 2003, Prof Brian Ripley wrote: I think you misunderstand how R uses memory. gc() does not free up all the memory used for the objects it frees, and repeated calls will free more. Don't speculate about how memory management works: do your homework! Are you saying that consecutive calls to gc() will free more memory than a single call, or am I misunderstanding? Reading ?gc and ?Memory I don't see anything about this mentioned. Where should I be looking to find more comprehensive info on R's memory management?? I'm not writing any packages, just would like to have a better handle on efficiently using memory as it is usually the limiting factor with R. FYI, I'm running R1.8.1 and RedHat9 on a P4 with 2GB of RAM in case there is any platform specific info that may be applicable. Thanks, Doug Grove Statistical Research Associate Fred Hutchinson Cancer Research Center In any case, you are using an outdated version of R, and your first course of action should be to compile up R-devel and try that, as there has been improvements to memory management under Windows. You could also try compiling using the native malloc (and that *is* described in the INSTALL file) as that has different compromises. On Sat, 6 Dec 2003, Richard Pugh wrote: Hi all, I am currently building an application based on R 1.7.1 (+ compiled C/C++ code + MySql + VB). I am building this application to work on 2 different platforms (Windows XP Professional (500mb memory) and Windows NT 4.0 with service pack 6 (1gb memory)). This is a very memory intensive application performing sophisticated operations on large matrices (typically 5000x1500 matrices). I have run into some issues regarding the way R handles its memory, especially on NT. In particular, R does not seem able to recollect some of the memory used following the creation and manipulation of large data objects. For example, I have a function which receives a (large) numeric matrix, matches against more data (maybe imported from MySql) and returns a large list structure for further analysis. A typical call may look like this . myInputData - matrix(sample(1:100, 750, T), nrow=5000) myPortfolio - createPortfolio(myInputData) It seems I can only repeat this code process 2/3 times before I have to restart R (to get the memory back). I use the same object names (myInputData and myPortfolio) each time, so I am not create more large objects .. I think the problems I have are illustrated with the following example from a small R session . # Memory usage for Rui process = 19,800 testData - matrix(rnorm(1000), 1000) # Create big matrix # Memory usage for Rgui process = 254,550k rm(testData) #
[R] C++: SET_LENGTH() Over Many Iterations?
In a C++ extension to R (v 1.8.1), I've been experimenting with a generic push back function to tack one value at a time onto the end of an R vector created within the extension. After calling this function a certain number of times Rgui.exe (I'm writing in Windows using Visual Studio .NET 2003) will fail with an Access Violation, which doesn't happen when I pre-allocate the R-vector memory and write to the reserved slots; i.e., I'm not trying to create an R object too big to be handled by R within the context of my OS's available memory. Here's some simple test code I've been running: CPP Code #define PUSH_BACK_INTEGER(v, x) \ do {\ UNPROTECT_PTR(v);\ SET_LENGTH(v, GET_LENGTH(v) + 1);\ PROTECT(v);\ INTEGER_POINTER(v)[GET_LENGTH(v) - 1] = x;\ }\ while (false) SEXP R_SimplePushBackTest(SEXP args) { SEXP arg1, arg2, int_vect; PROTECT(arg1 = AS_INTEGER(CADR(args))); int n_reps = INTEGER_POINTER(arg1)[0]; PROTECT(arg2 = AS_LOGICAL(CADDR(args))); bool full_alloc = (LOGICAL_POINTER(arg2)[0] ? true : false); if (full_alloc) PROTECT(int_vect = NEW_INTEGER(n_reps)); else PROTECT(int_vect = NEW_INTEGER(0)); for (int i = 0; i n_reps; ++i) { Rprintf( ** Iteration %d:\n, i + 1); if (full_alloc) INTEGER_POINTER(int_vect)[i] = i; else PUSH_BACK_INTEGER(int_vect, i); } SEXP out, names, cls; PROTECT(out = NEW_LIST(1)); SET_VECTOR_ELT(out, 0, int_vect); PROTECT(names = NEW_CHARACTER(1)); SET_STRING_ELT(names, 0, COPY_TO_USER_STRING(integer.vector)); SET_NAMES(out, names); PROTECT(cls = NEW_CHARACTER(1)); SET_STRING_ELT(cls, 0, COPY_TO_USER_STRING(pushback)); classgets(out, cls); UNPROTECT(6); return out; } /CPP Code R Code nreps=5 allocate=FALSE sink(pushback_test.txt) test.pushback=.External(R_SimplePushBackTest, as.integer(nreps), as.logical(allocate)) print(test.pushback) sink() /R Code If allocate=TRUE (vector memory is pre-allocated in the extension), the code proceeds normally on my system; if allocate=FALSE, Rgui.exe eventually crashes from an Access Violation. I've gathered only enough information from the R source code so far to think that R is ultimately calling realloc() through the SET_LENGTH macro; that would make my code rather inefficient, but I'm trying for genericness here. In C++, is there a better way than what I'm doing to concatenate values onto the end of an R vector of arbitrary length, especially over many iterations? Thanks for taking the time to read this. System specs: Pentium 4 2.5 GHz, 512 MB RAM, 40 GB hard drive, Win XP Yrs etc., Jim Java __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R]: global and local variables
Hi! I'm no guru in R. But I can think of 2 ways (have to be tried): 1) As Uwe Ligges said: just save return the stored variable a-preprocess(xdata) and return it (if you want to return more than 1 item use list) and give this variable to the next function. Example: func1-function(x) { y-x^2 # for testing: print(paste(func1, x, y)) return(y) } func2-function(x, valuefunc1=NA) { if (is.na(valuefunc1)) valuefunc1-func1(x) # calculate things with valuefunc1 print(paste(func2, x, valuefunc1)) return(list(valuefunc1=valuefunc1)) } func3-function(x, valuefunc1=NA) { if (is.na(valuefunc1)) valuefunc1-func1(x) # calculate things with valuefunc1 print(paste(func3, x, valuefunc1)) return(list(valuefunc1=valuefunc1)) } # use this a-list(NA) names(a)-c(valuefunc1) a-func2(x=3, valuefunc1=a$valuefunc1) a-func3(x=3, valuefunc1=a$valuefunc1) # be careful, makes only sense if the x is equal to former fuction call... a-func3(x=999, valuefunc1=a$valuefunc1) # will be the same a-func3(x=999) # will be different 2) use global variables like assign(stored.value, x, envir=.GlobalEnv) example see reply of my posting: RE: [R] LOCF - Last Observation Carried Forward Simon Fear (Sat 15 Nov 2003 - 03:28:03 EST) HTH, Karl __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] problem with pls(x, y, ..., ncomp = 16): Error in inherit s( x, data.frame) : subscript out of bounds
Help for pls says: ?pls [...] Arguments: ... ncomp: the numbers of latent variables to be assessed in the modelling. Default is from one to the rank of 'X'. [...] so my assumption was (maybe wrong) that idea of ncomp parameter is to limit number of assessed variables. also, if called without crossvalidation it gives the same error: m - pls(x, y, ncomp = 16) Error in inherits(x, data.frame) : subscript out of bounds R Liaw, Andy [EMAIL PROTECTED] 12/09/2003 12:50 PM To: Ryszard Czerminski/PH/[EMAIL PROTECTED], [EMAIL PROTECTED] cc: Subject:RE: [R] problem with pls(x, y, ..., ncomp = 16): Error in inherit s( x, data.frame) : subscript out of bounds I don't know the details of pls (in the pls.pcr package, I assume), but if you use validation=CV, that says you want to use CV to select the best number of components. Then why would you specify ncomp as well? Andy From: [EMAIL PROTECTED] When I try to use ncomp parameter in pls procedure I get following error: library(pls.pcr) m - pls(x, y, validation = CV, niter = 68, ncomp = 16) Error in inherits(x, data.frame) : subscript out of bounds Without ncomp parameter everything seems to work OK dim(x) [1] 68 116 dim(y) [1] 68 1 m - pls(x, y, validation = CV, niter = 68) length(m$ncomp) [1] 67 Ryszard __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Interfacing R and Python in MS Windows
I got farther this time, but it wasn't enough though... does anyone know what this errors are? Thanks! Hector E:\to_doRcmd INSTALL e:/to_do/RSPython.tar.gz -- Making package RSPython ** WARNING: this package has a configure script It probably needs manual configuration ** installing inst files adding build stamp to DESCRIPTION making DLL ... making GeneralConverters.d from GeneralConverters.c making PythonCall.d from PythonCall.c making PythonFunctionConverters.d from PythonFunctionConverters.c making PythonReferences.d from PythonReferences.c making PythonReflectance.d from PythonReflectance.c making RCall.d from RCall.c making RPythonConverters.d from RPythonConverters.c making RPythonReferences.d from RPythonReferences.c making UserConverters.d from UserConverters.c gcc -I../inst/include -Ic:/Python23/include -D_R_=1 -DUSE_R=1 -IC:/PROGRA~1/R/rw1080/src/include -Wall -O2 -c General Converters.c -o GeneralConverters.o In file included from c:/Python23/include/Python.h:75, from ../inst/include/RPythonModule.h:4, from ../inst/include/UserConverters.h:4, from GeneralConverters.c:1: c:/Python23/include/intobject.h:41: parse error before PyInt_AsUnsignedLongLongMask c:/Python23/include/intobject.h:41: warning: type defaults to `int' in declaration of `PyInt_AsUnsignedLongLongMask' c:/Python23/include/intobject.h:41: warning: data definition has no type or storage class In file included from c:/Python23/include/Python.h:77, from ../inst/include/RPythonModule.h:4, from ../inst/include/UserConverters.h:4, from GeneralConverters.c:1: c:/Python23/include/longobject.h:37: warning: parameter names (without types) in function declaration c:/Python23/include/longobject.h:39: parse error before PyLong_AsLongLong c:/Python23/include/longobject.h:39: warning: type defaults to `int' in declaration of `PyLong_AsLongLong' c:/Python23/include/longobject.h:39: warning: data definition has no type or storage class c:/Python23/include/longobject.h:40: parse error before PyLong_AsUnsignedLongLong c:/Python23/include/longobject.h:40: warning: type defaults to `int' in declaration of `PyLong_AsUnsignedLongLong' c:/Python23/include/longobject.h:40: warning: data definition has no type or storage class c:/Python23/include/longobject.h:41: parse error before PyLong_AsUnsignedLongLongMask c:/Python23/include/longobject.h:41: warning: type defaults to `int' in declaration of `PyLong_AsUnsignedLongLongMask' c:/Python23/include/longobject.h:41: warning: data definition has no type or storage class make[2]: *** [GeneralConverters.o] Error 1 make[1]: *** [srcDynlib] Error 2 make: *** [pkg-RSPython] Error 2 *** Installation of RSPython failed *** Liaw, Andy wrote: Check the Writing R Extensions manual for requirements on the DESCRIPTION file. HTH, Andy __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] Windows Memory Issues
On Tue, 9 Dec 2003 [EMAIL PROTECTED] wrote: Thanks for the reply. So are you saying that multiple calls to gc() frees up memory to Windows and then other processes can use that newly freed memory? No. You typically can't free memory back to Windows (or many other OSes). So multiple calls to gc() does not actually make more memory available to new R objects that I might create. Yes and no. It makes more memory available, but only memory that would have been made available in any case if you had tried to use it. R calls the garbage collector before requesting more memory from the operating system and before running out of memory. The reason I ask is because I want to know how to use all the available memory that I can to store object in R. ?gc says that garbage collection is run without user intervention so there is really nothing I can do to improve memory under Windows except increase the --max-mem-size at startup You can't do anything else to make more memory available, only to use less. -thomas __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] Windows Memory Issues
On Tue, 9 Dec 2003 [EMAIL PROTECTED] wrote: Thanks for the reply. So are you saying that multiple calls to gc() frees up memory to Windows and then other processes can use that newly freed memory? So multiple calls to gc() does not actually make more memory That is what I said. Why do people expect me to repeat myself? available to new R objects that I might create. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] Windows Memory Issues
On Tue, 9 Dec 2003, Thomas Lumley wrote: On Tue, 9 Dec 2003 [EMAIL PROTECTED] wrote: Thanks for the reply. So are you saying that multiple calls to gc() frees up memory to Windows and then other processes can use that newly freed memory? No. You typically can't free memory back to Windows (or many other OSes). At least using R under Windows NT/2000/XP you can. I've watched it do so whilst fixing memory leaks. There is another complication here: R for Windows uses a third-party malloc, and you can free memory back to that if not to the OS. The reason Windows is special is the issue of fragmentation, which OSes using mmap (and R-devel under Windows) typically do not suffer. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Interfacing R and Python in MS Windows
I believe RSPython is from Omegahat, so why not ask on the mailing list for Omegahat? On Tue, 9 Dec 2003, Héctor Villafuerte D. wrote: I got farther this time, but it wasn't enough though... does anyone know what this errors are? -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] packages for ecologists
Hello R-user, sorry for this very off-topic question. But I shall present R to my dept. (pro's and con's and what it can do). The pro's and con's are easy but not what R can do (additional to the normal statistics). I looked through the packages, but the enormous amount of packages makes it very difficult for me to decide which one is worth mentioning. I used only a small part of all R packages (mainly recommended packages and grasper) and would like to know which package for ecologist has to be mentioned. I would greatly appreciate if you can tell me which packages you think are very useful for ecolgical research in R e.g. vegan, ade4, ... thanks in advance, regards Martin __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] packages for ecologists
Definitely take a look at WiSP. http://www.ruwpa.st-and.ac.uk/estimating.abundance/WiSP/ Mark == Mark Herzog Post Doctoral Researcher Dept. of Natural Resources and Environmental Science University of Nevada Reno Reno, NV 89512 (775) 784-6984 (office) (775) 784-4583 (fax) [EMAIL PROTECTED] == -Original Message- From: Martin Wegmann [mailto:[EMAIL PROTECTED] Sent: Tuesday, December 09, 2003 12:26 PM To: R-list Subject: [R] packages for ecologists Hello R-user, sorry for this very off-topic question. But I shall present R to my dept. (pro's and con's and what it can do). The pro's and con's are easy but not what R can do (additional to the normal statistics). I looked through the packages, but the enormous amount of packages makes it very difficult for me to decide which one is worth mentioning. I used only a small part of all R packages (mainly recommended packages and grasper) and would like to know which package for ecologist has to be mentioned. I would greatly appreciate if you can tell me which packages you think are very useful for ecolgical research in R e.g. vegan, ade4, ... thanks in advance, regards Martin __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] histogram density division
Hi, I already sent this query but I didn't receive any answers. I try another time and try to explain another way my question... I want to decrease the height of my bands in my histogram by a factor of 5... how can I do that? Is there an argument to hist() that do the job? Thanks, Mathieu Hi, I would like to plot an histogram with modified density values. My Y-axis represent the occurences of the ranges (specified by the breaks argument) of my data (represented by a big vector). Now, I would like to divide by X the number of occurences in each bands and do a plot of lower occurences. How can I do that? Thanks, Mathieu __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] Johnson's Sb distribution
Hi all; I'm working with the library SuppDists trying to fit a Johnson's Sb distribution to a dataset. It works fine, but I need to set one of the location parameters (epsilon) to zero. How can I do this using the function JohnsonFit() or any other similar? ...and Is it possible to define the type (SN,SL,SB,SU) or the library assumes the type automatically depending on the data? Thanks for any hint Christian __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] histogram density division
Mathieu - That's easy. Assign the return value of hist() to some variable, say fixed, then go in and hack the value of fixed$counts however you like, and re-plot using plot(fixed). Example code: fixed - hist(rnorm(2000)) fixed$counts - fixed$counts / 5 plot(fixed) I confess I didn't quite understand your question the first time I saw it, so couldn't reply. - tom blackwell - u michigan medical school - ann arbor - On Tue, 9 Dec 2003, Mathieu Drapeau wrote: Hi, I already sent this query but I didn't receive any answers. I try another time and try to explain another way my question... I want to decrease the height of my bands in my histogram by a factor of 5... how can I do that? Is there an argument to hist() that do the job? Thanks, Mathieu Hi, I would like to plot an histogram with modified density values. My Y-axis represent the occurences of the ranges (specified by the breaks argument) of my data (represented by a big vector). Now, I would like to divide by X the number of occurences in each bands and do a plot of lower occurences. How can I do that? Thanks, Mathieu __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] R Interface handholding
I'm not sure exactly what you want to do but it seems you are going about it in an overly complicated way. If you are going to build a package you just need to put a NAMESPACE file at the top level and include a useDynLib specification or write a short R function called .First.Lib that calls library.dynam and include it in the R subdirectory. Typically this function is simply .First.lib = function(lib, pkg) { library.dynam(pkg, pkg, lib) } but it can do other initializations. After that you just put the source for myincr in the src subdirectory, compile and load the package, then call the function as .Call(myincr, x, PACKAGE=myPackageName) Pat Gunn [EMAIL PROTECTED] writes: Hello, I need a bit of handholding with R, specifically, with writing packages for it. I'm a systems programmer, and am, on the request of several users of our software, working on generating R interfaces. For starters, I've written the following R function (which compiles): SEXP myincr(SEXP Rinput) { // Returns input integer incremented by one int input; SEXP returner; PROTECT(Rinput = AS_NUMERIC(Rinput)); input = * INTEGER(Rinput); input++; PROTECT(returner = NEW_INTEGER(input)); Rprintf(Hey there\n); return returner; } I've made this into a package, by dropping it into a stub directory along with something called init.c: #include areone.h #include R_ext/Rdynload.h #include Rinternals.h R_NativePrimitiveArgType myincr_t[1] = {INTSXP}; static const R_CMethodDef cMethods[] = { {myincr, (DL_FUNC) myincr, 1, myincr_t} }; void R_init_myincr(DllInfo* dll) { R_registerRoutines(dll, cMethods, NULL, NULL, NULL); } R is happy to install this for me, but after doing a library(myincr), the function doesn't seem to be available, so I presume I'm missing something. Does R normally call, at library load, R_init_$MODULENAME() ? My other question is.. our software produces data structures (we call datsets) which resemble limited database tables, and I'd like some advice on exposing them to R -- columns, in our scheme, either hold doubles or strings, the columns have names, and we need the ability to export these into appropriate R structures as well as populate them from R. I notice that the R DBI, at least as according to its documentation, uses the database to hold the data (presumably in temporary tables) and returns parts of it as requested, via R functions. I could use a static global pointer (in C) into a storage space of datsets, and write bridge functions exporting them as R arrays, or I could attempt to find an appropriate native format and export to it.. any advice? It might be worth mentioning that all of the code to do this will eventually be auto-generated -- we already have code to do this for C that doesn't need to expose people linking our code to all of our custom structures and stuff. -- Pat Gunn Research/Systems Programmer, Auton Group, CMU __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help -- Douglas Bates[EMAIL PROTECTED] Statistics Department608/262-2598 University of Wisconsin - Madisonhttp://www.stat.wisc.edu/~bates/ __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] Key for custom lattice panel function
Hi all, I've created a custom lattice panel function for levelplot - instead of representing z by colour, it plot circles with radius proportional to z (in the style of the map plots of Jacques Bertin). I'm happy to email an example graph to anyone interested. The problem is now to create a key for the plot. This is difficult because all of the other lattice plots convey information through point colour, shape, and texture, not point size. For this reason (and having looked at the code) I don't think I can shoehorn $key or $colorkey to produce the type of key that I want. I'm using grid.circles() with native units that have been scaled in the same way as in panel.levelplot. Can anyone offer any suggestions as to how to create this key? Hadley __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] OT: BibTex year-only citation in text?
Sorry for the off-topic question, but I know there are some talented LaTeX users out there. Which bibliography style gives only the year in text citations (e.g for further details, see Anderson (1992) )? Thanks Jason -- Indigo Industrial Controls Ltd. http://www.indigoindustrial.co.nz 64-21-343-545 [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] Key for custom lattice panel function
On Tuesday 09 December 2003 16:37, Hadley Wickham wrote: Hi all, I've created a custom lattice panel function for levelplot - instead of representing z by colour, it plot circles with radius proportional to z (in the style of the map plots of Jacques Bertin). I'm happy to email an example graph to anyone interested. The problem is now to create a key for the plot. This is difficult because all of the other lattice plots convey information through point colour, shape, and texture, not point size. For this reason (and having looked at the code) I don't think I can shoehorn $key or $colorkey to produce the type of key that I want. I'm using grid.circles() with native units that have been scaled in the same way as in panel.levelplot. Can anyone offer any suggestions as to how to create this key? Exactly how do you want to your key to look like ? As things stand now, I don't think it would be possible to put arbitrary keys. However, this should be easy to fix. That is, we could allow the key to be an arbitrary grid object (as produced by draw.key and draw.colorkey), and use its grobheight or grobwidth to allocate the necessary space. I'll try to add this in the next major update. For now, you could fake it by printing your trellis object in a grid viewport that's smaller than the whole screen, and draw your key separately afterwards in the remaining space. HTH, Deepayan __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] nested analysis with lme - odd result?
Hello! When I simulate variance at only a single level in a nested analysis using lme (all levels are random effects), the results confuse me. Instead of lme reporting high variance in only that simulated level, substantial variance (10% of simulated level) often appears in other levels -- in some configurations, as often as 50% of the time. Usually this spurious variance shows up in levels of nesting above the level with high simulated variance. Am I doing something wrong or should I expect such 'over-detection'? I was expecting near zero variance in those other levels except say, 5% of the time. Here's some example code with variance simulated in level 2 and spurious variance appearing frequently in level 1: library(nlme) # 4 level nesting simVar - c(0,1,0,.1) # first is level one, last becomes error nAtLevel - c(5,5,5,5) # number of replicates at each level F1V - F2V - F3V - residV - numeric(0) for (rep in 1:100){ F1f - F2f - F3f - value - numeric(0) mn - numeric(length(nAtLevel)) # data generator for (F1 in 1:nAtLevel[1]) { mn[1] - rnorm(1,sd=simVar[1]) # set mean for level 1 for (F2 in 1:nAtLevel[2]) { mn[2] - rnorm(1,sd=simVar[2]) # set mean for level 2 for (F3 in 1:nAtLevel[3]) { mn[3] - rnorm(1,sd=simVar[3]) # set mean for level 3 for (F4 in 1:nAtLevel[4]) { mn[4] - rnorm(1,sd=simVar[4]) #set mean for lowest level value - c(value, sum(mn)) F1f - c(F1f,F1) F2f - c(F2f,F2) F3f - c(F3f,F3) } } } } y.lme - lme(value ~ 1,random = ~ 1 | as.factor(F1f)/ as.factor(F2f)/ as.factor(F3f)) v - as.numeric(VarCorr(y.lme)[,2]) v - as.numeric(na.omit(v)) F1V - c(F1V,v[1]) F2V - c(F2V,v[2]) F3V - c(F3V,v[3]) residV - c(residV,y.lme$sigma) } var.df - data.frame(F1V,F2V,F3V,residV) par(mfrow=c(2,2)) hist(var.df$F1V,main='Level 1 StdDev') hist(var.df$F2V,main='Level 2 StdDev') hist(var.df$F3V,main='Level 3 StdDev') hist(var.df$residV,main='Residual StdDev') txt - 'Simulated stddev:' for (i in 1:length(simVar)) { txt - paste(txt,' level ',i,'=',simVar[i],',',sep='')} mtext(txt, outer=T,line=-1,side=3) Summary plots for several sample sizes is at: http://david.science.oregonstate.edu/~allisong/R/sampSize.pdf Thanks for any suggestions! Gary -- Gary Allison Department of Evolution, Ecology and Organismal Biology Ohio State University __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] expressing functions
# Why does expressing one function require(ctest) t.test # return only function (x, ...) UseMethod(t.test) environment: namespace:ctest # but expressing another function shapiro.test # returns more complete code? function (x) { DNAME - deparse(substitute(x)) x - sort(x[complete.cases(x)]) n - length(x) if (n 3 || n 5000) stop(sample size must be between 3 and 5000) rng - x[n] - x[1] if (rng == 0) stop(all `x[]' are identical) if (rng 1e-10) x - x/rng n2 - n%/%2 sw - .C(swilk, init = FALSE, as.single(x), n, n1 = as.integer(n), as.integer(n2), a = single(n2), w = double(1), pw = double(1), ifault = integer(1), PACKAGE = ctest) if (sw$ifault sw$ifault != 7) stop(paste(ifault=, sw$ifault, . This should not happen)) RVAL - list(statistic = c(W = sw$w), p.value = sw$pw, method = Shapiro-Wilk normality test, data.name = DNAME) class(RVAL) - htest return(RVAL) } environment: namespace:ctest -- Richard E. Remington III Statistician KERN Statistical Services, Inc. PO Box 1046 Boise, ID 83701 Tel: 208.426.0113 KernStat.com __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] expressing functions
On Tue, 9 Dec 2003, Remington, Richard wrote: # Why does expressing one function require(ctest) t.test # return only function (x, ...) UseMethod(t.test) environment: namespace:ctest # but expressing another function shapiro.test # returns more complete code? [...] False hypothesis: both are the complete code. You are not understanding (S3-style) generic functions: see any good book on R/S. (`An Introduction to R' is based on notes that predate them, but they are covered in more detail in the draft R Language manual. `The reader is referred to the official references for a complete discussion of this mechanism.' which I think means Chambers Hastie, 1992,) -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] expressing functions
Remington, Richard wrote: # Why does expressing one function require(ctest) t.test # return only function (x, ...) UseMethod(t.test) environment: namespace:ctest # but expressing another function shapiro.test # returns more complete code? function (x) { DNAME - deparse(substitute(x)) x - sort(x[complete.cases(x)]) n - length(x) if (n 3 || n 5000) stop(sample size must be between 3 and 5000) ... Short answer: Unless you're programming your own functions, you don't need to worry about that. Long answer: Because the first is generic - it looks at what kind of data you're testing (two vectors, a formula, whatever, ...) and calls the appropriate sub-function. shapiro.test does not; it just takes one data format, and stops in its tracks if that's not what you've provided. The ideas behind this are documented in Writing R Extensions (R-exts.pdf) which is supplied with binary R distributions, and is available from CRAN. See chapter 6, Generic functions and methods, in the version that accompanies R-1.8.1. Cheers Jason -- Indigo Industrial Controls Ltd. http://www.indigoindustrial.co.nz 64-21-343-545 [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: [R] OT: BibTex year-only citation in text?
Jason, For many bibliography styles, the command \citeyear{key} will work. If this doesn't work for the style you are using, you can investigate style-specific methods or consider other styles. I find that natbib is good for author-year formats. If you use Latex more than on occasion, a good reference book is invaluable. I like The Latex Companion by Goossens, Mittelbach, and Samarin. Regards, Matt Matthew R. Nelson, Ph.D. Director, Biostatistics Sequenom, Inc. -Original Message- From: Jason Turner [mailto:[EMAIL PROTECTED] Sent: Tuesday, December 09, 2003 4:20 PM To: R-Help Subject: [R] OT: BibTex year-only citation in text? Sorry for the off-topic question, but I know there are some talented LaTeX users out there. Which bibliography style gives only the year in text citations (e.g for further details, see Anderson (1992) )? Thanks Jason -- Indigo Industrial Controls Ltd. http://www.indigoindustrial.co.nz 64-21-343-545 [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: [R] expressing functions
Perhaps what should be added to the previous answers is that you can find out where the real work is done like this: require(ctest) t.test methods(t.test) ctest:::t.test.default ctest:::t.test.formula If the class of the first argument to t.test is formula then t.test.formula gets invoked so that's where the real work is done; otherwise, t.test.default gets invoked so that's where the real work is done. --- Remington, Richard wrote: # Why does expressing one function require(ctest) t.test # return only function (x, ...) UseMethod(t.test) environment: namespace:ctest # but expressing another function shapiro.test # returns more complete code? function (x) { DNAME - deparse(substitute(x)) x - sort(x[complete.cases(x)]) n - length(x) if (n 3 || n 5000) stop(sample size must be between 3 and 5000) __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
[R] pvalues
dear all- If I have a vector of numbers (not necessarily normally distributed) how can I get the p-value of a number in this distribution. I am interested in the inverse of 'quantile' . thank you- Maya __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help