Re: [R] How to increase memory for R on Soliars 10 with 16GB and 64bit R
Dongseok Choi wrote: Thank you very much for your help!! Now, it runs without any problem. Is it going to be fixed in the next release? Of course, Brian Thanks again, Dongseok Dongseok Choi, Ph.D. Assistant Professor Division of Biostatistics Department of Public Health Preventive Medicine Oregon Health Science University 3181 SW Sam Jackson Park Road, CB-669 Portland, OR 97239-3098 TEL) 503-494-5336 FAX) 503-494-4981 [EMAIL PROTECTED] Prof Brian Ripley [EMAIL PROTECTED] 07/13/05 12:03 AM On Tue, 12 Jul 2005, Dongseok Choi wrote: My machine is SUN Java Workstation 2100 with 2 AMD Opteron CPUs and 16GB RAM. R is compiled as 64bit by using SUN compilers. I trying to fit quantile smoothing on my data and I got an message as below. fit1-rqss(z1~qss(cbind(x,y),lambda=la1),tau=t1) Error in as.matrix.csr(diag(n)) : cannot allocate memory block of size 2496135168 The lengths of vector x and y are both 17664. I tried and found that the same command ran with x[1:16008] and y[1:16008]. So, it looks to me a memory related problem, but I'm not sure how I can allocate memory block. I read the command line option but not sure what do to with it. Could you help me on this? It is trying to allocate a single memory block of size over 2^31-1 bytes. R internally uses ints for sizes of vectors and that is a limit (see help(Memory-limits) ). However, it is intended that on 64-bit systems that there is a limit here of 8*(2^31-1) but there was a typo. Please change line 1534 of src/main/memory.c to #if SIZEOF_LONG 4 and re-compile. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] How to increase memory for R on Soliars 10 with 16GB and 64bit R
Uwe Ligges wrote: Dongseok Choi wrote: Thank you very much for your help!! Now, it runs without any problem. Is it going to be fixed in the next release? Of course, Brian [hmmm, looks like some wrong shortcut has been used - and it must have been me who forgot to drink coffee before strating to post ...] I meant to say: Of course, Brian already did, see the svn log: r34947 | ripley | 2005-07-13 03:16:18 -0400 (Wed, 13 Jul 2005) | 1 line Changed paths: M /trunk/NEWS M /trunk/src/main/memory.c correct typo for R_alloc on 64-bit systems Uwe Ligges Thanks again, Dongseok Dongseok Choi, Ph.D. Assistant Professor Division of Biostatistics Department of Public Health Preventive Medicine Oregon Health Science University 3181 SW Sam Jackson Park Road, CB-669 Portland, OR 97239-3098 TEL) 503-494-5336 FAX) 503-494-4981 [EMAIL PROTECTED] Prof Brian Ripley [EMAIL PROTECTED] 07/13/05 12:03 AM On Tue, 12 Jul 2005, Dongseok Choi wrote: My machine is SUN Java Workstation 2100 with 2 AMD Opteron CPUs and 16GB RAM. R is compiled as 64bit by using SUN compilers. I trying to fit quantile smoothing on my data and I got an message as below. fit1-rqss(z1~qss(cbind(x,y),lambda=la1),tau=t1) Error in as.matrix.csr(diag(n)) : cannot allocate memory block of size 2496135168 The lengths of vector x and y are both 17664. I tried and found that the same command ran with x[1:16008] and y[1:16008]. So, it looks to me a memory related problem, but I'm not sure how I can allocate memory block. I read the command line option but not sure what do to with it. Could you help me on this? It is trying to allocate a single memory block of size over 2^31-1 bytes. R internally uses ints for sizes of vectors and that is a limit (see help(Memory-limits) ). However, it is intended that on 64-bit systems that there is a limit here of 8*(2^31-1) but there was a typo. Please change line 1534 of src/main/memory.c to #if SIZEOF_LONG 4 and re-compile. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] texture in barplots?
On Thursday 14 July 2005 00:51, Peter Dalgaard wrote: Adrian Dusa [EMAIL PROTECTED] writes: ...snip... This comes up every now and then, and while it seems that everyone thinks fill patterns would be nice to have, I suspect that every attempt to actually implement it have gotten killed in infancy. The thing that is tricky to design right is the cross-device issues. Only some devices support this at all, and when they do, the patterns tend to be device dependent too. Probably not impossible -- there are other bits of the device drivers that deal with missing capabilities, like string rotation and clipping -- just, well, tricky. All clear then; I'll try to find some colors that are different even in BW, or maybe some grays. Um... Romania, I suppose? What city? It's Bucharest; I have a different signature in English but I do forget to use it when sending e-mails abroad (bad habit). Thank you, Adrian -- Adrian Dusa Romanian Social Data Archive University of Bucharest, Romania 1, Schitu Magureanu Bd. Tel./Fax: +40 21 3126618 \ +40 21 3120210 / int.101 -- This message was scanned for spam and viruses by BitDefender. For more information please visit http://linux.bitdefender.com/ __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Coxph with factors
Hello, I am fitting a coxph model with factors. I am running into problems when using 'survfit'. I am unsure how R is treating the factors when I fit, say: DATA-data.frame(time.sec,done,f.pom=factor(f.pom),po,vo) final-coxph(Surv(time.sec,done)~f.pom*vo+po,data=DATA) final.surv-survfit((final), individual=T,conf.type=log-log) print(final.surv) Also, when I trying using survfit with presepecified values for the covariates and factor (i.e. '0' to get the baseline) I get an error message, which I think is a result of my inability to specify the factors in the correct manner. I would very much appreciate if somebody can help me out. Thanks very much, Kylie __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Efficient testing for +ve definiteness
On Wed, 13 Jul 2005, Makram Talih wrote: Dear R-users, Is there a preferred method for testing whether a real symmetric matrix is positive definite? [modulo machine rounding errors.] The obvious way of computing eigenvalues via E - eigen(A, symmetric=T, only.values=T)$values and returning the result of !any(E = 0) seems less efficient than going through the LU decomposition invoked in determinant.matrix(A) and checking the sign and (log) modulus of the determinant. I suppose this has to do with the underlying C routines. Any thoughts or anecdotes? It has to do with what exactly you want to test. Knowing the determinant does not tell you if the matrix is close to non-positive definite or not. For numerical work, a comparison of the smallest eigenvalue to the largest is usually the most useful indication of possible problems in computations. An alternative is to try a Choleski decomposition, which may be faster but is less informative. Given how fast eigenvalues can be computed by current algorithms (and note the comments in ?eigen) I would suggest not worrying about speed until you need to (probably never). -- 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 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] High resolution plots
Thanks for the pointer! .wmf is far superior, I was just in the dark about the format and R's ability to produce it (An Introduction to R Device drivers does not mention it and I had obviously missed the deciding last two words in ?device 'windows') Thanks again. -Luis Gabor Grothendieck wrote: On 7/13/05, *Luis Tercero* [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: Dear R-help community, would any of you have a (preferably simple) example of a presentation-quality .png plot, i.e. one that looks like the .eps plots generated by R? I am working with R 2.0.1 in WindowsXP and am having similar problems as Knut Krueger in printing high-quality plots. I have looked at the help file and examples therein as well as others I have been able to find online but to no avail. After many many tries I have to concede I cannot figure it out. I would be very grateful for your help. If you want the highest resolution use a vector format, not a bitmapped format such as png. See: http://maths.newcastle.edu.au/~rking/R/help/04/02/1168.html http://maths.newcastle.edu.au/%7Erking/R/help/04/02/1168.html for some background. [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Pearson dispersion statistic
Thank you for your reply. I am aware of the good reasons not to use the deviance estimate in binomial, Poisson, and gamma families. However, for the inverse Gaussian, the choice seems to me less clear cut. So I just wanted to compare two different options. I have used the dispersion parameter to compute the standardized deviance residuals: summary(model.gamma)$deviance.resid /(summary(model.gamma)$dispersion * (1 - hatvalues(model.gamma)))^0.5 I noticed differences with Genstat which outputs these stand. dev. residuals directly, and they are explained by the automatic use of deviance instead of Pearson. Kind regards, Robin Smit -Original Message- From: Prof Brian Ripley [mailto:[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] ] Sent: dinsdag 12 juli 2005 15:12 To: Smit, R. (Robin) Cc: r-help@stat.math.ethz.ch Subject: : Re: [R] Dispersion in glm (was (no subject)) Actually, glm() does not estimate the dispersion at all, so you will need to be more specific. For example, summary.glm() and predict.glm() use the Pearson statistic if dispersion=NULL (the default) for most families. You can supply any other value you choose, and the MASS package makes use of this for ML estimation of the dispersion parameter (related to the shape) of the gamma family. There are rather good reasons (serious bias) not to use the deviance estimate in the binomial and Poisson families (see the example plots in MASS4), and good reasons not to use either in the gamma family. As the Pearson and deviance estimates agree for the gaussian, that does leave begging the question of why you want to do this. Further, McCullagh Nelder have general arguments why the Pearson estimate might always be preferred to the deviance one. So that `another statastical package' appears to need justification for its choice. On Mon, 11 Jul 2005, Smit, R. (Robin) wrote: The estimate of glm dispersion can be based on the deviance or on the Pearson statistic. I have compared output from R glm() to another statastical package and it appears that R uses the Pearson statistic. I was wondering if it is possible to make use R the deviance instead by modifying the glm(...) function? Thanks for your attention. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ 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, UK Fax: +44 1865 272595 This e-mail and its contents are subject to the DISCLAIMER at http://www.tno.nl/disclaimer/email.html [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Name for factor's levels with contr.sum
One way to do this generally is to make a copy of contr.sum, rename it, and set the dimnames appropriately. I think contr.treatment is misleading (it labels contrasts of two levels by just one of them), and Christoph's labels are informative but impractically long. But if you want to label each contrast by the level it contrasts with the rest, cont - array(0, c(lenglev, lenglev - 1), list(levels, levels[-lenglev])) in a modified contr.sum will do it. E.g. z - factor(letters[1:3]) contr.treatment(z) b c a 0 0 b 1 0 c 0 1 contr.sum(z) [,1] [,2] a10 b01 c -1 -1 mycontr.sum(z) a b a 1 0 b 0 1 c -1 -1 On Wed, 13 Jul 2005, Christoph Buser wrote: Dear Ghislain I do not know a general elegant solution, but for some applications the following example may be helpful: ## Artificial data for demonstration: group is fixed, species is random dat - data.frame(group = c(rep(A,20),rep(B,17),rep(C,24)), species = c(rep(sp1, 4), rep(sp2,5), rep(sp3,5), rep(sp4,6), rep(sp5,2), rep(sp6,5), rep(sp7,3), rep(sp8,3), rep(sp9,4), rep(sp10,6), rep(sp11,6), rep(sp12,6), rep(sp13,6)), area = rnorm(61)) ## You can attach a contrast at your fixed factor of interest group ## Create the contrast you like to test (in our case contr.sum for 3 ## levels) mat - contr.sum(3) ## You can add the names you want to see in the output ## Be carefull that you give the correct names to the concerned ## column. Otherwise there is the big danger of misinterpretation. colnames(mat) - c(: A against rest, : B against rest) ## Attatch the contrast at your factor group dat[,group] - C(dat[,group],mat) ## Now calculate the lme library(nlme) reg.lme - lme(area ~ group, data = dat, random = ~ 1|species) summary(reg.lme) Maybe someone has a better idea how to do it generally. Hope this helps Christoph Buser -- Christoph Buser [EMAIL PROTECTED] Seminar fuer Statistik, LEO C13 ETH (Federal Inst. Technology)8092 Zurich SWITZERLAND phone: x-41-44-632-4673 fax: 632-1228 http://stat.ethz.ch/~buser/ -- Ghislain Vieilledent writes: Good morning, I used in R contr.sum for the contrast in a lme model: options(contrasts=c(contr.sum,contr.poly)) Septo5.lme-lme(Septo~Variete+DateSemi,Data4.Iso,random=~1|LieuDit) intervals(Septo5.lme)$fixed lower est. upper (Intercept) 17.0644033 23.106110 29.147816 Variete1 9.5819873 17.335324 25.088661 Variete2 -3.3794907 6.816101 17.011692 Variete3 -0.5636915 8.452890 17.469472 Variete4 -22.8923812 -10.914912 1.062558 Variete5 -10.7152821 -1.865884 6.983515 Variete6 0.2743390 9.492175 18.710012 Variete7 -23.7943250 -15.070737 -6.347148 Variete8 -21.7310554 -12.380475 -3.029895 Variete9 -27.9782575 -17.480555 -6.982852 DateSemi1 -5.7903419 -1.547875 2.694592 DateSemi2 3.6571596 8.428417 13.199675 attr(,label) [1] Fixed effects: How is it possible to obtain a return with the name of my factor's levels as with contr.treatment ? Thanks for you help. -- Ghislain Vieilledent 30, rue Bernard Ortet 31 500 TOULOUSE 06 24 62 65 07 [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] High resolution plots
Peter Dalgaard schrieb: What is the real issue here? Import trouble? If you're importing to Word/PowerPoint, why not use the Windows metafile? Perhaps they are too ugly compared to EPS by your taste? My reason was http://www.adobe.com/support/techdocs/328541.html I tried to plot emf files with pagemaker (not even a cheep or freeware pprogramm) and the lyaout was corrupted. I tried to use eps files the text was corrupted and I do not know how to transfer the *.ps file to tiff. Maybe there is any payware but if you have a layout application for more than 700 Euro you don´t like to by any further application only to change the *.ps format. So I am looking for any solution to increase the resolution. Maybe I do not understand the meanings of res, width and heigh or it is not working as expected (see answer from 12 Jul 2005) I still hope that Porf. Brian D. Ripley will answer to this question even though I got him on his nerves with a wrong answer. :-( with regards Knut Krueger http://www.biostatistic.de __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Ellipse Cosrrespondence Analysis(Internal Stability)
I'am a student of Educational Universiti of Indonesia. I take Internal Stabiliti in Correespondence Analysis of Normal Distributerd Data for my final task. I've had problem with ellipse contructing as STATISTICA 6.0 generate in Correspondence Analysis Session. I thing the ellipse in bivariate normal distribution can be applied in the plot of correspondence analysis. The ellipse is t(X-Xbar)(inverse(Sigma))(X-Xbar)= chi-square(alfa,2) and the probability is P[t(X-Xbar)(inverse(Sigma))(X-Xbar)= chi-square(alfa,2)]=1-alfa The point outside the ellipse become supplementary points as the point outside the ellipse in the bivariate normal is outlier. If i construct the ellipse from the row and column coordinates like t(F-Fbar)(inverse(Sigma(F)))(F-Fbar)= chi-square(alfa,2), wheter the ellipse valid? (for detecting outlier data) How can i create the valid ellipse ? Please Help Me. Thanks to you all.. - [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] High resolution plots
Knut Krueger [EMAIL PROTECTED] writes: Peter Dalgaard schrieb: What is the real issue here? Import trouble? If you're importing to Word/PowerPoint, why not use the Windows metafile? Perhaps they are too ugly compared to EPS by your taste? My reason was http://www.adobe.com/support/techdocs/328541.html Oh, lovely. I tried to plot emf files with pagemaker (not even a cheep or freeware pprogramm) and the lyaout was corrupted. I tried to use eps files the text was corrupted and I do not know how to transfer the *.ps file to tiff. Stupid question: You did remember to close the graphics device when you generated those plots (dev.off())? If an EPS file doesn't work with an Adobe product, there is some reason to suspect that the file is corrupted or that there is an issue with R's postscript driver. If you're seeing text corruption outside the plot, then surely the E in EPS (for encapsulated) is not working as intended. Did you try the pdf() driver? That usually comes out rather nicely in connection with Acrobat/Distiller. Maybe there is any payware but if you have a layout application for more than 700 Euro you don´t like to by any further application only to change the *.ps format. So I am looking for any solution to increase the resolution. Maybe I do not understand the meanings of res, width and heigh or it is not working as expected (see answer from 12 Jul 2005) I still hope that Porf. Brian D. Ripley will answer to this question even though I got him on his nerves with a wrong answer. :-( with regards Knut Krueger http://www.biostatistic.de -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] High resolution plots
Peter Dalgaard schrieb: Knut Krueger [EMAIL PROTECTED] writes: Peter Dalgaard schrieb: What is the real issue here? Import trouble? If you're importing to Word/PowerPoint, why not use the Windows metafile? Perhaps they are too ugly compared to EPS by your taste? My reason was http://www.adobe.com/support/techdocs/328541.html Oh, lovely. isn't it ? :-( Stupid question: You did remember to close the graphics device when you generated those plots (dev.off())? the first idea was no I didn´t ... but I found my script and I did :-) If an EPS file doesn't work with an Adobe product, there is some reason to suspect that the file is corrupted or that there is an issue with R's postscript driver. If you're seeing text corruption outside the plot, then surely the E in EPS (for encapsulated) is not working as intended. That's a normal behaviour in windows surrounding. Did you try the pdf() driver? That usually comes out rather nicely in connection with Acrobat/Distiller. No I did not actually. the plots are good enough for printing with an laser printer, but they are not good enough for the journal. It is curious. one of them wants only pdf the other one wants tiff and wordperfect or word .. maybe the next one will receive old latin stone plates ;-) Therefore for one journal it is enough to use the eps or bmp resolution with pdf file or just an pdf output, but for the next one I will need a tiff with 300 dpi at least. with regards Knut Krueger http://www.biostatistic.de [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Memory question
Thank you Dr. Spencer Graves for your answer. What kind of matrices? They come form an image of about 3000x5000, and I need to generate arround 1024 matrices of the same size, they are not sparse matrices. What function can I use to, once generated one matrix, I could save into disk and then use the same space for the following, and so on. Thank you very much for your help Kenneth Spencer Graves wrote: What kinds of matrices? There are facilities in the Matrix and SparseM packages that might help for sparse matrices. If they are N x k where N is large and k is not, can you compute something like the QR decomposition and get away with keeping only the R part for most of your matrices? One could potentially define a class of matrices that are only kept in memory only when needed; I think S-Plus may do that. It would take a lot of work to make that work generally, but you might be able to accomplish what you need with a much smaller effort. spencer graves Kenneth Roy Cabrera Torres wrote: Hi R users and developers: I want to know how can I save memory in R for example: - saving on disk a matrix. - using again the matrix (changing their values) - saving again the matrix on disk in a different file. The idea is that I have a process that generate several matrices, but if I keep them all in memory it will overflow. How can I save them in different files, so I use the same amount of memory for each processed matrix? Thank you for your help. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Name for factor's levels with contr.sum
This is exactly what I've been looking for (without success) when I was speaking about a more elegant and general solution. I agree with your argument that labels might be misleading. Nevertheless if a user is aware what contr.sum calculates, it is practical to have an annotation. Thank you very much for the solution. Regards, Christoph Buser -- Christoph Buser [EMAIL PROTECTED] Seminar fuer Statistik, LEO C13 ETH (Federal Inst. Technology) 8092 Zurich SWITZERLAND phone: x-41-44-632-4673 fax: 632-1228 http://stat.ethz.ch/~buser/ -- Prof Brian Ripley writes: One way to do this generally is to make a copy of contr.sum, rename it, and set the dimnames appropriately. I think contr.treatment is misleading (it labels contrasts of two levels by just one of them), and Christoph's labels are informative but impractically long. But if you want to label each contrast by the level it contrasts with the rest, cont - array(0, c(lenglev, lenglev - 1), list(levels, levels[-lenglev])) in a modified contr.sum will do it. E.g. z - factor(letters[1:3]) contr.treatment(z) b c a 0 0 b 1 0 c 0 1 contr.sum(z) [,1] [,2] a10 b01 c -1 -1 mycontr.sum(z) a b a 1 0 b 0 1 c -1 -1 On Wed, 13 Jul 2005, Christoph Buser wrote: Dear Ghislain I do not know a general elegant solution, but for some applications the following example may be helpful: ## Artificial data for demonstration: group is fixed, species is random dat - data.frame(group = c(rep(A,20),rep(B,17),rep(C,24)), species = c(rep(sp1, 4), rep(sp2,5), rep(sp3,5), rep(sp4,6), rep(sp5,2), rep(sp6,5), rep(sp7,3), rep(sp8,3), rep(sp9,4), rep(sp10,6), rep(sp11,6), rep(sp12,6), rep(sp13,6)), area = rnorm(61)) ## You can attach a contrast at your fixed factor of interest group ## Create the contrast you like to test (in our case contr.sum for 3 ## levels) mat - contr.sum(3) ## You can add the names you want to see in the output ## Be carefull that you give the correct names to the concerned ## column. Otherwise there is the big danger of misinterpretation. colnames(mat) - c(: A against rest, : B against rest) ## Attatch the contrast at your factor group dat[,group] - C(dat[,group],mat) ## Now calculate the lme library(nlme) reg.lme - lme(area ~ group, data = dat, random = ~ 1|species) summary(reg.lme) Maybe someone has a better idea how to do it generally. Hope this helps Christoph Buser -- Christoph Buser [EMAIL PROTECTED] Seminar fuer Statistik, LEO C13 ETH (Federal Inst. Technology) 8092 Zurich SWITZERLAND phone: x-41-44-632-4673fax: 632-1228 http://stat.ethz.ch/~buser/ -- Ghislain Vieilledent writes: Good morning, I used in R contr.sum for the contrast in a lme model: options(contrasts=c(contr.sum,contr.poly)) Septo5.lme-lme(Septo~Variete+DateSemi,Data4.Iso,random=~1|LieuDit) intervals(Septo5.lme)$fixed lower est. upper (Intercept) 17.0644033 23.106110 29.147816 Variete1 9.5819873 17.335324 25.088661 Variete2 -3.3794907 6.816101 17.011692 Variete3 -0.5636915 8.452890 17.469472 Variete4 -22.8923812 -10.914912 1.062558 Variete5 -10.7152821 -1.865884 6.983515 Variete6 0.2743390 9.492175 18.710012 Variete7 -23.7943250 -15.070737 -6.347148 Variete8 -21.7310554 -12.380475 -3.029895 Variete9 -27.9782575 -17.480555 -6.982852 DateSemi1 -5.7903419 -1.547875 2.694592 DateSemi2 3.6571596 8.428417 13.199675 attr(,label) [1] Fixed effects: How is it possible to obtain a return with the name of my factor's levels as with contr.treatment ? Thanks for you help. -- Ghislain Vieilledent 30, rue Bernard Ortet 31 500 TOULOUSE 06 24 62 65 07 [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford,
Re: [R] unexpected par('pin') behaviour
joerg == joerg van den hoff [EMAIL PROTECTED] on Wed, 13 Jul 2005 16:00:58 +0200 writes: joerg hi everybody, joerg I noticed the following: in one of my scripts 'layout' is used to joerg generate a (approx. square) grid of variable dimensions (depending on joerg no. of input files). if the no. of subplots (grid cells) becomes joerg moderately large (say 9) I use a construct like joerg ###layout grid computation and set up occurs here### joerg ... joerg opar - par(no.readonly = T); joerg on.exit(par(opar)) joerg par(mar=c(4.1, 4.1, 1.1, .1)) joerg ###plotting occurs here joerg ... joerg to reduce the figure margins to achieve a more joerg compact display. apart from 'mar' no other par() joerg setting is modified. yet another example of using par('no.readonly') when it's not needed and inefficient. Replacing the above by ###layout grid computation and set up occurs here### ... op - par(mar=c(4.1, 4.1, 1.1, .1)) on.exit(par(op)) ###plotting occurs here will be much more efficient and even solve your problem with pin. But then, yes, there might be another par() problem hidden in your code / example, but unfortunately you have not specified reproducible code. joerg this works fine until the total number of subplots becomes too large joerg (large depending on the current size of the X11() graphics device joerg window, e.g. 7 x 6 subplots for the default size fo x11()). joerg I then get the error message (only _after_ all plots are correctly joerg displayed, i.e. obviously during execution of the above on.exit() call) joerg Error in par(opar) : joerg invalid value specified for graphics parameter pin joerg and par(pin) yields: joerg [1] 0.34864 -0.21419 you mean *after* all the plotting , not the pin values in 'opar', right? joerg which indeed is invalid (negative 2nd component). joerg I'm aware of this note from ?par: joerg The effect of restoring all the (settable) graphics parameters as joerg in the examples is hard to predict if the device has been resized. joerg Several of them are attempting to set the same things in different joerg ways, and those last in the alphabet will win. In particular, the joerg settings of 'mai', 'mar', 'pin', 'plt' and 'pty' interact, as do joerg the outer margin settings, the figure layout and figure region joerg size. {{which shows you the known but not widely known fact that traditional par() based graphics are ``flawed by design'' and that's why there is the package grid for better designed graphics }} joerg but my problem occurs without any resizing of the joerg x11() window prior to resetting par to par(opar). It still would be interesting to get a reproducible example here, as the posting guide asks for. Martin joerg any ideas, what is going on? joerg platform powerpc-apple-darwin7.9.0 joerg arch powerpc joerg os darwin7.9.0 joerg system powerpc, darwin7.9.0 joerg status Patched ^^^ I hope that this is not the basic problem joerg major2 joerg minor1.0 joerg year 2005 joerg month05 joerg day 12 joerg language R joerg regards, joerg joerg __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] How to increase memory for R on Soliars 10 with 16GB and 64bit R
From: Uwe Ligges Uwe Ligges wrote: Dongseok Choi wrote: Thank you very much for your help!! Now, it runs without any problem. Is it going to be fixed in the next release? Of course, Brian [hmmm, looks like some wrong shortcut has been used - and it must have been me who forgot to drink coffee before strating to post ...] Is this indication that R-help is more addictive than coffee? Andy I meant to say: Of course, Brian already did, see the svn log: -- -- r34947 | ripley | 2005-07-13 03:16:18 -0400 (Wed, 13 Jul 2005) | 1 line Changed paths: M /trunk/NEWS M /trunk/src/main/memory.c correct typo for R_alloc on 64-bit systems Uwe Ligges Thanks again, Dongseok Dongseok Choi, Ph.D. Assistant Professor Division of Biostatistics Department of Public Health Preventive Medicine Oregon Health Science University 3181 SW Sam Jackson Park Road, CB-669 Portland, OR 97239-3098 TEL) 503-494-5336 FAX) 503-494-4981 [EMAIL PROTECTED] Prof Brian Ripley [EMAIL PROTECTED] 07/13/05 12:03 AM On Tue, 12 Jul 2005, Dongseok Choi wrote: My machine is SUN Java Workstation 2100 with 2 AMD Opteron CPUs and 16GB RAM. R is compiled as 64bit by using SUN compilers. I trying to fit quantile smoothing on my data and I got an message as below. fit1-rqss(z1~qss(cbind(x,y),lambda=la1),tau=t1) Error in as.matrix.csr(diag(n)) : cannot allocate memory block of size 2496135168 The lengths of vector x and y are both 17664. I tried and found that the same command ran with x[1:16008] and y[1:16008]. So, it looks to me a memory related problem, but I'm not sure how I can allocate memory block. I read the command line option but not sure what do to with it. Could you help me on this? It is trying to allocate a single memory block of size over 2^31-1 bytes. R internally uses ints for sizes of vectors and that is a limit (see help(Memory-limits) ). However, it is intended that on 64-bit systems that there is a limit here of 8*(2^31-1) but there was a typo. Please change line 1534 of src/main/memory.c to #if SIZEOF_LONG 4 and re-compile. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] High resolution plots
Knut Krueger [EMAIL PROTECTED] writes: Peter Dalgaard schrieb: Knut Krueger [EMAIL PROTECTED] writes: Peter Dalgaard schrieb: What is the real issue here? Import trouble? If you're importing to Word/PowerPoint, why not use the Windows metafile? Perhaps they are too ugly compared to EPS by your taste? My reason was http://www.adobe.com/support/techdocs/328541.html Oh, lovely. isn't it ? :-( Stupid question: You did remember to close the graphics device when you generated those plots (dev.off())? the first idea was no I didn´t ... but I found my script and I did :-) If an EPS file doesn't work with an Adobe product, there is some reason to suspect that the file is corrupted or that there is an issue with R's postscript driver. If you're seeing text corruption outside the plot, then surely the E in EPS (for encapsulated) is not working as intended. That's a normal behaviour in windows surrounding. Did you try the pdf() driver? That usually comes out rather nicely in connection with Acrobat/Distiller. No I did not actually. the plots are good enough for printing with an laser printer, but they are not good enough for the journal. It is curious. one of them wants only pdf the other one wants tiff and wordperfect or word .. maybe the next one will receive old latin stone plates ;-) Therefore for one journal it is enough to use the eps or bmp resolution with pdf file or just an pdf output, but for the next one I will need a tiff with 300 dpi at least. It's still a bit weird if they want a bitmapped format, given the difficulty of scaling such beasts, but what was ever wrong with installing ghostscript and doing bitmap(my.tiff, tiff24nc, res=300) ? -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] memory problem
I'm a beginner in R and, therefore, I don't know how serious my trouble is. After running a script: ** *t**-c(14598417794,649693)* *data**=data.frame(read.spss(C:\\Ginters\\Kalibracija\\cal_data.sav))* *Xs=**as.matrix(data[,1:2]) * *koef**=data.frame(read.spss(C:\\Ginters\\Kalibracija\\f.sav)) * *piks=**as.matrix(koef[,1**])* *g=regressionestimator(Xs,piks,t)* I get: *Error: cannot allocate vector of size 1614604 Kb In addition: Warning messages: 1: Reached total allocation of 255Mb: see help(memory.size) 2: Reached total allocation of 255Mb: see help(memory.size) * My OS is Win 2000 Proffesional. Those 2 objects are of sizes * object.size(Xs) [1] 805404 object.size(piks) [1] 115128* accordingly. The 2 files use only 142KB and 60KB accordingly. Why does memory need so much (1.6 GB) space? How can I enlarge it? Is it possible to allocate a part of memory used to the hard drive? Or, is the trouble only with my script? Best regards, Ginters [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Memory question
On Thu, 14 Jul 2005, Kenneth Cabrera wrote: Thank you Dr. Spencer Graves for your answer. What kind of matrices? They come form an image of about 3000x5000, and I need to generate arround 1024 matrices of the same size, they are not sparse matrices. What function can I use to, once generated one matrix, I could save into disk and then use the same space for the following, and so on. You can use either save or .saveRDS/serialize followed by rm() and gc(). You cannot use the same space, but you can free up the space. Then when you need the data again, load/.readRDS/unserialize can pull the object back. (If you arrange this right the object will only go into a temporary frame and so only be needed one at a time.) Thank you very much for your help Kenneth Spencer Graves wrote: What kinds of matrices? There are facilities in the Matrix and SparseM packages that might help for sparse matrices. If they are N x k where N is large and k is not, can you compute something like the QR decomposition and get away with keeping only the R part for most of your matrices? One could potentially define a class of matrices that are only kept in memory only when needed; I think S-Plus may do that. It would take a lot of work to make that work generally, but you might be able to accomplish what you need with a much smaller effort. spencer graves Kenneth Roy Cabrera Torres wrote: Hi R users and developers: I want to know how can I save memory in R for example: - saving on disk a matrix. - using again the matrix (changing their values) - saving again the matrix on disk in a different file. The idea is that I have a process that generate several matrices, but if I keep them all in memory it will overflow. How can I save them in different files, so I use the same amount of memory for each processed matrix? Thank you for your help. -- 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 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] memory problem
On 7/14/2005 7:19 AM, Ginters wrote: I'm a beginner in R and, therefore, I don't know how serious my trouble is. After running a script: ** *t**-c(14598417794,649693)* *data**=data.frame(read.spss(C:\\Ginters\\Kalibracija\\cal_data.sav))* *Xs=**as.matrix(data[,1:2]) * *koef**=data.frame(read.spss(C:\\Ginters\\Kalibracija\\f.sav)) * *piks=**as.matrix(koef[,1**])* *g=regressionestimator(Xs,piks,t)* I get: *Error: cannot allocate vector of size 1614604 Kb In addition: Warning messages: 1: Reached total allocation of 255Mb: see help(memory.size) 2: Reached total allocation of 255Mb: see help(memory.size) * My OS is Win 2000 Proffesional. Those 2 objects are of sizes * object.size(Xs) [1] 805404 object.size(piks) [1] 115128* accordingly. The 2 files use only 142KB and 60KB accordingly. Why does memory need so much (1.6 GB) space? How can I enlarge it? Is it possible to allocate a part of memory used to the hard drive? Or, is the trouble only with my script? This sounds like a problem with the regressionestimator function, which I think comes from the sampling package. You'll need to contact the maintainer of the package to find out why it needs so much memory, and whether there's a way to get what you want without it. Duncan Murdoch __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] help: how to change the size of a window after it has been created
Hello, I wish to plot some figures in a window in turn, but the size of these figures is different, so how can I change the size of the window by resetting the parameters before each plotting? Thank you, Shengzhe __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Plotting greek symbols in plot titles, labels, etc.
Hi, Can anyone please tell me how to use the escape sequences (e.g. \\*a for alfa) for plotting Greek symbols in for example plot titles? I have read the reference manual, but can't seem to get it to work. Thanx! Roy -- The information contained in this communication and any atta...{{dropped}} __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Plotting greek symbols in plot titles, labels, etc.
Take a look at '?plotmath'. -roger Roy Werkman wrote: Hi, Can anyone please tell me how to use the escape sequences (e.g. \\*a for alfa) for plotting Greek symbols in for example plot titles? I have read the reference manual, but can't seem to get it to work. Thanx! Roy -- Roger D. Peng http://www.biostat.jhsph.edu/~rpeng/ __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] help: how to change the size of a window after it has been created
Not sure you can do this. You might have to launch separate graphics windows. -roger wu sz wrote: Hello, I wish to plot some figures in a window in turn, but the size of these figures is different, so how can I change the size of the window by resetting the parameters before each plotting? Thank you, Shengzhe __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Roger D. Peng http://www.biostat.jhsph.edu/~rpeng/ __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] help: how to change the size of a window after it has been created
On Thu, 2005-07-14 at 14:38 +0200, wu sz wrote: Hello, I wish to plot some figures in a window in turn, but the size of these figures is different, so how can I change the size of the window by resetting the parameters before each plotting? Thank you, Shengzhe Other than dragging a plot window with a mouse, I do not think that there is a way to change the size of an open display device via code (though somebody will no doubt correct me if I am wrong). See ?Devices Depending upon your OS, there is likely a 'width' and 'height' argument for the screen display device for newly opened devices. For example, under systems using X, where X11() is the screen device: X11(width = 11, height = 8) plot(1:10) X11(width = 5, height = 5) barplot(1:5) In the above example, two separate plot windows will be created. If you want to only have one open at a time, you can close the device first using dev.off() before the subsequent plots. However, you lose the current plot as soon as you close it. X11(width = 11, height = 8) plot(1:10) dev.off() X11(width = 5, height = 5) barplot(1:5) dev.off() HTH, Marc Schwartz __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] help: how to change the size of a window after it has been created
Marc Schwartz wrote: On Thu, 2005-07-14 at 14:38 +0200, wu sz wrote: Hello, I wish to plot some figures in a window in turn, but the size of these figures is different, so how can I change the size of the window by resetting the parameters before each plotting? Thank you, Shengzhe Other than dragging a plot window with a mouse, I do not think that there is a way to change the size of an open display device via code (though somebody will no doubt correct me if I am wrong). Perhaps some lateral thinking is called for. Rather than change the size of the window, how about changing the size of the plotting region within the window. For example: split.screen(matrix(c(0,0.5, 0,1, 0.5,1,0,1), ncol=4, byrow=T)) plot(1:5) close.screen(all=T) split.screen(matrix(c(0,0.5, 0,0.5, 0.5,1,0.5,1), ncol=4, byrow=T)) plot(1:5) close.screen(all=T) Bob -- Bob O'Hara Dept. of Mathematics and Statistics P.O. Box 68 (Gustaf Hällströmin katu 2b) FIN-00014 University of Helsinki Finland Telephone: +358-9-191 51479 Mobile: +358 50 599 0540 Fax: +358-9-191 51400 WWW: http://www.RNI.Helsinki.FI/~boh/ Journal of Negative Results - EEB: http://www.jnr-eeb.org __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Fieller's Conf Limits and EC50's
Hi, I didn't verify your formulas for Fieller's method of computing the confidence interval. A slightly simpler approach is to use the Delta method to compute the CI. It is also valid for any link function. It yields a simpler formula for the variance of EC50 (for any link function): varEC50 - 1/b1^2 * (var.b0 + EC50^2*var.b1 + 2*EC50*cov.b0.b1) So, you can compute the CIs as: LCL - EC50 - zalpha.2 * sqrt(varEC50) UCL - EC50 + zalpha.2 * sqrt(varEC50) This works for any EC_p, where p is the probability of getting a positive response, and for link function. The CI from the Delta method should be very nearly the same as that obtained using Fieller's method for EC50. For smaller probabilities (e.g., p 0.1), CIs for the EC_p values obtained using the two methods can be slightly different. Ravi. -Original Message- From: [EMAIL PROTECTED] [mailto:r-help- [EMAIL PROTECTED] On Behalf Of Stephen B. Cox Sent: Wednesday, July 13, 2005 12:43 PM To: r-help@stat.math.ethz.ch Subject: [R] Fieller's Conf Limits and EC50's Folks I have modified an existing function to calculate 'ec/ld/lc' 50 values and their associated Fieller's confidence limits. It is based on EC50.calc (writtien by John Bailer) - but also borrows from the dose.p (MASS) function. My goal was to make the original EC50.calc function flexible with respect to 1) probability at which to calculate the expected dose, and 2) the link function. I would appreciate comments about the validity of doing so! In particular - I want to make sure that the confidence limit calculations are still valid when changing the link function. ec.calc-function(obj,conf.level=.95,p=.5) { # calculates confidence interval based upon Fieller's thm. # modified version of EC50.calc found in PB Fig 7.22 # now allows other link functions, using the calculations # found in dose.p (MASS) # SBC 19 May 05 call - match.call() coef = coef(obj) vcov = summary.glm(obj)$cov.unscaled b0-coef[1] b1-coef[2] var.b0-vcov[1,1] var.b1-vcov[2,2] cov.b0.b1-vcov[1,2] alpha-1-conf.level zalpha.2 - -qnorm(alpha/2) gamma - zalpha.2^2 * var.b1 / (b1^2) eta = family(obj)$linkfun(p) #based on calcs in VR's dose.p EC50 - (eta-b0)/b1 const1 - (gamma/(1-gamma))*(EC50 + cov.b0.b1/var.b1) const2a - var.b0 + 2*cov.b0.b1*EC50 + var.b1*EC50^2 - gamma*(var.b0 - cov.b0.b1^2/var.b1) const2 - zalpha.2/( (1-gamma)*abs(b1) )*sqrt(const2a) LCL - EC50 + const1 - const2 UCL - EC50 + const1 + const2 conf.pts - c(LCL,EC50,UCL) names(conf.pts) - c(Lower,EC50,Upper) return(conf.pts,conf.level,call=call) } Thanks Stephen Cox __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting- guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Keeping memory usage low for a big script
Hi all, I've got a script that I run on several computers, some with much less ram than others, and I would like to try and keep the memory usage as low as possible. The script creates variables to store intermediate results of calculations, and then at the end writes a text file with the results of the calculations. Is a quick and dirty way to accomplish this to use rm() to delete objects no longer in use, and then run gc()? ...or is there a better way? Thanks in advance, Ken __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Where's iris?
-Original Message- From: Gabor Grothendieck [mailto:[EMAIL PROTECTED] Sent: 13 July 2005 20:30 To: Ruben Roa Cc: R-help@stat.math.ethz.ch Subject: Re: [R] Where's iris? On 7/13/05, Ruben Roa [EMAIL PROTECTED] wrote: Hi: Where is the iris data set actually located in the R 2.1.0 folder (under W XP)? Is it a text file or it is a binary file? Ruben Uwe has already explained how to get it in text form; however, if you are curious about its original format in R then its actually stored in iris.R as R source code which you can view at: https://svn.r-project.org/R/trunk/src/library/datasets/data/iris.R (or download the entire R source and get it from there). The iris.R in the website and the iris.R i get from dump(iris,file=iris.R) from R 2.1.0 W XP are different (the file from dump() describes the factors using structure(as.integer(c()),.Label=c(),class=factor) whereas the file in the website uses gl=(,,label=c()), but the difference did not matter. I wanted to emulate the file format in a txt *.R file with my own data and i could do it with the most helpful advice by you and Uwe. Thank you! Ruben __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Keeping memory usage low for a big script
On 7/14/2005 9:46 AM, Ken Termiso wrote: Hi all, I've got a script that I run on several computers, some with much less ram than others, and I would like to try and keep the memory usage as low as possible. The script creates variables to store intermediate results of calculations, and then at the end writes a text file with the results of the calculations. Is a quick and dirty way to accomplish this to use rm() to delete objects no longer in use, and then run gc()? ...or is there a better way? Using rm() is good. gc() will run automatically as needed; the main reason to run it yourself is just to make calculation timings more reproducible. It will also affect memory fragmentation, most likely beneficially, but not necessarily... Duncan Murdoch __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Fwd: Re: Problem installing R packages
Hi, I am trying to install the R libraries rmutil and repeated on a Mac OS X version 10.4.1 (which has the latest version of the Mac Developer tools installed) and I am having trouble compiling the libraries. The error message I receive is as follows (I have only included the error message when I try and install the rmutil library): * Installing *source* package 'rmutil' ... ** libs gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c cutil.c -o cutil.o gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c dist.c -o dist.o g77 -fno-common -g -O2 -c gettvc.f -o gettvc.o gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c romberg.c -o romberg.o gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c toms614.c -o toms614.o gcc-3.3 -bundle -flat_namespace -undefined suppress -o rmutil.so cutil.o dist.o gettvc.o romberg.o toms614.o -L/usr/local/lib/gcc/powerpc-apple-darwin6.8/3.4.2 -lg2c -lSystem package successfully installed ld: dist.o has external relocation entries in non-writable section (__TEXT,__text) for symbols: restFP saveFP make: *** [rmutil.so] Error 1 ERROR: compilation failed for package 'rmutil' .. I contacted Jim Lindsey (the package author) who wasn't able to suggest a solution. Hence, does anyone have an idea what could be causing this error or does anyone have a Mac binary for these libraries? Thanks, John Marioni -- John Marioni PhD Student Computational Biology Group Department of Applied Mathematics and Theoretical Physics University of Cambridge Wilberforce Road, Cambridge, CB3 0WA Email: [EMAIL PROTECTED] Phone: +44 (0)1223 763391 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Plotting greek symbols in plot titles, labels, etc.
Type demo(plotmath) at the R prompt. At 2:45 PM +0200 7/14/05, Roy Werkman wrote: Hi, Can anyone please tell me how to use the escape sequences (e.g. \\*a for alfa) for plotting Greek symbols in for example plot titles? I have read the reference manual, but can't seem to get it to work. Thanx! Roy At 8:49 AM -0400 7/14/05, Roger D. Peng wrote: Take a look at '?plotmath'. -roger Roy Werkman wrote: Hi, Can anyone please tell me how to use the escape sequences (e.g. \\*a for alfa) for plotting Greek symbols in for example plot titles? I have read the reference manual, but can't seem to get it to work. Thanx! Roy -- -- Don MacQueen Environmental Protection Department Lawrence Livermore National Laboratory Livermore, CA, USA __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] read.xport
I am trying to import data from a SAS XPORT file that contains 24 SAS files. When I use the read.xport procedure only about 16 data frames (components) are created. Any suggestions? * Gary A. Nelson, Ph.D Massachusetts Division of Marine Fisheries 30 Emerson Avenue Gloucester, MA 01930 Phone: (978) 282-0308 x114 Fax: (617) 727-3337 Email: [EMAIL PROTECTED] [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] maps drawing
There is also the package maptools if you want or need to read ESRI shapefiles. Regards, Christoph Buser -- Christoph Buser [EMAIL PROTECTED] Seminar fuer Statistik, LEO C13 ETH (Federal Inst. Technology) 8092 Zurich SWITZERLAND phone: x-41-44-632-4673 fax: 632-1228 http://stat.ethz.ch/~buser/ -- m p writes: Hello, is there a package in R that would allow map drawing: coastlines, country/state boundaries, maybe topography, rivers etc? Thanks for any guidance, Mark __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] maps drawing
GIS-related packages in R: ade4 -- Analysis of Environmental Data : Exploratory and Euclidean methods in Environmental sciences adehabitat -- Analysis of habitat selection by animals fields -- Tools for spatial data GRASS -- Interface between GRASS 5.0 geographical information system and R Mapdata -- Extra Map Databases Mapproj -- Map Projections Maps -- Draw Geographical Maps Maptools -- tools for reading and handling shapefiles Maptree -- Mapping, pruning, and graphing tree models PBSmapping -- PBS Mapping 2 Shapefiles -- Read and Write ESRI Shapefiles Sp -- classes and methods for spatial data Spatial -- Functions for Kriging and Point Pattern Analysis Spatstat -- Spatial Point Pattern analysis, model-fitting and simulation Spdep -- Spatial dependence: weighting schemes, statistics and models etc. On 7/13/05, m p [EMAIL PROTECTED] wrote: Hello, is there a package in R that would allow map drawing: coastlines, country/state boundaries, maybe topography, rivers etc? Thanks for any guidance, Mark __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Xiaohua Dai, Dr. * Postdoctoral in elephant-tree ecosystem simulation * Vegetation community analysis with GIS * Animal movement and foraging study using GIS - Centre for Systems Research, Durban Institute of Technology P.O.Box 953, Durban 4000, South Africa Tel: +27-31-2042737(O) Fax: +27-31-2042736(O) Mobile: +27-723682954 Personal Blog: http://www.blog.edu.cn/blog.asp?name=ecoinformatics [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] unexpected par('pin') behaviour
Martin Maechler wrote: joerg == joerg van den hoff [EMAIL PROTECTED] on Wed, 13 Jul 2005 16:00:58 +0200 writes: joerg hi everybody, joerg I noticed the following: in one of my scripts 'layout' is used to joerg generate a (approx. square) grid of variable dimensions (depending on joerg no. of input files). if the no. of subplots (grid cells) becomes joerg moderately large (say 9) I use a construct like joerg ###layout grid computation and set up occurs here### joerg ... joerg opar - par(no.readonly = T); joerg on.exit(par(opar)) joerg par(mar=c(4.1, 4.1, 1.1, .1)) joerg ###plotting occurs here joerg ... joerg to reduce the figure margins to achieve a more joerg compact display. apart from 'mar' no other par() joerg setting is modified. yet another example of using par('no.readonly') when it's not needed and inefficient. might be. but at least it is immune against modifying some more 'par' settings in the course of modfications at some other place in the programm. inefficiency: should be at the ppm level of total cpu-usage in my case, :-). what's so bad with copying back and forth this moderately large vector? Replacing the above by ###layout grid computation and set up occurs here### ... op - par(mar=c(4.1, 4.1, 1.1, .1)) on.exit(par(op)) ###plotting occurs here will be much more efficient and even solve your problem with pin. right (solves the problem). I'll adopt this change for the time being. thank you. But then, yes, there might be another par() problem hidden in your code / example, but unfortunately you have not specified reproducible code. joerg this works fine until the total number of subplots becomes too large joerg (large depending on the current size of the X11() graphics device joerg window, e.g. 7 x 6 subplots for the default size fo x11()). joerg I then get the error message (only _after_ all plots are correctly joerg displayed, i.e. obviously during execution of the above on.exit() call) joerg Error in par(opar) : joerg invalid value specified for graphics parameter pin joerg and par(pin) yields: joerg [1] 0.34864 -0.21419 you mean *after* all the plotting , not the pin values in 'opar', right? yes joerg which indeed is invalid (negative 2nd component). joerg I'm aware of this note from ?par: joerg The effect of restoring all the (settable) graphics parameters as joerg in the examples is hard to predict if the device has been resized. joerg Several of them are attempting to set the same things in different joerg ways, and those last in the alphabet will win. In particular, the joerg settings of 'mai', 'mar', 'pin', 'plt' and 'pty' interact, as do joerg the outer margin settings, the figure layout and figure region joerg size. {{which shows you the known but not widely known fact that traditional par() based graphics are ``flawed by design'' and that's why there is the package grid for better designed graphics ... which seems to my simple mind a lot more complicated to come to terms with than the graphics package. I understand that grid is more powerful but the subset of functionality provided by 'graphics' seems more difficult to use in 'grid'. wrong or right? }} joerg but my problem occurs without any resizing of the joerg x11() window prior to resetting par to par(opar). It still would be interesting to get a reproducible example here, as the posting guide asks for. ===cut graphics.off() f - function(n=7, m=6) { nm - n*m layout(matrix(1:(nm),n,m)) opar - par(no.readonly = T) on.exit(par(opar)) par(mar = c(4.1, 4.1, 1.1, 0.1)) for (i in 1:nm) plot(i, pch=(i-1)%%25+1) layout(1) } f(5) #good par('pin') f() #bad (at least for x11() default size) par('pin') ===cut Martin thanks for bothering. joerg joerg any ideas, what is going on? joerg platform powerpc-apple-darwin7.9.0 joerg arch powerpc joerg os darwin7.9.0 joerg system powerpc, darwin7.9.0 joerg status Patched ^^^ I hope that this is not the basic problem no, don't think so. that concerned the MacOS GUI, I believe. joerg major2 joerg minor1.0 joerg year 2005 joerg month05 joerg day 12 joerg language R joerg regards, joerg joerg __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] memory problem
On Thu, 14 Jul 2005, Duncan Murdoch wrote: On 7/14/2005 7:19 AM, Ginters wrote: Why does memory need so much (1.6 GB) space? How can I enlarge it? Is it possible to allocate a part of memory used to the hard drive? Or, is the trouble only with my script? This sounds like a problem with the regressionestimator function, which I think comes from the sampling package. You'll need to contact the maintainer of the package to find out why it needs so much memory, and whether there's a way to get what you want without it. You can do regression calibration of sampling weights using the calibrate() function in the survey package. That might use less memory. -thomas __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Fwd: Re: Problem installing R packages
What Fortran compiler (g77) are you using, obtained from where? See http://www.astro.gla.ac.uk/users/norman/note/2004/restFP/ Your lines are wrapped in ways that make parsing impossible, but my guess at the link line does not include -lcc_dynamic. See also the list archives, e.g. http://tolstoy.newcastle.edu.au/R/help/05/06/7345.html and of course the r-sig-mac list for the mysteries of MacOS X. People seem to be moving to gcc4/gfortran on that platform. On Thu, 14 Jul 2005, J-C. Marioni wrote: Hi, I am trying to install the R libraries rmutil and repeated on a Mac OS X version 10.4.1 (which has the latest version of the Mac Developer tools installed) and I am having trouble compiling the libraries. The error message I receive is as follows (I have only included the error message when I try and install the rmutil library): * Installing *source* package 'rmutil' ... ** libs gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c cutil.c -o cutil.o gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c dist.c -o dist.o g77 -fno-common -g -O2 -c gettvc.f -o gettvc.o gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c romberg.c -o romberg.o gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c toms614.c -o toms614.o gcc-3.3 -bundle -flat_namespace -undefined suppress -o rmutil.so cutil.o dist.o gettvc.o romberg.o toms614.o -L/usr/local/lib/gcc/powerpc-apple-darwin6.8/3.4.2 -lg2c -lSystem package successfully installed ld: dist.o has external relocation entries in non-writable section (__TEXT,__text) for symbols: restFP saveFP make: *** [rmutil.so] Error 1 ERROR: compilation failed for package 'rmutil' .. I contacted Jim Lindsey (the package author) who wasn't able to suggest a solution. Hence, does anyone have an idea what could be causing this error or does anyone have a Mac binary for these libraries? Thanks, John Marioni -- John Marioni PhD Student Computational Biology Group Department of Applied Mathematics and Theoretical Physics University of Cambridge Wilberforce Road, Cambridge, CB3 0WA Email: [EMAIL PROTECTED] Phone: +44 (0)1223 763391 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] DSC 2005 registration
Early registration discount for DSC 2005: Directions in Statistical Computing, August 13-14, in Seattle closes on July 18. Registration is online, from the conference web page at http://depts.washington.edu/dsc2005/ Titles and abstracts for presentations are at http://depts.washington.edu/dsc2005/program.html -thomas Thomas Lumley Assoc. Professor, Biostatistics [EMAIL PROTECTED] University of Washington, Seattle __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Coxph with factors
Kylie-Anne Richards kar at itga.com.au writes: I am fitting a coxph model with factors. I am running into problems when using 'survfit'. I am unsure how R is treating the factors when I fit, say: DATA-data.frame(time.sec,done,f.pom=factor(f.pom),po,vo) final-coxph(Surv(time.sec,done)~f.pom*vo+po,data=DATA) final.surv-survfit((final), individual=T,conf.type=log-log) print(final.surv) Better chances to get a reply when you tell us what problems you are running into. Dieter Menne __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] stripchart usage and alternatives
Mike R mike.rstat at gmail.com writes: . stripchart(r~u,vertical=TRUE,pch=21) stripchart(rm~levels(uf),vertical=TRUE,pch=3,add=TRUE) the above code creates a scatter plot of nominal data are there alternatives to generate the same or similar kind of figure? How about stripplot in lattice/trellis? Dieter __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Fwd: Re: Problem installing R packages
We used the Fortran compiler that came from the customised option when installing from the mac dmg g77 --version GNU Fortran (GCC) 3.4.2 We will look into the links you provided. Thanks, John On Jul 14 2005, Prof Brian Ripley wrote: What Fortran compiler (g77) are you using, obtained from where? See http://www.astro.gla.ac.uk/users/norman/note/2004/restFP/ Your lines are wrapped in ways that make parsing impossible, but my guess at the link line does not include -lcc_dynamic. See also the list archives, e.g. http://tolstoy.newcastle.edu.au/R/help/05/06/7345.html and of course the r-sig-mac list for the mysteries of MacOS X. People seem to be moving to gcc4/gfortran on that platform. On Thu, 14 Jul 2005, J-C. Marioni wrote: Hi, I am trying to install the R libraries rmutil and repeated on a Mac OS X version 10.4.1 (which has the latest version of the Mac Developer tools installed) and I am having trouble compiling the libraries. The error message I receive is as follows (I have only included the error message when I try and install the rmutil library): * Installing *source* package 'rmutil' ... ** libs gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c cutil.c -o cutil.o gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c dist.c -o dist.o g77 -fno-common -g -O2 -c gettvc.f -o gettvc.o gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c romberg.c -o romberg.o gcc-3.3 -no-cpp-precomp -I/Library/Frameworks/R.framework/Resources/include -I/usr/local/include -fno-common -g -O2 -c toms614.c -o toms614.o gcc-3.3 -bundle -flat_namespace -undefined suppress -o rmutil.so cutil.o dist.o gettvc.o romberg.o toms614.o -L/usr/local/lib/gcc/powerpc-apple-darwin6.8/3.4.2 -lg2c -lSystem package successfully installed ld: dist.o has external relocation entries in non-writable section (__TEXT,__text) for symbols: restFP saveFP make: *** [rmutil.so] Error 1 ERROR: compilation failed for package 'rmutil' .. I contacted Jim Lindsey (the package author) who wasn't able to suggest a solution. Hence, does anyone have an idea what could be causing this error or does anyone have a Mac binary for these libraries? Thanks, John Marioni -- John Marioni PhD Student Computational Biology Group Department of Applied Mathematics and Theoretical Physics University of Cambridge Wilberforce Road, Cambridge, CB3 0WA Email: [EMAIL PROTECTED] Phone: +44 (0)1223 763391 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] New Package: TeachingDemos
This is to announce the new package TeachingDemos now available on CRAN. This is a collection of functions that I have used to demonstrate different ideas in various classes taught. Some of the first ones I wrote just to help me visualize what was happening, they latter were usefull in teaching others (both in a 1 on 1 setting, and in classroom type settings). A few examples (more are available in the package): loess.demo: This was one of the first I wrote, mainly for myself, but others have liked it also. This function creates a scatterplot and shows the loess fit, then when you click on a point in the plot it shows the window around that point, the weights used for the local regression at that point and the local regression used to predict at that point. Additional clicks will show the same info for other x-values. vis.gamma, vis.binom, vis.norm, and vis.t: These functions plot the given distribution (and related distributions) then creates a Tk slider box that allows you to change the parameters to see how the distribution function changes. With overlayed plots you can also see the relationship between related distributions. put.points.demo: Start with a scatterplot that shows the correlation coefficient and least squares line, then add, delete, move points interactively to see how the correlation and regression line change. Hopefully others will find these demonstrations useful, let me know if you have any suggestions for additional demonstrations or improvements on the current ones. Greg Snow, Ph.D. Statistical Data Center, LDS Hospital Intermountain Health Care [EMAIL PROTECTED] (801) 408-8111 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] A statistical modeling problem
Hi, This is not an R related question and I apologize for that, but given the brain power of the R community it is hard for me to resist posting this here. I have a problem where each participant is shown a series of visual cues (displayed on a computer screen in a random order) and asked to respond by pressing a button (from a finite number of buttons) that corresponds to the correct answer. The data is of the form (r_ij , t_ij), where r_ij = 1 indicates correct response by the i-th subject on the j-th cue, and t_ij is the time taken to respond. Are there statistical models to describe this set up? Any references or suggestions would be greatly appreciated. Thanks very much, Ravi. [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] read.xport
How about avoiding SAS XPORT altogether and exporting everything in the simple, clean, non-proprietary, extremely reliable, platform-independent ... etc text format (CSV, tab delimited etc)? -Original Message- From: Nelson, Gary (FWE) [mailto:[EMAIL PROTECTED] Sent: Thursday, July 14, 2005 10:31 AM To: r-help@stat.math.ethz.ch Subject: [R] read.xport I am trying to import data from a SAS XPORT file that contains 24 SAS files. When I use the read.xport procedure only about 16 data frames (components) are created. Any suggestions? ** *** Gary A. Nelson, Ph.D Massachusetts Division of Marine Fisheries 30 Emerson Avenue Gloucester, MA 01930 Phone: (978) 282-0308 x114 Fax: (617) 727-3337 Email: [EMAIL PROTECTED] [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] read.xport
bogdan romocea wrote: How about avoiding SAS XPORT altogether and exporting everything in the simple, clean, non-proprietary, extremely reliable, platform-independent ... etc text format (CSV, tab delimited etc)? I hope the problem is fixed in the latest version of foreign (no version info was given). In case it's not, you may want to look at the sasxport.get function in the Hmisc package. Frank -Original Message- From: Nelson, Gary (FWE) [mailto:[EMAIL PROTECTED] Sent: Thursday, July 14, 2005 10:31 AM To: r-help@stat.math.ethz.ch Subject: [R] read.xport I am trying to import data from a SAS XPORT file that contains 24 SAS files. When I use the read.xport procedure only about 16 data frames (components) are created. Any suggestions? ** *** Gary A. Nelson, Ph.D Massachusetts Division of Marine Fisheries 30 Emerson Avenue Gloucester, MA 01930 Phone: (978) 282-0308 x114 Fax: (617) 727-3337 Email: [EMAIL PROTECTED] [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Frank E Harrell Jr Professor and Chair School of Medicine Department of Biostatistics Vanderbilt University __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] plot the number of replicates at the same point
Thank you for thinking about the problem for me. However, I have found that your method doesn't work at all. You may test the following example: x1=c(0.6,0.4,.4,.4,.2,.2,.2,0,0) x2=c(0.4,.2,.4,.6,0,.2,.4,0,.2) x1=rep(x1,4) x2=rep(x2,4) temp=data.frame(x1,x2) temp1=table(temp) plot(temp$x1,temp$x2,cex=0) text(as.numeric(rownames(temp1)), as.numeric(colnames(temp1)), temp1) what I got here is not what I wanted. You may compare with plot(x1,x2) I actually want some plots similar to what SAS proc plot produced. Does anybody have a clue of how to do this easily in R? --- Jean Eid [EMAIL PROTECTED] wrote: You can do the following (don't know it this is the most efficient way but it works) temp-read.table(your file to read the data, header=T) temp1-table(temp) plot(temp$x, temp$y, cex=0) text(as.numeric(rownames(temp1)), as.numeric(colnames(temp1)), temp1) HTH On Wed, 13 Jul 2005, Kerry Bush wrote: Dear R-helper, I want to plot the following-like data: x y 1 1 1 1 1 2 1 3 1 3 1 4 .. In the plot that produced, I don't want to show the usual circles or points. Instead, I want to show the number of replicates at that point. e.g. at the position of (1,1), there are 2 obsevations, so a number '2' will be displayed in the plot. Is my narrative clear? Is there a way to make the plot in R? __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Calculate of data frame
Hello, I have a dataframe which I read from a file: df - readtable(myFile); The dataframe has 4 columns: model, count, value and date where model and date are made of charactors and count and value are numbers. An example is like: modelcountvaluedate A420.87/1/2005 A310.27/2/2005 B714.27/1/2005 B16.2 7/2/2005 A2117/1/2005 B965.2 7/2/2005 Now I want to get the total count and value for each model/date pair, like modelcountvaluedate A631.87/1/2005 A310.27/2/2005 B714.27/1/2005 B1067.4 7/2/2005 Anyone can tell me how to do this? Thanks. Fan [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Error running lme.
I am trying to fit lme using R 2.1.1 under Windows 2k. I am getting the following message noted below. Any suggestions that would help me correct my error would be greatly appreciated. Thanks, John fit1lme-lme(Velocity~time,data=gate) Error in getGroups.data.frame(dataMix, groups) : Invalid formula for groups John Sorkin M.D., Ph.D. Chief, Biostatistics and Informatics Baltimore VA Medical Center GRECC and University of Maryland School of Medicine Claude Pepper OAIC University of Maryland School of Medicine Division of Gerontology Baltimore VA Medical Center 10 North Greene Street GRECC (BT/18/GR) Baltimore, MD 21201-1524 410-605-7119 - NOTE NEW EMAIL ADDRESS: [EMAIL PROTECTED] [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] read.xport
I have the latest version of foreign, but it still doesn't work. I quickly tried the Hmisc package, but the same issue arose. I will delve into the Hmisc package further. Thanks. -Original Message- From: Frank E Harrell Jr [mailto:[EMAIL PROTECTED] Sent: Thursday, July 14, 2005 11:46 AM To: bogdan romocea Cc: Nelson, Gary (FWE); R-help@stat.math.ethz.ch Subject: Re: [R] read.xport bogdan romocea wrote: How about avoiding SAS XPORT altogether and exporting everything in the simple, clean, non-proprietary, extremely reliable, platform-independent ... etc text format (CSV, tab delimited etc)? I hope the problem is fixed in the latest version of foreign (no version info was given). In case it's not, you may want to look at the sasxport.get function in the Hmisc package. Frank -Original Message- From: Nelson, Gary (FWE) [mailto:[EMAIL PROTECTED] Sent: Thursday, July 14, 2005 10:31 AM To: r-help@stat.math.ethz.ch Subject: [R] read.xport I am trying to import data from a SAS XPORT file that contains 24 SAS files. When I use the read.xport procedure only about 16 data frames (components) are created. Any suggestions? ** *** Gary A. Nelson, Ph.D Massachusetts Division of Marine Fisheries 30 Emerson Avenue Gloucester, MA 01930 Phone: (978) 282-0308 x114 Fax: (617) 727-3337 Email: [EMAIL PROTECTED] [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Frank E Harrell Jr Professor and Chair School of Medicine Department of Biostatistics Vanderbilt University __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Partek has Dunn-Sidak Multiple Test Correction. Is this the same/similar to any of R's p.adjust.methods?
The Partek package (www.partek.com) allows only two selections for Multiple Test Correction: Bonferroni and Dunn-Sidak. Can anyone suggest why Partek implemented Dunn-Sidak and not the other methods that R has? Is there any particular advantage to the Dunn-Sidak method? R knows about these methods (in R 2.1.1): p.adjust.methods [1] holm hochberg hommel bonferroni BH BY fdr [8] none BH is Benjamini Hochberg (1995) and is also called fdr (I wish R's documentation said this clearly). BY is Benjamini Yekutieli (2001). I found a few hits from Google on Dunn-Sidak, but I'm curious if anyone can tell me on a conservative-liberal scale, where the Dunn-Sidak method falls? My guess is it's less conservative than Bonferroni (but aren't all the other methods?), but how does it compare to the other methods? A limited numerical experiment suggested this order to me: bonferroni (most conservative), hochberg and holm about the same, BY, BH (also called fdr), and then none. Thanks for any of thoughts on this. efg __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] East Asian language
Dear all, I just installed R 2.1.1. The installation program automatically recognized my Windows XP was using Chinese language, so now my R console displays everything in Chinese. How can I still let R console display in English without modifying my Window XP language setup? Thank you so much! Best, Nan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Does R have ANOVA permutation tests?
Hi list, Does anybody know if R has functions to do the ANOVA permutation test? I googled and found R has the vegan package to do ANOVA like permutation test for Constrained Correspondence Analysis. But does R have a function for general ANOVA-like permutation tests? Thanks in advance! Yuefeng __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Calculate of data frame
Zhang, Fan [EMAIL PROTECTED] writes: Now I want to get the total count and value for each model/date pair, like modelcountvaluedate A631.87/1/2005 A310.27/2/2005 B714.27/1/2005 B1067.4 7/2/2005 Anyone can tell me how to do this? aggregate(df[,c(count,value)],df[,c(date,model)],sum) date model count value 1 7/1/2005 A 6 31.8 2 7/2/2005 A 3 10.2 3 7/1/2005 B 7 14.2 4 7/2/2005 B10 71.4 -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] plot the number of replicates at the same point
On 7/14/05, Kerry Bush [EMAIL PROTECTED] wrote: Thank you for thinking about the problem for me. However, I have found that your method doesn't work at all. You may test the following example: x1=c(0.6,0.4,.4,.4,.2,.2,.2,0,0) x2=c(0.4,.2,.4,.6,0,.2,.4,0,.2) x1=rep(x1,4) x2=rep(x2,4) temp=data.frame(x1,x2) temp1=table(temp) plot(temp$x1,temp$x2,cex=0) text(as.numeric(rownames(temp1)), as.numeric(colnames(temp1)), temp1) what I got here is not what I wanted. You may compare with plot(x1,x2) I actually want some plots similar to what SAS proc plot produced. Does anybody have a clue of how to do this easily in R? Try this: x1=c(0.6,0.4,.4,.4,.2,.2,.2,0,0) x2=c(0.4,.2,.4,.6,0,.2,.4,0,.2) x1=rep(x1,4) x2=rep(x2,4) temp=data.frame(x1,x2) foo - subset(as.data.frame(table(temp)), Freq 0) foo$x1 - as.numeric(as.character(foo$x1)) foo$x2 - as.numeric(as.character(foo$x2)) with(foo, plot(x1, x2, type = n)) with(foo, text(x1, x2, lab = Freq)) -Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Error running lme.
John Your model is not properly specified for lme. You have not included any random effects or a grouping variable. Let me assume just for sake of argument that you want to include a random effect for the intercept and for time. Your lme specification would be fm1 - lme(Velocity~time, random=~time|ID, gate) Where ID is a grouping variable. I would encourage you to work with lmer, which is now found in the Matrix package. The call to lmer differs slightly from lme, fm1 - lmer(Velocity~time + (time|ID), gate) I hope this is helpful, Harold -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of John Sorkin Sent: Thursday, July 14, 2005 1:05 PM To: r-help@stat.math.ethz.ch Subject: [R] Error running lme. I am trying to fit lme using R 2.1.1 under Windows 2k. I am getting the following message noted below. Any suggestions that would help me correct my error would be greatly appreciated. Thanks, John fit1lme-lme(Velocity~time,data=gate) Error in getGroups.data.frame(dataMix, groups) : Invalid formula for groups John Sorkin M.D., Ph.D. Chief, Biostatistics and Informatics Baltimore VA Medical Center GRECC and University of Maryland School of Medicine Claude Pepper OAIC University of Maryland School of Medicine Division of Gerontology Baltimore VA Medical Center 10 North Greene Street GRECC (BT/18/GR) Baltimore, MD 21201-1524 410-605-7119 -- NOTE NEW EMAIL ADDRESS: [EMAIL PROTECTED] [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] East Asian language
Nan Lin wrote: Dear all, I just installed R 2.1.1. The installation program automatically recognized my Windows XP was using Chinese language, so now my R console displays everything in Chinese. How can I still let R console display in English without modifying my Window XP language setup? Thank you so much! E.g. by setting the envrionment variable LANGUAGE=en Uwe Ligges Best, Nan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] plot the number of replicates at the same point
On Thu, 2005-07-14 at 12:30 -0500, Deepayan Sarkar wrote: On 7/14/05, Kerry Bush [EMAIL PROTECTED] wrote: Thank you for thinking about the problem for me. However, I have found that your method doesn't work at all. You may test the following example: x1=c(0.6,0.4,.4,.4,.2,.2,.2,0,0) x2=c(0.4,.2,.4,.6,0,.2,.4,0,.2) x1=rep(x1,4) x2=rep(x2,4) temp=data.frame(x1,x2) temp1=table(temp) plot(temp$x1,temp$x2,cex=0) text(as.numeric(rownames(temp1)), as.numeric(colnames(temp1)), temp1) what I got here is not what I wanted. You may compare with plot(x1,x2) I actually want some plots similar to what SAS proc plot produced. Does anybody have a clue of how to do this easily in R? Try this: x1=c(0.6,0.4,.4,.4,.2,.2,.2,0,0) x2=c(0.4,.2,.4,.6,0,.2,.4,0,.2) x1=rep(x1,4) x2=rep(x2,4) temp=data.frame(x1,x2) foo - subset(as.data.frame(table(temp)), Freq 0) foo$x1 - as.numeric(as.character(foo$x1)) foo$x2 - as.numeric(as.character(foo$x2)) with(foo, plot(x1, x2, type = n)) with(foo, text(x1, x2, lab = Freq)) -Deepayan Great solution Deepayan. I was just in the process of working on something similar. One quick modification is that the two plot()/text() functions can be combined into: with(foo, plot(x1, x2, pch = as.character(Freq))) Best regards, Marc Schwartz __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Fwd: Re: East Asian language
Original message Date: Thu, 14 Jul 2005 19:50:41 +0200 From: Uwe Ligges [EMAIL PROTECTED] Subject: Re: [R] East Asian language To: Nan Lin [EMAIL PROTECTED] Cc: r-help@stat.math.ethz.ch Nan Lin wrote: Dear all, I just installed R 2.1.1. The installation program automatically recognized my Windows XP was using Chinese language, so now my R console displays everything in Chinese. How can I still let R console display in English without modifying my Window XP language setup? Thank you so much! E.g. by setting the envrionment variable LANGUAGE=en Or, like what I did, change XP language setting back to English from the control panel. Uwe Ligges Best, Nan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Efficient testing for +ve definiteness
To reinforce Prof. Ripley's comment that, Knowing the determinant does not tell you if the matrix is close to non-positive definite, note that the determinant of the negative of the identity matrix, (-diag(k)), is (-1)^k; if k is even, the determinant is positive. This silly example connects to real cases, as for example a 3x3 matrix of rank 1 with eigenvalues (1, -1e-20, -1e-21). The last two eigenvalues are buried in the noise relative to the largest (under standard double precision arithmatec). spencer graves Prof Brian Ripley wrote: On Wed, 13 Jul 2005, Makram Talih wrote: Dear R-users, Is there a preferred method for testing whether a real symmetric matrix is positive definite? [modulo machine rounding errors.] The obvious way of computing eigenvalues via E - eigen(A, symmetric=T, only.values=T)$values and returning the result of !any(E = 0) seems less efficient than going through the LU decomposition invoked in determinant.matrix(A) and checking the sign and (log) modulus of the determinant. I suppose this has to do with the underlying C routines. Any thoughts or anecdotes? It has to do with what exactly you want to test. Knowing the determinant does not tell you if the matrix is close to non-positive definite or not. For numerical work, a comparison of the smallest eigenvalue to the largest is usually the most useful indication of possible problems in computations. An alternative is to try a Choleski decomposition, which may be faster but is less informative. Given how fast eigenvalues can be computed by current algorithms (and note the comments in ?eigen) I would suggest not worrying about speed until you need to (probably never). -- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA [EMAIL PROTECTED] www.pdf.com http://www.pdf.com Tel: 408-938-4420 Fax: 408-280-7915 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Fwd: Re: East Asian language
Baoqiang Cao wrote: Original message Date: Thu, 14 Jul 2005 19:50:41 +0200 From: Uwe Ligges [EMAIL PROTECTED] Subject: Re: [R] East Asian language To: Nan Lin [EMAIL PROTECTED] Cc: r-help@stat.math.ethz.ch Nan Lin wrote: Dear all, I just installed R 2.1.1. The installation program automatically recognized my Windows XP was using Chinese language, so now my R console displays everything in Chinese. How can I still let R console display in English without modifying my Window XP language setup? Thank you so much! E.g. by setting the envrionment variable LANGUAGE=en Or, like what I did, change XP language setting back to English from the control panel. No, please read the question... Uwe Ligges Uwe Ligges Best, Nan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Using system to run a stand alone program that requires input in Windows
Under Windows I want to run a stand alone program that takes a number of commands from the user. I've been running the program from the command line using C:\Data\ foo.exe params.txt where foo is the program is params.txt is a text file with a few lines (9) of parameters. I want to run this from R using system a la: system(foo.exe, input = params.txt) but that doesn't do it. What am I missing? platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major2 minor1.0 year 2005 month04 day 18 language R __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Using system to run a stand alone program that requires input in Windows
Dr Carbon wrote: Under Windows I want to run a stand alone program that takes a number of commands from the user. I've been running the program from the command line using C:\Data\ foo.exe params.txt where foo is the program is params.txt is a text file with a few lines (9) of parameters. I want to run this from R using system a la: system(foo.exe, input = params.txt) but that doesn't do it. What am I missing? You misinterpreted input: system(foo.exe params.txt) should be sufficient. See also ?shell. Uwe Ligges platform i386-pc-mingw32 arch i386 os mingw32 system i386, mingw32 status major2 minor1.0 year 2005 month04 day 18 language R __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Calculate of data frame
Thanks for your answer, Peter. Now I have a new data frame which is sorted by (model,date) pair. Since the sorting is done to charactors, I have 7/1/2005 followed by 7/10/2005 instead of 7/2/2005. How can I do the aggregation and at the same time, make the result sorted the way I want (i.e, 7/1/2005 followd by 7/2/2005)? Thanks, Fan -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Peter Dalgaard Sent: Thursday, July 14, 2005 1:29 PM To: Zhang, Fan Cc: r-help@stat.math.ethz.ch Subject: Re: [R] Calculate of data frame Zhang, Fan [EMAIL PROTECTED] writes: Now I want to get the total count and value for each model/date pair, like modelcountvaluedate A631.87/1/2005 A310.27/2/2005 B714.27/1/2005 B1067.4 7/2/2005 Anyone can tell me how to do this? aggregate(df[,c(count,value)],df[,c(date,model)],sum) date model count value 1 7/1/2005 A 6 31.8 2 7/2/2005 A 3 10.2 3 7/1/2005 B 7 14.2 4 7/2/2005 B10 71.4 -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Using system to run a stand alone program that requires input in Windows
system(foo.exe params.txt) Alas that hangs just opens a window with foo.exe; params.txt is not passed to it. That's why I thought input might work. Any other ideas. I'm trying to think of a DOS program that requires input to test this. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Partek has Dunn-Sidak Multiple Test Correction. Is this the same/similar to any of R's p.adjust.methods?
Earl F. Glynn [EMAIL PROTECTED] writes: The Partek package (www.partek.com) allows only two selections for Multiple Test Correction: Bonferroni and Dunn-Sidak. Can anyone suggest why Partek implemented Dunn-Sidak and not the other methods that R has? Is there any particular advantage to the Dunn-Sidak method? R knows about these methods (in R 2.1.1): p.adjust.methods [1] holm hochberg hommel bonferroni BH BY fdr [8] none BH is Benjamini Hochberg (1995) and is also called fdr (I wish R's documentation said this clearly). BY is Benjamini Yekutieli (2001). I found a few hits from Google on Dunn-Sidak, but I'm curious if anyone can tell me on a conservative-liberal scale, where the Dunn-Sidak method falls? My guess is it's less conservative than Bonferroni (but aren't all the other methods?), but how does it compare to the other methods? As far as I gather, D-S is exact for independent tests, conservative for comparisons of group means, and liberal for mutually exclusive tests (in which case Bonferroni is exact). It is always less conservative than Bonferroni, but the difference is small for typical significance levels: when the Bonferroni level is p, the D-S level is 1 - (1-p/N)^N and if you put p=0.05 and vary N you'll find that it varies from 0.05 at N=1 down to 0.04877 at N=10. (Exercise for the students: what is the limit as N goes to infinity?) The three H-methods play a somewhat different game, basically by only requiring multiple-testing adjustment for non-significant tests. The FDR methods play yet differently by allowing the per test level to increase with the number of significant tests. A limited numerical experiment suggested this order to me: bonferroni (most conservative), hochberg and holm about the same, BY, BH (also called fdr), and then none. Thanks for any of thoughts on this. I'd expect the differences to be fairly small in scenarios where the global null hypothesis is true (excluding none). The main difference comes in when some of the nulls are actually false. Also, it depends on your definitions: With the exception of BY and none the p.adjust methods agree on the smallest adjusted p value, so have the same familywise error rate under the global null. If you count the total number of rejected tests, then you get a difference due to cascading in the non-bonferroni cases. -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Using system to run a stand alone program that requires input in Windows
Well, I'm confused because I tried to cook up an exaple and it appears to work as I think it should. If I have a simple script named myInput.bat: @echo off echo Enter y set /p Input= if /i %Input%==y (goto Proceed) echo y not entered exit /b :Proceed echo y entered and a text file with just a y in it called y.txt then this works: system(myInput y.txt, show.output.on.console = T) Enter y y entered So, I'm baffled. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] read.xport
On Thu, 14 Jul 2005, Nelson, Gary (FWE) wrote: I have the latest version of foreign, but it still doesn't work. I Are you sure: a new version was released a few hours ago? It may not answer your question, but please do give actual version numbers (as the posting guide asks). -- 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 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Using system to run a stand alone program that requires input in Windows
You want shell() not system(), I suspect, as you specified a shell command and I gather you are using Windows (not `DOS'). On Thu, 14 Jul 2005, Dr Carbon wrote: system(foo.exe params.txt) Alas that hangs just opens a window with foo.exe; params.txt is not passed to it. That's why I thought input might work. Any other ideas. I'm trying to think of a DOS program that requires input to test this. -- 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 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] test for difference in the order of self generated sequence
Hello, I have an experiment in which we ask subjects to generate a list of thoughts after being exposed to a stimuli. The thoughts were then coded into two categories (e.g. A B). The objective is to show that the order in which thoughts are generated is affected by the experimental conditions. Can somebody tell me what functions in R can help me do the test or point me to the appropriate reference? Thanks! Example data: Subject 1 in condition 1 generated 5 thoughts. The first, second, and 5th thought is in category A, thought 3 and 4 are in category B. Subject 2 only generated 4 thoughts. ---Condition-A-B n1 11, 2, 5 3, 4 n2 23, 4 1, 2 --- Kaiya Liu __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] read.xport
Nelson, Gary (FWE) wrote: I have the latest version of foreign, but it still doesn't work. I quickly tried the Hmisc package, but the same issue arose. I will delve into the Hmisc package further. This is disappointing as we rely on read.xport quite a bit. You can use the option with sasxport.get that reads csv files. The help file has a URL with a full howto. Frank Thanks. -Original Message- From: Frank E Harrell Jr [mailto:[EMAIL PROTECTED] Sent: Thursday, July 14, 2005 11:46 AM To: bogdan romocea Cc: Nelson, Gary (FWE); R-help@stat.math.ethz.ch Subject: Re: [R] read.xport bogdan romocea wrote: How about avoiding SAS XPORT altogether and exporting everything in the simple, clean, non-proprietary, extremely reliable, platform-independent ... etc text format (CSV, tab delimited etc)? I hope the problem is fixed in the latest version of foreign (no version info was given). In case it's not, you may want to look at the sasxport.get function in the Hmisc package. Frank -Original Message- From: Nelson, Gary (FWE) [mailto:[EMAIL PROTECTED] Sent: Thursday, July 14, 2005 10:31 AM To: r-help@stat.math.ethz.ch Subject: [R] read.xport I am trying to import data from a SAS XPORT file that contains 24 SAS files. When I use the read.xport procedure only about 16 data frames (components) are created. Any suggestions? ** *** Gary A. Nelson, Ph.D Massachusetts Division of Marine Fisheries 30 Emerson Avenue Gloucester, MA 01930 Phone: (978) 282-0308 x114 Fax: (617) 727-3337 Email: [EMAIL PROTECTED] [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Frank E Harrell Jr Professor and Chair School of Medicine Department of Biostatistics Vanderbilt University __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] visual event programming in R
Hello, One of my students is building a visual environment for the package dprep developed by my research group at the University of Puerto Rico. The environment will be in the style of the Orange (specifically channels and tokens design) and Weka (specifically Knowledge flow environment) machine learning/data mining software. She is using only the tools provided by tcl/tk (active tcl). I wonder if somebody else has built such kind of environment for a R package, and which tools has used? Best, Edgar Acuna CASTLE group UPR-Mayaguez -- Open WebMail Project (http://openwebmail.org) __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] plot the number of replicates at the same point
On 7/14/05, Marc Schwartz (via MN) [EMAIL PROTECTED] wrote: On Thu, 2005-07-14 at 12:30 -0500, Deepayan Sarkar wrote: On 7/14/05, Kerry Bush [EMAIL PROTECTED] wrote: Thank you for thinking about the problem for me. However, I have found that your method doesn't work at all. You may test the following example: x1=c(0.6,0.4,.4,.4,.2,.2,.2,0,0) x2=c(0.4,.2,.4,.6,0,.2,.4,0,.2) x1=rep(x1,4) x2=rep(x2,4) temp=data.frame(x1,x2) temp1=table(temp) plot(temp$x1,temp$x2,cex=0) text(as.numeric(rownames(temp1)), as.numeric(colnames(temp1)), temp1) what I got here is not what I wanted. You may compare with plot(x1,x2) I actually want some plots similar to what SAS proc plot produced. Does anybody have a clue of how to do this easily in R? Try this: x1=c(0.6,0.4,.4,.4,.2,.2,.2,0,0) x2=c(0.4,.2,.4,.6,0,.2,.4,0,.2) x1=rep(x1,4) x2=rep(x2,4) temp=data.frame(x1,x2) foo - subset(as.data.frame(table(temp)), Freq 0) foo$x1 - as.numeric(as.character(foo$x1)) foo$x2 - as.numeric(as.character(foo$x2)) with(foo, plot(x1, x2, type = n)) with(foo, text(x1, x2, lab = Freq)) -Deepayan Great solution Deepayan. I was just in the process of working on something similar. One quick modification is that the two plot()/text() functions can be combined into: with(foo, plot(x1, x2, pch = as.character(Freq))) Only as long as all(Freq 10) (I've been bitten by this before. :-) ). Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] plot the number of replicates at the same point
On Thu, 2005-07-14 at 15:08 -0500, Deepayan Sarkar wrote: On 7/14/05, Marc Schwartz (via MN) [EMAIL PROTECTED] wrote: On Thu, 2005-07-14 at 12:30 -0500, Deepayan Sarkar wrote: On 7/14/05, Kerry Bush [EMAIL PROTECTED] wrote: Thank you for thinking about the problem for me. However, I have found that your method doesn't work at all. You may test the following example: x1=c(0.6,0.4,.4,.4,.2,.2,.2,0,0) x2=c(0.4,.2,.4,.6,0,.2,.4,0,.2) x1=rep(x1,4) x2=rep(x2,4) temp=data.frame(x1,x2) temp1=table(temp) plot(temp$x1,temp$x2,cex=0) text(as.numeric(rownames(temp1)), as.numeric(colnames(temp1)), temp1) what I got here is not what I wanted. You may compare with plot(x1,x2) I actually want some plots similar to what SAS proc plot produced. Does anybody have a clue of how to do this easily in R? Try this: x1=c(0.6,0.4,.4,.4,.2,.2,.2,0,0) x2=c(0.4,.2,.4,.6,0,.2,.4,0,.2) x1=rep(x1,4) x2=rep(x2,4) temp=data.frame(x1,x2) foo - subset(as.data.frame(table(temp)), Freq 0) foo$x1 - as.numeric(as.character(foo$x1)) foo$x2 - as.numeric(as.character(foo$x2)) with(foo, plot(x1, x2, type = n)) with(foo, text(x1, x2, lab = Freq)) -Deepayan Great solution Deepayan. I was just in the process of working on something similar. One quick modification is that the two plot()/text() functions can be combined into: with(foo, plot(x1, x2, pch = as.character(Freq))) Only as long as all(Freq 10) (I've been bitten by this before. :-) ). Deepayan D'oh! Indeed you are correct, since 'pch' is a single character: plot(1:20, pch = as.character(1:20)) as opposed to: plot(1:20, type = n) text(1:20, lab = 1:20) Thanks for the correction. Marc __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] read.xport
Sorry about the version. I have version 0.8-8 of foreign. I downloaded (and installed) it this afternoon. The XPORT files were created in SAS using PROC COPY. I used the code: temp-read.xport(c:/temp/int82ag.xpt). I still get only 16 file of the 24 SAS placed into the XPORT file. The XPORT files come from ftp://cusk.nmfs.noaa.gov/mrfss/intercept/ag/ if anyone wants to try it. I used the .xpt in int82ag.zip and it appears they were created using SASV5XPT. -Original Message- From: Prof Brian Ripley [mailto:[EMAIL PROTECTED] Sent: Thursday, July 14, 2005 2:01 PM To: Nelson, Gary (FWE) Cc: r-help@stat.math.ethz.ch Subject: Re: [R] read.xport On Thu, 14 Jul 2005, Nelson, Gary (FWE) wrote: I have the latest version of foreign, but it still doesn't work. I Are you sure: a new version was released a few hours ago? It may not answer your question, but please do give actual version numbers (as the posting guide asks). -- 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 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] integer codes of factors
U = c(b, b, b, c, d, e, e) F1 = factor( U, levels=c(a, b, c, d, e) ) as.numeric(F1) [1] 2 2 2 3 4 5 5 Here, the integer code of b in F1 is 2 K = factor( levels(F1) ) as.numeric(K) [1] 1 2 3 4 5 K [1] a b c d e Levels: a b c d e And again, the integer code of b in K is 2. Great! I am wondering how modify that usage such that the correspondence between the two numeric vectors can this be trusted. for example, the correspondence can be corrupted by placing the a at the end: F2 = factor( U, levels=c(b, c, d, e, a) ) as.numeric(F2) [1] 1 1 1 2 3 4 4 Placing the a at the end changed the integer code of b in F2 to 1, which is not a problem. But .. K = factor( levels(F2) ) as.numeric( K ) [1] 2 3 4 5 1 K [1] b c d e a Levels: a b c d e But the integer code of b in K is now 2, which does not correspond to its code in F2. One would think that ordered=TRUE ought to avoid the corruption, but it does not seem to accomplish that: K = factor( levels(F2), ordered=TRUE ) as.numeric(K) [1] 2 3 4 5 1 K [1] b c d e a Levels: a b c d e But the integer code of b in K is still 2. However, corruption can be avoided with this idiom: K = factor( levels(F2), levels=levels(F2) ) as.numeric(K) [1] 1 2 3 4 5 K [1] b c d e a Levels: b c d e a Now the integer code of b in K is 1, which, as desired, is in correspondence with its code in F2. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] read.xport
Frank E Harrell Jr [EMAIL PROTECTED] writes: Nelson, Gary (FWE) wrote: I have the latest version of foreign, but it still doesn't work. I quickly tried the Hmisc package, but the same issue arose. I will delve into the Hmisc package further. This is disappointing as we rely on read.xport quite a bit. Indeed. Can SAS read everyhting in the XPORT lib? If so, what can we do to figure out what goes wrong in read.xport. Can you paraphrase the situation using simulated data and put the file somewhere? snippage -Original Message- From: Nelson, Gary (FWE) [mailto:[EMAIL PROTECTED] Sent: Thursday, July 14, 2005 10:31 AM To: r-help@stat.math.ethz.ch Subject: [R] read.xport I am trying to import data from a SAS XPORT file that contains 24 SAS files. When I use the read.xport procedure only about 16 data frames (components) are created. Any suggestions? ** *** Gary A. Nelson, Ph.D Massachusetts Division of Marine Fisheries 30 Emerson Avenue Gloucester, MA 01930 Phone: (978) 282-0308 x114 Fax: (617) 727-3337 Email: [EMAIL PROTECTED] [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Frank E Harrell Jr Professor and Chair School of Medicine Department of Biostatistics Vanderbilt University __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] question from environmental statistics
Dear R users I want to knw if there is a way in which a raw dataset can be modelled by some distribution. besides the gof test is there any test involving gamma or lognormal that would fit the data. thank you -dev __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] read.xport
Peter Dalgaard [EMAIL PROTECTED] writes: Frank E Harrell Jr [EMAIL PROTECTED] writes: Nelson, Gary (FWE) wrote: I have the latest version of foreign, but it still doesn't work. I quickly tried the Hmisc package, but the same issue arose. I will delve into the Hmisc package further. This is disappointing as we rely on read.xport quite a bit. Indeed. Can SAS read everyhting in the XPORT lib? If so, what can we do to figure out what goes wrong in read.xport. Can you paraphrase the situation using simulated data and put the file somewhere? Lines crossed... Gary anwered the last bits already, and I can confirm both that SAS will read all 24 elements, whereas lookup.xport leaves off 7 of them - in the middle. Specifically: names(lookup.xport(int82ag.xpt)) [1] I1_821 I1_822 I1_823 I1_824 I1_825 I1_826 I2_821 [8] I2_822 I2_823 I2_824 I2_825 I2_826 GP4_822 GP4_823 [15] GP4_824 GP4_825 GP4_826 whereas PROC CONTENTS shows grep 'Data Set Name' readxport.lst | awk '{print $4}' FOO.I1_821 FOO.I1_822 FOO.I1_823 FOO.I1_824 FOO.I1_825 FOO.I1_826 FOO.I2_821 FOO.I2_822 FOO.I2_823 FOO.I2_824 FOO.I2_825 FOO.I2_826 FOO.I3_821 FOO.I3_822 FOO.I3_823 FOO.I3_824 FOO.I3_825 FOO.I3_826 FOO.GP4_821 FOO.GP4_822 FOO.GP4_823 FOO.GP4_824 FOO.GP4_825 FOO.GP4_826 Notice that all the I3's and GP4_821 has gone AWOL in lookup.xport -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] plot the number of replicates at the same point
Hi all, This is nowhere near as elegant as Deepayan's solution, but I read the spec as plotting symbols except where there were overlays. You are welcome to improve the following... Jim count.overplot-function(x,y,tol=NULL,...) { if(missing(x)) stop(Usage: count.overplot(x,y,tol=NULL,...)) dimx-dim(x) # if x is a data frame or matrix with at least two columns, split it if(missing(y) !is.null(dimx)) { y-x[,2] x-x[,1] } xlen-length(x) if(xlen != length(y)) stop(x and y must be the same length.) if(is.null(tol)) tol-c((max(x)-min(x))/1000,(max(y)-min(y))/1000) flags-1:xlen xsep-ysep-xdup-ydup-xydup-rep(0,xlen) nsep-ndup-0 for(i in 1:xlen) { if(!is.na(flags[i])) { dups-abs(x - x[i]) = tol[1] abs(y - y[i]) = tol[2] ndups-sum(dups) if(ndups 1) { ndup-ndup+1 xydup-ndups xdup[ndup]-x[i] ydup[ndup]-y[i] } else { nsep-nsep+1 xsep[nsep]-x[i] ysep[nsep]-y[i] } } flags[dups]-NA } plot(xsep[1:nsep],ysep[1:nsep],...) text(xdup[1:ndup],ydup[1:ndup],xydup[1:ndup]) } __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Variance components from lm?
I often use simple nested random-effect models for interlaboratory data. The variance components are important things to know. Is there an R function or package that gets variance components from lm objects? Or can someone point me to a method of doing so? __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] plot the number of replicates at the same point
See ?sunflowerplot for a graphic indication of the number of replications at each point. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Variance components from lm?
Under normal circumstances, 'RSiteSearch(variance components)' would likely identify lme in package nlme and lmer in package lme4. I recommend lme, as it comes with a fairly complete set of helper functions described in Pinheiro and Bates (2000) Mixed-Effect Models in S and S-Plus (Springer). lmer is Doug Bates' next generation product currently under development and lacking some of the helper functions available with lme. spencer graves p.s. I said, Under normal circumstances, because when I tried RSiteSearch just now, it didn't work. When I went to www.r-project.org - Search - R site search, I got Computer trouble, July 13, 2005. When it works (which is 99.7% of the time), it is my favorite tool for finding things about R for which I don't already know where to look. Steve Ellison wrote: I often use simple nested random-effect models for interlaboratory data. The variance components are important things to know. Is there an R function or package that gets variance components from lm objects? Or can someone point me to a method of doing so? __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA [EMAIL PROTECTED] www.pdf.com http://www.pdf.com Tel: 408-938-4420 Fax: 408-280-7915 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Calculation of group summaries
Several people suggested specific functions (by, tapply, sapply and others); thanks for not blowing off a simple question regarding how to do the following SQL in R: select year, site_id, visit_no, mean(undercut) AS meanUndercut, count(undercut) AS nUndercut, std(undercut) AS stdUndercut from channelMorphology group by year, site_id, visit_no ; I'd spent quite a bit of time with the suggested functions earlier but had no luck as I'd misread the docs and put the entire dataframe where it only wants the columns to be processed. Sometimes it's the simplest of things. This has lead to another confoundment-- sd() acts differently than mean() for some reason, at least with R 1.9.0. For some reason, means generate NA results and a warning message for each group: argument is not numeric or logical: returning NA in: mean.default(data[x, ], ...) Of course, the argument is numeric, or there'd be no sd value. Or more likely, I'm still missing something really basic. If I wrap the value in as.numeric() things work fine. Why should I have to do this for mean and median, but not sd? The code below should reproduce this error # Fake data for demo: nsites-6 yearList-1999:2001 fakesub-as.data.frame(cbind( year =rep(yearList,nsites/length(yearList),each=11) ,site_id =rep(c('site1','site2'),each=11*nsites) ,visit_no =rep(1,11*2*nsites) ,transect =rep(LETTERS[1:11],nsites,each=2) ,transdir =rep(c('LF','RT'),11*nsites) ,undercut =abs(rnorm(11*2*nsites,10)) ,angle=runif(11*2*nsites,0,180) )) # Create group summaries: sdmets-by(fakesub$undercut ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,sd ) nmets-by(fakesub$undercut ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,length ) xmets-by(fakesub$undercut ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,mean ) xmets-by(as.numeric(fakesub$undercut) ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,mean ) # Put site id values (year, site_id and visit_no) into results: # List unique id combinations as a list of lists. Then # reorganize that into 3 vectors for final results. # Certainly, there MUST be a better way... foo-strsplit(unique(paste(fakesub$year ,fakesub$site_id ,fakesub$visit_no ,sep='#')) ,split='#' ) year-list() for(i in 1:length(foo)) {year-rbind(year,foo[[i]][1])} site_id-list() for(i in 1:length(foo)) {site_id-rbind(site_id,foo[[i]][2])} visit_no-list() for(i in 1:length(foo)) {visit_no-rbind(visit_no,foo[[i]][3])} # Final result, more or less data.frame(cbind(a=year,b=site_id,c=visit_no,sdmets,nmets,xmets)) cur -- Curt Seeliger, Data Ranger CSC, EPA/WED contractor 541/754-4638 [EMAIL PROTECTED] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Variance components from lm?
From: Spencer Graves Under normal circumstances, 'RSiteSearch(variance components)' would likely identify lme in package nlme and lmer in package lme4. I recommend lme, as it comes with a fairly complete set of helper functions described in Pinheiro and Bates (2000) Mixed-Effect Models in S and S-Plus (Springer). lmer is Doug Bates' next generation product currently under development and lacking some of the helper functions available with lme. spencer graves p.s. I said, Under normal circumstances, because when I tried RSiteSearch just now, it didn't work. When I went to www.r-project.org - Search - R site search, I got Computer trouble, July 13, 2005. When it works (which is 99.7% of the time), it is my favorite tool for finding things about R for which I don't already know where to look. The server (graciously created and maintained by Prof. Jonathan Baron) crashed and a hard drive failure. I was told it would likely be back online next Monday. I missed this search engine, too, as I've grown to depend on it quite a bit. Cheers, Andy Steve Ellison wrote: I often use simple nested random-effect models for interlaboratory data. The variance components are important things to know. Is there an R function or package that gets variance components from lm objects? Or can someone point me to a method of doing so? __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA [EMAIL PROTECTED] www.pdf.com http://www.pdf.com Tel: 408-938-4420 Fax: 408-280-7915 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] question from environmental statistics
Take a look at this document by Vito Ricci: http://cran.r-project.org/doc/contrib/Ricci-distributions-en.pdf Did you try RSiteSearch(Fit distribution) or a Google search? That will lead you to fit.dist{gnlm} and fitdistr{MASS} Cheers Francisco From: [EMAIL PROTECTED] To: r-help@stat.math.ethz.ch Subject: [R] question from environmental statistics Date: Thu, 14 Jul 2005 14:06:45 -0700 Dear R users I want to knw if there is a way in which a raw dataset can be modelled by some distribution. besides the gof test is there any test involving gamma or lognormal that would fit the data. thank you -dev __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Calculation of group summaries
1. Try using more spaces so your code is easier to read. 2. Use data.frame to define your data frame (since the method in your post creates data frames of factors rather than the desired classes). 3. Given the appropriate function, f, a single 'by' statement rbind'ed together, as shown, will create the result. nsites - 6 yearList - 1999:2001 fakesub - data.frame( year = rep(yearList, nsites/length(yearList), each = 11), site_id = rep(c('site1','site2'), each = 11*nsites), visit_no = rep(1, 11*2*nsites), transect = rep(LETTERS[1:11], nsites, each = 2), transdir = rep(c('LF','RT'), 11*nsites), undercut = abs(rnorm(11*2*nsites, 10)), angle= runif(11*2*nsites, 0, 180) ) f - function(x) cbind(year = x[1,1], site_id = x[1,2], visit_no = x[1,3], mean = mean(x[,6]), sd = sd(x[,6]), length = length(x[,6])) do.call(rbind, by(fakesub, fakesub[,1:3], f)) On 7/14/05, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote: Several people suggested specific functions (by, tapply, sapply and others); thanks for not blowing off a simple question regarding how to do the following SQL in R: select year, site_id, visit_no, mean(undercut) AS meanUndercut, count(undercut) AS nUndercut, std(undercut) AS stdUndercut from channelMorphology group by year, site_id, visit_no ; I'd spent quite a bit of time with the suggested functions earlier but had no luck as I'd misread the docs and put the entire dataframe where it only wants the columns to be processed. Sometimes it's the simplest of things. This has lead to another confoundment-- sd() acts differently than mean() for some reason, at least with R 1.9.0. For some reason, means generate NA results and a warning message for each group: argument is not numeric or logical: returning NA in: mean.default(data[x, ], ...) Of course, the argument is numeric, or there'd be no sd value. Or more likely, I'm still missing something really basic. If I wrap the value in as.numeric() things work fine. Why should I have to do this for mean and median, but not sd? The code below should reproduce this error # Fake data for demo: nsites-6 yearList-1999:2001 fakesub-as.data.frame(cbind( year =rep(yearList,nsites/length(yearList),each=11) ,site_id =rep(c('site1','site2'),each=11*nsites) ,visit_no =rep(1,11*2*nsites) ,transect =rep(LETTERS[1:11],nsites,each=2) ,transdir =rep(c('LF','RT'),11*nsites) ,undercut =abs(rnorm(11*2*nsites,10)) ,angle=runif(11*2*nsites,0,180) )) # Create group summaries: sdmets-by(fakesub$undercut ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,sd ) nmets-by(fakesub$undercut ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,length ) xmets-by(fakesub$undercut ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,mean ) xmets-by(as.numeric(fakesub$undercut) ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,mean ) # Put site id values (year, site_id and visit_no) into results: # List unique id combinations as a list of lists. Then # reorganize that into 3 vectors for final results. # Certainly, there MUST be a better way... foo-strsplit(unique(paste(fakesub$year ,fakesub$site_id ,fakesub$visit_no ,sep='#')) ,split='#' ) year-list() for(i in 1:length(foo)) {year-rbind(year,foo[[i]][1])} site_id-list() for(i in 1:length(foo)) {site_id-rbind(site_id,foo[[i]][2])} visit_no-list() for(i in 1:length(foo)) {visit_no-rbind(visit_no,foo[[i]][3])} # Final result, more or less data.frame(cbind(a=year,b=site_id,c=visit_no,sdmets,nmets,xmets)) cur -- Curt Seeliger, Data Ranger CSC, EPA/WED contractor 541/754-4638 [EMAIL PROTECTED] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Calculation of group summaries
There was an error in my code (after advising you to use data.frame rather than cbind I used it myself!). Here it is again: nsites - 6 yearList - 1999:2001 fakesub - data.frame( year = rep(yearList, nsites/length(yearList), each = 11), site_id = rep(c('site1','site2'), each = 11*nsites), visit_no = rep(1, 11*2*nsites), transect = rep(LETTERS[1:11], nsites, each = 2), transdir = rep(c('LF','RT'), 11*nsites), undercut = abs(rnorm(11*2*nsites, 10)), angle= runif(11*2*nsites, 0, 180) ) f - function(x) data.frame(year = x[1,1], site_id = x[1,2], visit_no = x[1,3], mean = mean(x[,6]), sd = sd(x[,6]), length = length(x[,6])) do.call(rbind, by(fakesub, fakesub[,1:3], f)) On 7/14/05, Gabor Grothendieck [EMAIL PROTECTED] wrote: 1. Try using more spaces so your code is easier to read. 2. Use data.frame to define your data frame (since the method in your post creates data frames of factors rather than the desired classes). 3. Given the appropriate function, f, a single 'by' statement rbind'ed together, as shown, will create the result. nsites - 6 yearList - 1999:2001 fakesub - data.frame( year = rep(yearList, nsites/length(yearList), each = 11), site_id = rep(c('site1','site2'), each = 11*nsites), visit_no = rep(1, 11*2*nsites), transect = rep(LETTERS[1:11], nsites, each = 2), transdir = rep(c('LF','RT'), 11*nsites), undercut = abs(rnorm(11*2*nsites, 10)), angle= runif(11*2*nsites, 0, 180) ) f - function(x) cbind(year = x[1,1], site_id = x[1,2], visit_no = x[1,3], mean = mean(x[,6]), sd = sd(x[,6]), length = length(x[,6])) do.call(rbind, by(fakesub, fakesub[,1:3], f)) On 7/14/05, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote: Several people suggested specific functions (by, tapply, sapply and others); thanks for not blowing off a simple question regarding how to do the following SQL in R: select year, site_id, visit_no, mean(undercut) AS meanUndercut, count(undercut) AS nUndercut, std(undercut) AS stdUndercut from channelMorphology group by year, site_id, visit_no ; I'd spent quite a bit of time with the suggested functions earlier but had no luck as I'd misread the docs and put the entire dataframe where it only wants the columns to be processed. Sometimes it's the simplest of things. This has lead to another confoundment-- sd() acts differently than mean() for some reason, at least with R 1.9.0. For some reason, means generate NA results and a warning message for each group: argument is not numeric or logical: returning NA in: mean.default(data[x, ], ...) Of course, the argument is numeric, or there'd be no sd value. Or more likely, I'm still missing something really basic. If I wrap the value in as.numeric() things work fine. Why should I have to do this for mean and median, but not sd? The code below should reproduce this error # Fake data for demo: nsites-6 yearList-1999:2001 fakesub-as.data.frame(cbind( year =rep(yearList,nsites/length(yearList),each=11) ,site_id =rep(c('site1','site2'),each=11*nsites) ,visit_no =rep(1,11*2*nsites) ,transect =rep(LETTERS[1:11],nsites,each=2) ,transdir =rep(c('LF','RT'),11*nsites) ,undercut =abs(rnorm(11*2*nsites,10)) ,angle=runif(11*2*nsites,0,180) )) # Create group summaries: sdmets-by(fakesub$undercut ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,sd ) nmets-by(fakesub$undercut ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,length ) xmets-by(fakesub$undercut ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,mean ) xmets-by(as.numeric(fakesub$undercut) ,list(fakesub$year,fakesub$site_id,fakesub$visit_no) ,mean ) # Put site id values (year, site_id and visit_no) into results: # List unique id combinations as a list of lists. Then # reorganize that into 3 vectors for final results. # Certainly, there MUST be a better way... foo-strsplit(unique(paste(fakesub$year ,fakesub$site_id ,fakesub$visit_no ,sep='#')) ,split='#' ) year-list() for(i in 1:length(foo)) {year-rbind(year,foo[[i]][1])} site_id-list() for(i in 1:length(foo)) {site_id-rbind(site_id,foo[[i]][2])} visit_no-list() for(i in 1:length(foo)) {visit_no-rbind(visit_no,foo[[i]][3])} # Final result, more or less data.frame(cbind(a=year,b=site_id,c=visit_no,sdmets,nmets,xmets)) cur -- Curt Seeliger, Data Ranger CSC,
Re: [R] test for difference in the order of self generated sequence
Have you considered glmmPQL in library(MASS)? spencer graves Kaiya Liu wrote: Hello, I have an experiment in which we ask subjects to generate a list of thoughts after being exposed to a stimuli. The thoughts were then coded into two categories (e.g. A B). The objective is to show that the order in which thoughts are generated is affected by the experimental conditions. Can somebody tell me what functions in R can help me do the test or point me to the appropriate reference? Thanks! Example data: Subject 1 in condition 1 generated 5 thoughts. The first, second, and 5th thought is in category A, thought 3 and 4 are in category B. Subject 2 only generated 4 thoughts. ---Condition-A-B n1 11, 2, 5 3, 4 n2 23, 4 1, 2 --- Kaiya Liu __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA [EMAIL PROTECTED] www.pdf.com http://www.pdf.com Tel: 408-938-4420 Fax: 408-280-7915 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] problems with nls function
Hello, I was trying to fit a generalized logistic curve and my code for testing if the function nls is working is like this ### #define the generalized logistic curve function glogit=function(a, b, c, m, t, x) { tmp = 1/(1 + t * exp(-b * (x - m)))^(1/t) model.func = a + tmp * c #define the gradient vector Z = cbind(1, c*(x-m)*tmp^(t+1)*exp(-b*(x-m)), tmp, -c*b*exp(-b*(x-m)) *temp^(t+1), -c*tmp*(1/t)* (log(tmp) + tmp^(-t) * exp(-b(x-m attr(model.func,gradient)=Z model.func } a= 0 b=3 c=6 m=2 t=0.75 x=c(1:10) y=a + c/(1 + t * exp(-b * (x - m)))^(1/t) + rnorm(16, sd=0.01) plot(data.frame(x=x, y=y), type = p, col = red) nls(y ~ glogit(a, b, c, m, t, x), data=data.frame(y=y,x=x), start=list(a=1, b=2, c=3, m=1, t=0, x), trace=T) #End. When I ran this code, I got the error Error in assign(i, temp, envir = env) : attempt to use zero-length variable name. What does this error mean? How should I fix it? Many thanks. Hanna Lu Dept. of Biostatistics Columbia University [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] integer codes of factors
What's the problem? As suggested by the help page, the numeric codes are assigned in the order the names appear in the levels argument. Consider the following example from the help page plus a minor modification: (ff - factor(substring(statistics, 1:10, 1:10), levels=letters)) [1] s t a t i s t i c s Levels: a b c d e f g h i j k l m n o p q r s t u v w x y z as.numeric(ff) [1] 19 20 1 20 9 19 20 9 3 19 (ff - factor(substring(statistics, 1:10, 1:10), levels=letters[3:1])) [1] NA NA aNA NA NA NA NA cNA Levels: c b a as.numeric(ff) [1] NA NA 3 NA NA NA NA NA 1 NA I highly recommend Venable and Ripley (2002) Modern Applied Statistics with S (Springer) and V R (2000) S Programming (Springer). spencer graves Mike R wrote: U = c(b, b, b, c, d, e, e) F1 = factor( U, levels=c(a, b, c, d, e) ) as.numeric(F1) [1] 2 2 2 3 4 5 5 Here, the integer code of b in F1 is 2 K = factor( levels(F1) ) as.numeric(K) [1] 1 2 3 4 5 K [1] a b c d e Levels: a b c d e And again, the integer code of b in K is 2. Great! I am wondering how modify that usage such that the correspondence between the two numeric vectors can this be trusted. for example, the correspondence can be corrupted by placing the a at the end: F2 = factor( U, levels=c(b, c, d, e, a) ) as.numeric(F2) [1] 1 1 1 2 3 4 4 Placing the a at the end changed the integer code of b in F2 to 1, which is not a problem. But .. K = factor( levels(F2) ) as.numeric( K ) [1] 2 3 4 5 1 K [1] b c d e a Levels: a b c d e But the integer code of b in K is now 2, which does not correspond to its code in F2. One would think that ordered=TRUE ought to avoid the corruption, but it does not seem to accomplish that: K = factor( levels(F2), ordered=TRUE ) as.numeric(K) [1] 2 3 4 5 1 K [1] b c d e a Levels: a b c d e But the integer code of b in K is still 2. However, corruption can be avoided with this idiom: K = factor( levels(F2), levels=levels(F2) ) as.numeric(K) [1] 1 2 3 4 5 K [1] b c d e a Levels: b c d e a Now the integer code of b in K is 1, which, as desired, is in correspondence with its code in F2. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA [EMAIL PROTECTED] www.pdf.com http://www.pdf.com Tel: 408-938-4420 Fax: 408-280-7915 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] problems with nls function
Sorry that I specified the x as x=c(0, 0.4, 0.7, 1, 1.4, 1.6, 1.8, 2:10) The error message is the same. Thanks. Hanna - Original Message - From: Yimeng Lu To: r-help@stat.math.ethz.ch Sent: Thursday, July 14, 2005 11:19 PM Subject: problems with nls function Hello, I was trying to fit a generalized logistic curve and my code for testing if the function nls is working is like this ### #define the generalized logistic curve function glogit=function(a, b, c, m, t, x) { tmp = 1/(1 + t * exp(-b * (x - m)))^(1/t) model.func = a + tmp * c #define the gradient vector Z = cbind(1, c*(x-m)*tmp^(t+1)*exp(-b*(x-m)), tmp, -c*b*exp(-b*(x-m)) *temp^(t+1), -c*tmp*(1/t)* (log(tmp) + tmp^(-t) * exp(-b(x-m attr(model.func,gradient)=Z model.func } a= 0 b=3 c=6 m=2 t=0.75 x=c(1:10) y=a + c/(1 + t * exp(-b * (x - m)))^(1/t) + rnorm(16, sd=0.01) plot(data.frame(x=x, y=y), type = p, col = red) nls(y ~ glogit(a, b, c, m, t, x), data=data.frame(y=y,x=x), start=list(a=1, b=2, c=3, m=1, t=0, x), trace=T) #End. When I ran this code, I got the error Error in assign(i, temp, envir = env) : attempt to use zero-length variable name. What does this error mean? How should I fix it? Many thanks. Hanna Lu Dept. of Biostatistics Columbia University [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] A statistical modeling problem
Let's do the simplest things first: I'd start with normal probability plots of log(t_ij). If that all looked sensible, I'd then use lme to model log(t_ij) with subject as a random effect and (probably) que as a fixed effect. Then I'd probably try something like glmmPQL in library(MASS) to leverage whatever I got into logistic regression models with, again, subject as a random effect and que and log(t_ij) as possible explanatory variables. Regarding lme, I highly recommend Pinheiro and Bates (2000) Mixed-Effect Models in S and S-Plus (Springer). For glmmPQL, see Venables and Ripley (2002) Modern Applied Statistics with S, 4th ed. (Springer). I'd also search the archives for generalized linear mixed models, as there are other R packages that offer related capabilities. spencer graves Ravi Varadhan wrote: Hi, This is not an R related question and I apologize for that, but given the brain power of the R community it is hard for me to resist posting this here. I have a problem where each participant is shown a series of visual cues (displayed on a computer screen in a random order) and asked to respond by pressing a button (from a finite number of buttons) that corresponds to the correct answer. The data is of the form (r_ij , t_ij), where r_ij = 1 indicates correct response by the i-th subject on the j-th cue, and t_ij is the time taken to respond. Are there statistical models to describe this set up? Any references or suggestions would be greatly appreciated. Thanks very much, Ravi. [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA [EMAIL PROTECTED] www.pdf.com http://www.pdf.com Tel: 408-938-4420 Fax: 408-280-7915 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] problems with nls function
The start list should only contain parameters, not x. Also your function appears to have multiple errors including reference to temp (presumably intended as tmp?) and b(x-m) which presumably should be b*(x-m). On 7/14/05, Yimeng Lu [EMAIL PROTECTED] wrote: Sorry that I specified the x as x=c(0, 0.4, 0.7, 1, 1.4, 1.6, 1.8, 2:10) The error message is the same. Thanks. Hanna - Original Message - From: Yimeng Lu To: r-help@stat.math.ethz.ch Sent: Thursday, July 14, 2005 11:19 PM Subject: problems with nls function Hello, I was trying to fit a generalized logistic curve and my code for testing if the function nls is working is like this ### #define the generalized logistic curve function glogit=function(a, b, c, m, t, x) { tmp = 1/(1 + t * exp(-b * (x - m)))^(1/t) model.func = a + tmp * c #define the gradient vector Z = cbind(1, c*(x-m)*tmp^(t+1)*exp(-b*(x-m)), tmp, -c*b*exp(-b*(x-m)) *temp^(t+1), -c*tmp*(1/t)* (log(tmp) + tmp^(-t) * exp(-b(x-m attr(model.func,gradient)=Z model.func } a= 0 b=3 c=6 m=2 t=0.75 x=c(1:10) y=a + c/(1 + t * exp(-b * (x - m)))^(1/t) + rnorm(16, sd=0.01) plot(data.frame(x=x, y=y), type = p, col = red) nls(y ~ glogit(a, b, c, m, t, x), data=data.frame(y=y,x=x), start=list(a=1, b=2, c=3, m=1, t=0, x), trace=T) #End. When I ran this code, I got the error Error in assign(i, temp, envir = env) : attempt to use zero-length variable name. What does this error mean? How should I fix it? Many thanks. Hanna Lu Dept. of Biostatistics Columbia University [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Variance components from lm?
Yes, I'm very thankful for the support of Prof. Jonathan Baron and all the others who have contributed their time, creative energies and (for some) money to make R what it is today. It is beyond the budgets of most people on this planet to purchase licenses for every piece of statistical software that purports to do anything useful, not to mention the amount of time required to install and learn how to use fundamentally incompatible software, overcome the sometimes near impossiblity of porting data between the different platforms, etc. It is therefore an enormous contribution to have this common platform that includes capabilities to perform a increasing porportion of the statistical analyses that people in different disciplines have found useful. Indeed, this is not only a service to the statistical community, it is an enormous contribution to the future of humanity, because it makes it easier for people to learn a variety of statistical techniques, thereby making it easier for them to get better answers to the questions that concern them -- and through their increased statistical literacy, making it easier for them effectively pressure others to improve the quality of their problem solving through better collection and analysis of data. spencer graves Liaw, Andy wrote: From: Spencer Graves Under normal circumstances, 'RSiteSearch(variance components)' would likely identify lme in package nlme and lmer in package lme4. I recommend lme, as it comes with a fairly complete set of helper functions described in Pinheiro and Bates (2000) Mixed-Effect Models in S and S-Plus (Springer). lmer is Doug Bates' next generation product currently under development and lacking some of the helper functions available with lme. spencer graves p.s. I said, Under normal circumstances, because when I tried RSiteSearch just now, it didn't work. When I went to www.r-project.org - Search - R site search, I got Computer trouble, July 13, 2005. When it works (which is 99.7% of the time), it is my favorite tool for finding things about R for which I don't already know where to look. The server (graciously created and maintained by Prof. Jonathan Baron) crashed and a hard drive failure. I was told it would likely be back online next Monday. I missed this search engine, too, as I've grown to depend on it quite a bit. Cheers, Andy Steve Ellison wrote: I often use simple nested random-effect models for interlaboratory data. The variance components are important things to know. Is there an R function or package that gets variance components from lm objects? Or can someone point me to a method of doing so? __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Spencer Graves, PhD Senior Development Engineer PDF Solutions, Inc. 333 West San Carlos Street Suite 700 San Jose, CA 95110, USA [EMAIL PROTECTED] www.pdf.com http://www.pdf.com Tel: 408-938-4420 Fax: 408-280-7915 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html -- Notice: This e-mail message, together with any attachment...{{dropped}} __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html