RE: [R] special symobol / character
Hi Eric, Thanks for your speedy reply, I should have noted that I'm using a Linux machine. However, when I copy the symbol from Windows to Linux (using R/emacs) via an x-win 32 window it replaces the per mille symbol with a /211. R then produces the character (not a per mille symbol!) that can be found in the equivalent place using character.table() in the Hmisc package. I'd use windows, except we have huge output datasets generated by a climate model, and its not possible to move it to a machine running windows. Any other suggestions are welcome!, cheers, Matt. -Original Message- From: Eric Lecoutre [mailto:[EMAIL PROTECTED] Sent: Tuesday, 15 February 2005 18:54 To: FISCHER, Matthew; 'r-help@stat.math.ethz.ch' Subject: Re: [R] special symobol / character Hi Matthew, Most systems allow to enter any ASCII (or extended ASCII) character directly using a key combination. Accessing ANSI charcaters under Windows is possible with: ALT+0xxx (press ALT, hold it down, press 0 and the number of the character, release ALT) Thus: ALT+0137 makes: ? The future seems promising with the Unicode support: kudo R core team! Eric At 08:23 15/02/2005, FISCHER, Matthew wrote: Hi all, Is it possible to add a permil (or per mille) symbol to an R plot (I couldn't find this symbol under demo(Hershey) or the plotmath information). In some ascii tables it is symbol no. 137. cheers, Matt. [[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 Eric Lecoutre UCL / Institut de Statistique Voie du Roman Pays, 20 1348 Louvain-la-Neuve Belgium tel: (+32)(0)10473050 [EMAIL PROTECTED] http://www.stat.ucl.ac.be/ISpersonnel/lecoutre If the statistics are boring, then you've got the wrong numbers. -Edward Tufte [[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 :Off topic -- large data sets. Was RE: [R] 64 Bit R Background Question
Hi, Also I agree those cases are relatively rare in STATISTICAL analysis, you can encounter them for simulation topics (natural catalysm a 5 meter in the topographics can change all the simulations) Two ideas (in addition to loading several sections) is 1- to search for duplicate cases and estimate your model upon a frequency weighted shema, perhaps you don't have 200millions different 'cases' 2- take into account your data and model the used algorythm precision/accuracy, (i.e. No need to take into account 1million case, a precision close to .001, if the gradient, or any other function used, has a .01 accuracy) ... Best regards Naji Le 14/02/05 18:41, « Berton Gunter » [EMAIL PROTECTED] a écrit : read all 200 million rows a pipe dream no matter what platform I'm using? In principle R can handle this with enough memory. However, 200 million rows and three columns is 4.8Gb of storage, and R usually needs a few times the size of the data for working space. You would likely be better off not reading the whole data set at once, but loading sections of it from Oracle as needed. -thomas Thomas's comment raises a question: Can comeone give me an example (perhaps in a private response, since I'm off topic here) where one actually needs all cases in a large data set (large being 1e6, say) to do a STATISTICAL analysis? By statistical I exclude, say searching for some particular characteristic like an adverse event in a medical or customer repair database, etc. Maybe a definition of statistical is: anything that cannot be routinely done in a single pass database query. The reason I ask this is that it seems to me that with millions of cases, (careful, perhaps stratified or in some other not completely at random way) sampling should always suffice to reduce a dataset to manageable size sufficient for the data analysis needs at hand. But my ignorance and naivete probably show here. Thanks. -- Bert __ 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] special symobol / character
On Tue, 15 Feb 2005, FISCHER, Matthew wrote: Thanks for your speedy reply, I should have noted that I'm using a Linux machine. However, when I copy the symbol from Windows to Linux (using R/emacs) via an x-win 32 window it replaces the per mille symbol with a /211. \211 I hope. (That is octal for 137.) R then produces the character (not a per mille symbol!) that can be found in the equivalent place using character.table() in the Hmisc package. I'd use windows, except we have huge output datasets generated by a climate model, and its not possible to move it to a machine running windows. Any other suggestions are welcome!, Note that this is not as you claimed an ASCII symbol and it is not even an ISO Latin-1 symbol. The area 128 to 143 is a control area in almost all encodings except WinAnsi, which is what you are using on Windows, I believe. Try `man ascii' on your Linux system. per mille is not AFAICS in the Adobe symbol font that plotmath uses, and so it cannot be added there. In any case, it is text-like, not a symbol (you would want it in the same font as %). It is the Unicode character U+2030, so you will be able to use it in a UTF-8 locale in R-2.1.0 via \u2030. -Original Message- From: Eric Lecoutre [mailto:[EMAIL PROTECTED] Sent: Tuesday, 15 February 2005 18:54 To: FISCHER, Matthew; 'r-help@stat.math.ethz.ch' Subject: Re: [R] special symobol / character Hi Matthew, Most systems allow to enter any ASCII (or extended ASCII) character directly using a key combination. Accessing ANSI charcaters under Windows is possible with: ALT+0xxx (press ALT, hold it down, press 0 and the number of the character, release ALT) Thus: ALT+0137 makes: ? The future seems promising with the Unicode support: kudo R core team! Eric At 08:23 15/02/2005, FISCHER, Matthew wrote: Hi all, Is it possible to add a permil (or per mille) symbol to an R plot (I couldn't find this symbol under demo(Hershey) or the plotmath information). In some ascii tables it is symbol no. 137. -- 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] convolution of functions
Dear sir, we would like to know if there exist any R package containing the computational performance of the n-fold Stieljes' convolution of functions. We look forward to hearing from you. Thank you in advance. M.Luz Gámiz Pérez Dpt. Estadística e Investigación Operativa Facultad de Ciencias Universidad de Granada Telf.: 958-243156 e-mail: [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] update.packages - delete downloaded - N
Hello, Using Update Packages from CRAN in R Version 2.0.1 Patched (2005-01-15) under Windows 98, I found it tricky to save the downloaded files. Even if I answer N to the question, if the downloaded files should be deleted, they are deleted after R is quitted. I understood that I have to copy them to a different location before quitting R. Is this an intended behavoiur of R or did I miss some instruction? Heinz Tüchler __ 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] update.packages - delete downloaded - N
Heinz Tuechler wrote: Hello, Using Update Packages from CRAN in R Version 2.0.1 Patched (2005-01-15) under Windows 98, I found it tricky to save the downloaded files. Even if I answer N to the question, if the downloaded files should be deleted, they are deleted after R is quitted. I understood that I have to copy them to a different location before quitting R. Is this an intended behavoiur of R or did I miss some instruction? Yes, intended, yes, you missed destdir: It is intended to save in tempdir() by default. You can specify argument destdir, if you want something different. Uwe Ligges Heinz Tüchler __ 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] Tests on contingency tables
Dear all, I have a dataset with qualitative variables (factors) and I want to test the null hypothesis of independance between two variables for each pair by using appropriate tests on contingency tables. I first applied chisq.test and obtained dependance in almost all cases with extremely small p-values and warning messages. chisq.test(table(data$ins.f, data$ins.st))$p.val [1] 4.811263e-100 Warning message: Chi-squared approximation may be incorrect in: chisq.test(table(data$ins.f, data$ins.st)) I then turned to Fisher's Exact Test for Count Data, but I got only error messages such as: Error in fisher.test(table(data$ins.f, data$ins.st)) : FEXACT error 501. The hash table key cannot be computed because the largest key is larger than the largest representable int. The algorithm cannot proceed. Reduce the workspace size or use another algorithm. maybe cause the dimensions of contingency tables are too large (?). dim(table(data$ins.f, data$ins.st)) [1] 10 8 I then tried likelihood-ratio G-statistic on contingency table (g.stats() from hierfstat package), as follows: g.stats(data.frame(as.numeric(data$ins.f),as.numeric(data$ins.s)))$g.stats [1] 486.1993 and I replaced in Chi2 distribution function to get p-value: 1-pchisq(486.199, df=63) [1] 0 Is there a better way to perform this or a more appropriate function dedicated to tests on large-dimensioned contingency tables ? Thanks in advance, Jacques VESLOT __ 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] Problem with R CMD Rd2dvi: Rd.sty not found
On Thu, 10 Feb 2005 21:27:03 + (UTC), Gabor Grothendieck [EMAIL PROTECTED] wrote : Tuszynski, Jaroslaw W. JAROSLAW.W.TUSZYNSKI at saic.com writes: : : Hi, : : I run into a problem with R CMD Rd2dvi command: it gives me File `Rd.sty' : not found error (See the output message on the bottom). I get the same : error when running RCMD check. My system is: Windows 2000, R version : 2.0.1, MiKTeX version 2.4. I do have a Rd.sty file in R_HOME/share/texm : directory. : ... There is some information on this at: http://www.murdoch-sutherland.com/Rtools/miktex.html The information there is slightly out of date: the last version of Miktex that I've tried has broken that solution. I asked the author about it, and he replied as below. I think this is pretty poor behaviour on the part of Miktex, but in other respects I prefer it to fptex. For what you're doing, you need to set up your $R_HOME/bin/helpprint.bat file so that it makes use of the TEXINPUTS environment variable. You want the pdftex command to have something like -include-directory=%TEXINPUTS% in it and then it will look in the right places for the includes. Duncan Murdoch To: Duncan Murdoch [EMAIL PROTECTED] Subject: Re: Help with e-tex? From: MiKTeX Support [EMAIL PROTECTED] Date: Sun, 6 Feb 2005 22:04:15 +0100 Duncan Murdoch wrote: Unfortunately, these instructions do not seem to work any more with pdfTeX. Sometime between these two releases: MiKTeX-pdfTeX 2.4.1603 (1.20a-rc4) (MiKTeX 2.4) MiKTeX-pdfTeX 2.4.1817 (1.20b) (MiKTeX 2.4) it seems that support for TEXINPUTS was dropped from pdfTeX. Sorry, I can confirm that TEXINPUTS isn't supported by newer versions of MiKTeX-pdfTeX. No promise that this will be fixed in the 2.4 branch. Sorry! Kind regards, Christian Schenk __ 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] R + MacOSX + Emacs(XEmacs) + ESS
Hi, I try to use Emacs or XEmacs with R in a MacOS X Panter without X11. Anybody can make this work? How? Thanks Ronaldo -- Um mapa-mundi que nao inclua a utopia, nao e digno de consulta, pois deixa de fora as terras a que a Humanidade esta sempre apontando. -- Oscar Wilde -- | // | \\ [***] | ( õ õ ) [Ronaldo Reis Júnior] | V [UFV/DBA-Entomologia] |/ \ [36571-000 Viçosa - MG ] | /(.''`.)\ [Fone: 31-3899-2532 ] | /(: :' :)\ [EMAIL PROTECTED]] |/ (`. `'` ) \[ICQ#: 5692561 | LinuxUser#: 205366 ] |( `- ) [***] | _/ \_Powered by GNU/Debian Woody/Sarge __ 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] R + MacOSX + Emacs(XEmacs) + ESS
On Tue, 2005-02-15 at 10:34 -0200, Ronaldo Reis-Jr. wrote: Hi, I try to use Emacs or XEmacs with R in a MacOS X Panter without X11. Anybody can make this work? Did you try googling for macos X emacs? That's the way you get it. I have found two different versions, both work graphically without X11. ESS installs quite smoothly. Depending on your configuration, you may have to use ESC for Meta instead of Alt of some other systems. So start R in ESS using ESC-R. (The emacs that comes with MacOS X also is GNU Emacs, but works only within terminal window.) cheers, jari oksanen -- Jari Oksanen [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] update.packages - delete downloaded - N
Thank you for the hint regarding the destdir parameter of update.packages. Heinz Tüchler At 12:15 15.02.2005 +0100, Uwe Ligges wrote: Heinz Tuechler wrote: Hello, Using Update Packages from CRAN in R Version 2.0.1 Patched (2005-01-15) under Windows 98, I found it tricky to save the downloaded files. Even if I answer N to the question, if the downloaded files should be deleted, they are deleted after R is quitted. I understood that I have to copy them to a different location before quitting R. Is this an intended behavoiur of R or did I miss some instruction? Yes, intended, yes, you missed destdir: It is intended to save in tempdir() by default. You can specify argument destdir, if you want something different. Uwe Ligges Heinz Tüchler __ 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] using poly in a linear regression in the presence of NA f ails (despite subsetting them out)
My apologies: It's another case of me not thinking statistically... It may also help those of us whose brains run at slow clock speeds to have ?poly, ?bs and ?ns mention how they react to NAs. Best, Andy From: Prof Brian Ripley Andy, I don't think it is a bug. The problem is that poly(x, 2) depends on the possible set of x values, and so needs to know all of them, unlike e.g. log(x) which is observation-by-observation. Silently omitting missing values is not a good idea in such cases, especially if the values are missing in other variables (which is what na.action is likely to do). I would say models with poly, ns, bs etc are inadvisable in the presence of missing values in their argument. We could make poly() give an informative message, though. Brian On Mon, 14 Feb 2005, Liaw, Andy wrote: This smells like a bug to me. The error is triggered by the line: variables - eval(predvars, data, env) inside model.frame.default(). At that point, na.action has not been applied, so poly() ended being called on data that still contains missing values. The qr() that issued the error is for generating the orthogonal basis when evaluating poly(), not for fitting the linear model itself. Essentially, calling model.frame(y ~ poly(x, 2), data=data.frame(x=c(NA, 1:3), y=rnorm(4)), na.action=na.omit) would show the same error. The obvious workaround is to omit cases with NAs before calling lm(). Andy From: Markus Jäntti I ran into a to me surprising result on running lm with an orthogonal polynomial among the predictors. The lm command resulted in Error in qr(X) : NA/NaN/Inf in foreign function call (arg 1) Error during wrapup: despite my using a subset in the call to get rid of NA's. poly is apparently evaluated before any NA's are subsetted out of the data. Example code (attached to this e-mail as as a script): ## generate some data n - 50 x - runif(n) a0 - 10 a1 - .5 sigma.e - 1 y - a0 + a1*x + rnorm(n)*sigma.e tmp.d - data.frame(y, x) rm(list=c(n, x, a0, a1, sigma.e, y)) print(lm.1 - lm(y ~ poly(x, 2), data = tmp.d) + + ## now make a few NA's + + tmp.d$x[1:2] - rep(NA, 2) Error: syntax error Error during wrapup: ## this fails, just as it should print(lm.1 - lm(y ~ poly(x, 2), data = tmp.d)) Call: lm(formula = y ~ poly(x, 2), data = tmp.d) Coefficients: (Intercept) poly(x, 2)1 poly(x, 2)2 10.380 -0.242 -1.441 ## these also fail, but should not? print(lm.2 - lm(y ~ poly(x, 2), data = tmp.d, subset = !is.na(x))) Call: lm(formula = y ~ poly(x, 2), data = tmp.d, subset = !is.na(x)) Coefficients: (Intercept) poly(x, 2)1 poly(x, 2)2 10.380 -0.242 -1.441 print(lm.3 - lm(y ~ poly(x, 2), data = tmp.d, na.action = na.omit)) Call: lm(formula = y ~ poly(x, 2), data = tmp.d, na.action = na.omit) Coefficients: (Intercept) poly(x, 2)1 poly(x, 2)2 10.380 -0.242 -1.441 ## but this works print(lm.3 - lm(y ~ poly(x, 2), data = subset(tmp.d, subset = !is.na(x Call: lm(formula = y ~ poly(x, 2), data = subset(tmp.d, subset = !is.na(x))) Coefficients: (Intercept) poly(x, 2)1 poly(x, 2)2 10.380 -0.242 -1.441 The documentation of lm is *not* misleading at this point, saying that subset an optional vector specifying a subset of observations to be used in the fitting process. which implies that data are subsetted once lm.fit is called. All the same, this behavior is a little unexpected to me. Is it to be considered a feature, that is, does it produce beneficial side effects which explain why it works as it does? Regards, Markus I am running R on a Debian testing system with kernel 2.6.10 and version _ platform i386-pc-linux-gnu arch i386 os linux-gnu system i386, linux-gnu status major2 minor0.1 year 2004 month11 day 15 language R -- Markus Jantti Abo Akademi University [EMAIL PROTECTED] http://www.iki.fi/~mjantti __ 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!
Re: [R] Tests on contingency tables
You can test independence via a log-linear model. More importantly, you can model that dependence and learn something useful about the data. I don't see your point here: the two factors are clearly highly dependent: who cares what the exact p value is? Did you do e.g. a mosaicplot as I suspect the dependence is obvious in any reasonable plot? On Tue, 15 Feb 2005, Jacques VESLOT wrote: Dear all, I have a dataset with qualitative variables (factors) and I want to test the null hypothesis of independance between two variables for each pair by using appropriate tests on contingency tables. I first applied chisq.test and obtained dependance in almost all cases with extremely small p-values and warning messages. chisq.test(table(data$ins.f, data$ins.st))$p.val [1] 4.811263e-100 Warning message: Chi-squared approximation may be incorrect in: chisq.test(table(data$ins.f, data$ins.st)) I then turned to Fisher's Exact Test for Count Data, but I got only error messages such as: Error in fisher.test(table(data$ins.f, data$ins.st)) : FEXACT error 501. The hash table key cannot be computed because the largest key is larger than the largest representable int. The algorithm cannot proceed. Reduce the workspace size or use another algorithm. maybe cause the dimensions of contingency tables are too large (?). The help file does says Note this fails (with an error message) when the entries of the table are too large. Note, the _entries_, not the dimensions. The issue is how many tables need to be enumerated. dim(table(data$ins.f, data$ins.st)) [1] 10 8 I then tried likelihood-ratio G-statistic on contingency table (g.stats() from hierfstat package), as follows: g.stats(data.frame(as.numeric(data$ins.f),as.numeric(data$ins.s)))$g.stats [1] 486.1993 and I replaced in Chi2 distribution function to get p-value: 1-pchisq(486.199, df=63) [1] 0 Is there a better way to perform this or a more appropriate function dedicated to tests on large-dimensioned contingency tables ? -- 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] Correct effect plots from lme() objects
Hello all! R2.0.1, W2k I posted this question to the list last Sunday without getting any replies on the list. I got two off the list though, suggesting me to plot manually as a second step, from estimable() or intervals() objects respectively. As this was not really what I wished for, I take the risk to upset somebody with a trivial question, and re-post it (just a little edited). So, here it is again: I use lme() from nlme to make mean estimates from series of field experiments where not all treatments (like variety) are present in all experiments. (An old problem in variety evaluation). A typical call is: yield.lme - lme(Yield ~ Variety, data = data, na.action = na.omit, random = ~ 1 | Experiment/Block) This works well, even when observations are lacking. I have checked against the accepted method for doing this in Sweden, which is PROC MIXED in SAS, and the fitted fixed effects are more or less identical. I use estimable() from gmodels (gregmisc package) to extract estimates, standard errors and such. I use matrices with the variety names as row names, it works smooth. What I am unable to, as yet, is to make nice plots of the estimates for a given set of varieties. To use only the fixed call directly on the dataset works for many plooting functions, but produces the wrong graphs, as the structure is not used. The structure (the random call) has to be used, as there are NA:s in the dataset. Pinheiro Bates have a lot of graphics on lme objects, but they try to illustrate more sophisticated relations than my need. I´ve looked through gplots and the graphic parts of nlme without any hits. Probably, my difficulties are just due to my own lack of skill. Some standard plotting facility plotting directly from the lme object ought to work, but I don´t understand how. How should I do this? Cheers /CG -- CG Pettersson MSci. PhD.Stud. Swedish University of Agricultural Sciences (SLU) Dep. of Ecology and Crop production sciences (EVP). http://www.slu.se/ [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] Tests on contingency tables
I forgot to mention a crucial statistical point: As you are doing this pairwise, remember Simpson's paradox. or if you don't know about it, Google it (it is not really due to Simpson, an instance of Stigler's Law of Eponymy). On Tue, 15 Feb 2005, Prof Brian Ripley wrote: You can test independence via a log-linear model. More importantly, you can model that dependence and learn something useful about the data. I don't see your point here: the two factors are clearly highly dependent: who cares what the exact p value is? Did you do e.g. a mosaicplot as I suspect the dependence is obvious in any reasonable plot? On Tue, 15 Feb 2005, Jacques VESLOT wrote: Dear all, I have a dataset with qualitative variables (factors) and I want to test the null hypothesis of independance between two variables for each pair by using appropriate tests on contingency tables. I first applied chisq.test and obtained dependance in almost all cases with extremely small p-values and warning messages. chisq.test(table(data$ins.f, data$ins.st))$p.val [1] 4.811263e-100 Warning message: Chi-squared approximation may be incorrect in: chisq.test(table(data$ins.f, data$ins.st)) I then turned to Fisher's Exact Test for Count Data, but I got only error messages such as: Error in fisher.test(table(data$ins.f, data$ins.st)) : FEXACT error 501. The hash table key cannot be computed because the largest key is larger than the largest representable int. The algorithm cannot proceed. Reduce the workspace size or use another algorithm. maybe cause the dimensions of contingency tables are too large (?). The help file does says Note this fails (with an error message) when the entries of the table are too large. Note, the _entries_, not the dimensions. The issue is how many tables need to be enumerated. dim(table(data$ins.f, data$ins.st)) [1] 10 8 I then tried likelihood-ratio G-statistic on contingency table (g.stats() from hierfstat package), as follows: g.stats(data.frame(as.numeric(data$ins.f),as.numeric(data$ins.s)))$g.stats [1] 486.1993 and I replaced in Chi2 distribution function to get p-value: 1-pchisq(486.199, df=63) [1] 0 Is there a better way to perform this or a more appropriate function dedicated to tests on large-dimensioned contingency tables ? -- 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 -- 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] R, MacOSX G5 bi-proc
You can use the snow package on CRAN. I don't know if RPVM/RMPI work on the Macs, but they are also on CRAN. Andy From: Naji Hi all, I've downloaded binary files from CRAN to install R and it's working quite fine..I'm wandering if I can take advantage of two processors from the Mac. Where can I find lecture/help for this? Thanks in advance Naji __ 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] memory problem with package mix
Hello, I think we have a memory problem with em.mix. We have done: library(mix) Manq - read.table(C:/.../file.txt) attach(Manq) Manq V1 V2 V3 V4 .V27 1 1 1 1 1... 2 1 NA 3 6 3 1 2 6 2 ... ... 300 2 NA 6 2... Essaimanq -prelim.mix(as.matrix(Manq),5) test - em.mix(Essaimanq) error cannot allocated vector of size 535808 KB in addition : warning message reached total allocation of 509MB Thank you __ 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] matlab norm(h) command in R: sqrt(sum(h^2)) - use in an expression
Hi in matlab I defined a function (double gamma, parameters at the end of this mail) as h(i)=((t/d1)^a1)*exp(-(t-d1)/b1)-c*((t/d2)^a2)*exp(-(t-d2)/b2); h=h/norm(h); I do know that norm() in matlab is equal to: sqrt(sum(x^2)) in R so in R I do it like: #function (double gamama) h - expression((t/d1)^a1*exp(-(t-d1)/b1)-c*(t/d2)^a2*exp(-(t-d2)/b2)) # plot it t - seq(0, 2, by = 100) t - t/1000 plot(eval(h), type = 'l') # however this yields an error h - h/sqrt(sum(h^2)) Error in h^2 : non-numeric argument to binary operator what shall I do to get the matlab: h = h/norm(h) implemented in R? thanks for a hint christoph # parameters peak1 - 5.4 fwhm1 - 5.2 peak2 - 10.8 fwhm2 - 7.35 dip - 0.35 b1 - 0.9 # dispersion b2 - 0.9 #dispersion a1 - peak1/b1 a2 - peak2/b2 d1 - a1*b1 d2 - a2*b2 c - dip __ 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] matlab norm(h) command in R: sqrt(sum(h^2)) - use in an e xpression
From: Christoph Lehmann Hi in matlab I defined a function (double gamma, parameters at the end of this mail) as h(i)=((t/d1)^a1)*exp(-(t-d1)/b1)-c*((t/d2)^a2)*exp(-(t-d2)/b2); h=h/norm(h); I do know that norm() in matlab is equal to: sqrt(sum(x^2)) in R so in R I do it like: #function (double gamama) h - expression((t/d1)^a1*exp(-(t-d1)/b1)-c*(t/d2)^a2*exp(-(t-d2)/b2)) # plot it t - seq(0, 2, by = 100) t - t/1000 plot(eval(h), type = 'l') # however this yields an error h - h/sqrt(sum(h^2)) ^ ^ Those are still expressions, not vectors. You need to wrap those in eval() as you did in plot(), or do h - eval(h) beforehand. Andy Error in h^2 : non-numeric argument to binary operator what shall I do to get the matlab: h = h/norm(h) implemented in R? thanks for a hint christoph # parameters peak1 - 5.4 fwhm1 - 5.2 peak2 - 10.8 fwhm2 - 7.35 dip - 0.35 b1 - 0.9 # dispersion b2 - 0.9 #dispersion a1 - peak1/b1 a2 - peak2/b2 d1 - a1*b1 d2 - a2*b2 c - dip __ 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] Tests on contingency tables
Thanks a lot for your help ! Right ! According to tables, most factors look indeed highly dependent... but, because of strange p-values and warning messages, as I tried to test it with Chisquare test, and because Fisher's Exact Test function doesn't work on my data, I wondered whether there were other functions to perform such tests. I will try with test independence via a log-linear model. Is this code correct ? (I can't catch exactly how to put 'formula' argument) z - table(data$fac1, data$fac2) names(dimnames(z)) - c(fac1, fac2) fm - loglm(~ins+pl,z) fm Call: loglm(formula = ~ins + pl, data = z) Statistics: X^2 df P( X^2) Likelihood Ratio 286.1236 490 Pearson 450.5332 490 Jacques -Message d'origine- De : Prof Brian Ripley [mailto:[EMAIL PROTECTED] Envoye : mardi 15 fevrier 2005 17:41 A : Jacques VESLOT Cc : R-HELP; [EMAIL PROTECTED] Objet : Re: [R] Tests on contingency tables You can test independence via a log-linear model. More importantly, you can model that dependence and learn something useful about the data. I don't see your point here: the two factors are clearly highly dependent: who cares what the exact p value is? Did you do e.g. a mosaicplot as I suspect the dependence is obvious in any reasonable plot? On Tue, 15 Feb 2005, Jacques VESLOT wrote: Dear all, I have a dataset with qualitative variables (factors) and I want to test the null hypothesis of independance between two variables for each pair by using appropriate tests on contingency tables. I first applied chisq.test and obtained dependance in almost all cases with extremely small p-values and warning messages. chisq.test(table(data$ins.f, data$ins.st))$p.val [1] 4.811263e-100 Warning message: Chi-squared approximation may be incorrect in: chisq.test(table(data$ins.f, data$ins.st)) I then turned to Fisher's Exact Test for Count Data, but I got only error messages such as: Error in fisher.test(table(data$ins.f, data$ins.st)) : FEXACT error 501. The hash table key cannot be computed because the largest key is larger than the largest representable int. The algorithm cannot proceed. Reduce the workspace size or use another algorithm. maybe cause the dimensions of contingency tables are too large (?). The help file does says Note this fails (with an error message) when the entries of the table are too large. Note, the _entries_, not the dimensions. The issue is how many tables need to be enumerated. dim(table(data$ins.f, data$ins.st)) [1] 10 8 I then tried likelihood-ratio G-statistic on contingency table (g.stats() from hierfstat package), as follows: g.stats(data.frame(as.numeric(data$ins.f),as.numeric(data$ins.s)))$g.stats [1] 486.1993 and I replaced in Chi2 distribution function to get p-value: 1-pchisq(486.199, df=63) [1] 0 Is there a better way to perform this or a more appropriate function dedicated to tests on large-dimensioned contingency tables ? -- 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] matlab norm(h) command in R: sqrt(sum(h^2)) - use in anexpression
you are trying to square an expression not an number! try this: h.fun - function(tt) ((tt/d1)^a1) * exp(-(tt-d1)/b1) - cc*((tt/d2)^a2) * exp(-(tt-d2)/b2) h - h.fun(tt) h - h/sqrt(sum(h*h)) I hope it helps. Best, Dimitris Dimitris Rizopoulos Ph.D. Student Biostatistical Centre School of Public Health Catholic University of Leuven Address: Kapucijnenvoer 35, Leuven, Belgium Tel: +32/16/336899 Fax: +32/16/337015 Web: http://www.med.kuleuven.ac.be/biostat/ http://www.student.kuleuven.ac.be/~m0390867/dimitris.htm - Original Message - From: Christoph Lehmann [EMAIL PROTECTED] To: R-help@stat.math.ethz.ch Sent: Tuesday, February 15, 2005 3:45 PM Subject: [R] matlab norm(h) command in R: sqrt(sum(h^2)) - use in anexpression Hi in matlab I defined a function (double gamma, parameters at the end of this mail) as h(i)=((t/d1)^a1)*exp(-(t-d1)/b1)-c*((t/d2)^a2)*exp(-(t-d2)/b2); h=h/norm(h); I do know that norm() in matlab is equal to: sqrt(sum(x^2)) in R so in R I do it like: #function (double gamama) h - expression((t/d1)^a1*exp(-(t-d1)/b1)-c*(t/d2)^a2*exp(-(t-d2)/b2)) # plot it t - seq(0, 2, by = 100) t - t/1000 plot(eval(h), type = 'l') # however this yields an error h - h/sqrt(sum(h^2)) Error in h^2 : non-numeric argument to binary operator what shall I do to get the matlab: h = h/norm(h) implemented in R? thanks for a hint christoph # parameters peak1 - 5.4 fwhm1 - 5.2 peak2 - 10.8 fwhm2 - 7.35 dip - 0.35 b1 - 0.9 # dispersion b2 - 0.9 #dispersion a1 - peak1/b1 a2 - peak2/b2 d1 - a1*b1 d2 - a2*b2 c - dip __ 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] D(eval(g)) problem, since Function `eval' is not in the derivatives table
thanks Andy and Dimitris for your reply to my expression/eval - problem starting with the resulting expression g I need g's derivative as expression, but I get: Function `eval' is not in the derivatives table: #function (double gamama) h - expression((t/d1)^a1*exp(-(t-d1)/b1)-c*(t/d2)^a2*exp(-(t-d2)/b2)) # plot it t - seq(0, 2, by = 100) t - t/1000 g - expression(eval(h)/sqrt(sum(eval(h)^2))) plot(eval(g), type = 'l') g.deriv - D(g, t) Error in D(g, t) : Function `eval' is not in the derivatives table is there any way one can solve this problem? thanks a lot christoph ## -- ## parameters peak1 - 5.4 fwhm1 - 5.2 peak2 - 10.8 fwhm2 - 7.35 dip - 0.35 b1 - 0.9 # dispersion b2 - 0.9 #dispersion a1 - peak1/b1 a2 - peak2/b2 d1 - a1*b1 d2 - a2*b2 c - dip __ 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] convolution of functions
If you mean the convolution (f*g)(x) = integral f(x-y)g(y) dy for integrable functions f and g on R^n, then I think using the fact that the Fourier transform of the convolution is the product of the Fourier transforms of f and g is the most efficient approach, unless f or g have some special structure. For this you just need fft() in base R. You do have to do a little bookkeeping to manage the discretizations. Reid Huntsinger -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Mari Luz Gamiz Perez Sent: Tuesday, February 15, 2005 5:19 AM To: r-help@stat.math.ethz.ch Subject: [R] convolution of functions Dear sir, we would like to know if there exist any R package containing the computational performance of the n-fold Stieljes' convolution of functions. We look forward to hearing from you. Thank you in advance. M.Luz Gámiz Pérez Dpt. Estadística e Investigación Operativa Facultad de Ciencias Universidad de Granada Telf.: 958-243156 e-mail: [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] Correct effect plots from lme() objects
On Tuesday 15 February 2005 08:15, CG Pettersson wrote: Hello all! R2.0.1, W2k I posted this question to the list last Sunday without getting any replies on the list. I got two off the list though, suggesting me to plot manually as a second step, from estimable() or intervals() objects respectively. As this was not really what I wished for, I take the risk to upset somebody with a trivial question, and re-post it (just a little edited). So, here it is again: I use lme() from nlme to make mean estimates from series of field experiments where not all treatments (like variety) are present in all experiments. (An old problem in variety evaluation). A typical call is: yield.lme - lme(Yield ~ Variety, data = data, na.action = na.omit, random = ~ 1 | Experiment/Block) This works well, even when observations are lacking. I have checked against the accepted method for doing this in Sweden, which is PROC MIXED in SAS, and the fitted fixed effects are more or less identical. I use estimable() from gmodels (gregmisc package) to extract estimates, standard errors and such. I use matrices with the variety names as row names, it works smooth. What I am unable to, as yet, is to make nice plots of the estimates for a given set of varieties. To use only the fixed call directly on the dataset works for many plooting functions, but produces the wrong graphs, as the structure is not used. The structure (the random call) has to be used, as there are NA:s in the dataset. It's not clear to me what you are looking for. In particular, exactly what 'estimates' do you want to plot? Could you give us a reproducible example (preferably using a dataset from nlme) and describe the plot you are looking for in terms of that example? Deepayan Pinheiro Bates have a lot of graphics on lme objects, but they try to illustrate more sophisticated relations than my need. I´ve looked through gplots and the graphic parts of nlme without any hits. Probably, my difficulties are just due to my own lack of skill. Some standard plotting facility plotting directly from the lme object ought to work, but I don´t understand how. How should I do this? Cheers /CG __ 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] shrinkage estimates in lme
Hello. Slope estimates in lme are shrinkage estimates which pull the OLS slope estimates towards the population estimates, the degree of which depends on the group sample size and the distance between the group-based estimate and the overall population estimate. Although these shrinkage estimates as said to be more precise with respect to the true values, they are also biased. So there is a tradeoff between precision and bias. Are there rules of thumb to help determine when it is better to use the OLS slope estimates and when to use the mixed model (lme) shrinkage estimates? I have 35 groups but the numbers per group vary from over 50 to as low as 4. Thanks for any help. Bill Shipley Subject Matter Editor, Ecology North American Editor, Annals of Botany Département de biologie, Université de Sherbrooke, Sherbrooke (Québec) J1K 2R1 CANADA [EMAIL PROTECTED] http://callisto.si.usherb.ca:8080/bshipley/ http://callisto.si.usherb.ca:8080/bshipley/ [[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] lattice multiple plots per page
On Monday 14 February 2005 18:12, Jeff Jorgensen wrote: Dear R-sters, I was wondering if anyone has encountered the following issues. I've figured out how to get multiple levelplots [library(lattice)] on a single plot. However, when I add text (adding axis labels for the entire four panel plot) the text is missing when I insert the *.eps file I've created into my LaTeX document (via MikTeX-WinEdt). And, I've just upgraded to R v2.0.1 from v1.8.1 (Windows), and each individual levelplot is smaller compared to the older R release. Any clues as to 1) how I can recover the lost text, and Your use is incorrect, in the sense that the return value of grid.locator() is not a valid input of ltext. You could do the following to fix this, library(grid) ltext(lapply(grid.locator(), convertX, native, TRUE), lab=x-axis label, where I click,cex=1.5) ltext(lapply(grid.locator(), convertX, native, TRUE), lab = y-axis label, where I click, cex=1.5, srt=90) but that doesn't seem to help either. So, there seems to be a bug somewhere. However, you presumably want something like grid.text(x = .05, y = .5, lab=y-axis label, default.units = npc, gp = gpar(cex=1.5), rot = 90) grid.text(x = .5, y = .05, lab=x-axis label, default.units = npc, gp = gpar(cex=1.5)) for the last two steps, which seems to work. 2) increase the size of each of the levelplots? You need to change the following settings, in particular the entries that end with 'padding' (similarly for 'layout.heights'). str(trellis.par.get(layout.widths)) List of 13 $ left.padding : num 1 $ key.left : num 1 $ key.ylab.padding : num 1 $ ylab : num 1 $ ylab.axis.padding: num 1 $ axis.left: num 1 $ axis.panel : num 1 $ panel: num 1 $ between : num 1 $ axis.right : num 1 $ axis.key.padding : num 1 $ key.right: num 1 $ right.padding: num 1 Deepayan Cheers, Jeff Jorgensen ~~~ Sample code that illustrates what I'm trying to do: #create a levelplot x-seq(pi/4, 5*pi, length = 100) y-seq(pi/4, 5*pi, length = 100) r-as.vector(sqrt(outer(x^2, y^2, +))) grid-expand.grid(x=x, y=y) grid$z-cos(r^2) * exp(-r/(pi^3)) a-levelplot(z~x*y, grid, cuts = 50, xlab=, ylab=, colorkey = FALSE) #create the multiple panel plot, here using all the same levelplot trellis.par.set(list(background=list(col=white))) #white background #using position to scale the plots up and to the right ~10% #to make room for the axis labels print(a,position=c(0.1,0.1,1,1),split=c(1,1,2,2),more=T) print(a,position=c(0.1,0.1,1,1),split=c(1,2,2,2),more=T) print(a,position=c(0.1,0.1,1,1),split=c(2,1,2,2),more=T) print(a,position=c(0.1,0.1,1,1),split=c(2,2,2,2),more=F) #commands that let you click where you want the labels centered ltext(grid::grid.locator(),lab=x-axis label, where I click,cex=1.5) ltext(grid::grid.locator(),lab=y-axis label, where I click,cex=1.5,srt=90) #save device to an *.eps file, to be called later by a \includegraphics command dev.copy2eps(file=twobytwoplot.eps) __ 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 with package mix
On 15-Feb-05 [EMAIL PROTECTED] wrote: Hello, I think we have a memory problem with em.mix. We have done: library(mix) Manq - read.table(C:/.../file.txt) attach(Manq) Manq V1 V2 V3 V4 .V27 1 1 1 1 1... 2 1 NA 3 6 3 1 2 6 2 ... ... 300 2 NA 6 2... Essaimanq -prelim.mix(as.matrix(Manq),5) test - em.mix(Essaimanq) error cannot allocated vector of size 535808 KB in addition : warning message reached total allocation of 509MB Hmm. According to the above, it seems you might have 5 categorical variables V1...V5 with at least 6 levels, so since your call to em.mix does not specify any model restriction (for which you need to call ecm.mix insead) you may have at least 6^5 = 46656 cells for the different combinations of levels. This will require 46655 parameters for the probabilities of these cells. For each cell, you have a separate vector of means for the multivariate normal distribution to be fitted to the (27-5)=22 continuous variables. This requires 22*46656 = 1026432 parameters. Sub-total: 1073087 Then, as a bit of sugar on the cake, you have the 22*21/2 = 121 parameters for the covariance matrix. Sub-total: 1073208 Since em.mix does quite complicated things, it is perhaps not surpising that it demands more than 509MB (corresponding to about 500 bytes per parameter or, with 8 bytes per number, about 60 numbers per parameter). Not to mention the 8100 numbers (about 65000 bytes) required for each working copy of the representation of the data. In any case, apparently you only have 300*27 = 8100 data, quite inadequate for this unrestricted model! Even if you could have allocated enough memory, you would then have found that the EM fit would not get anywhere. Suggested solution: think about restricting the number of parameters in the model, using the parameter margins to ecnm.mix to restrict the number of independent combinations of categorical levels, and also design to specify a simpler model for the dependence of the continuous variables on the categoricals (e.g. the matrix corresponding to the model ~ V1+V2+V3+V4+V5 only introduces 5*6*22=660 new parameters, namely a simple additive effect of level of each Vi on the mean of each of the 22 continuous variables). Hoping this helps, Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Feb-05 Time: 16:24:18 -- XFMail -- __ 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] how many 7th of the month is there between two dates
This is a eaeir way to ask my prior question: I want to caculate how many an exact day of the month there is between two dates. For example; How many 7th of the month is there between 1998/12/17 and 2000/1/7. To make the problem simple, the day of the month (7) is the day in the 2nd date. __ 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] reading a 40kb csv file in R
I am trying to read a matrix of 5000*2000 from a csv file of real numbers in R using the scan function. I am getting an error saying that R can not read a vector 32KB and it can't allocate a memory of 191 MB etc. What should I do. I am using R1.7.1. Thanks = Thanks Fairouz Makhlouf __ 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] shrinkage estimates in lme
What exactly do you mean by slope estimates? Marginal or subject-specific? Presuming that you mean the fixed-effects (since you refer to OLS) then these are unbiased since they are Weighted Least Squares estimators based on the marginal model. They are unbiased even in the case where you misspecify the correlation structure. Moreover if you use a reasonably well chosen covariance structure then they are also very efficient. The subject-specific fitted values are shrunken toward the mean in the sense: \hat{y}_i = \sum_i V_i^{-1}X_i\hat{\bfbeta} + (I-\sum_i V_i^{-1})y_i and thus it is a weighted average of the population average profile X_i\hat{\bfbeta} and the observed data y_i with weights \sum_i V_i^{-1} and (I-\sum_i V_i^{-1}), respectively (where \sum_i V_i^{-1} is the residual covariance matrix). I hope this helps. Best, Dimitris Dimitris Rizopoulos Ph.D. Student Biostatistical Centre School of Public Health Catholic University of Leuven Address: Kapucijnenvoer 35, Leuven, Belgium Tel: +32/16/336899 Fax: +32/16/337015 Web: http://www.med.kuleuven.ac.be/biostat/ http://www.student.kuleuven.ac.be/~m0390867/dimitris.htm - Original Message - From: Bill Shipley [EMAIL PROTECTED] To: R help list r-help@stat.math.ethz.ch Sent: Tuesday, February 15, 2005 5:17 PM Subject: [R] shrinkage estimates in lme Hello. Slope estimates in lme are shrinkage estimates which pull the OLS slope estimates towards the population estimates, the degree of which depends on the group sample size and the distance between the group-based estimate and the overall population estimate. Although these shrinkage estimates as said to be more precise with respect to the true values, they are also biased. So there is a tradeoff between precision and bias. Are there rules of thumb to help determine when it is better to use the OLS slope estimates and when to use the mixed model (lme) shrinkage estimates? I have 35 groups but the numbers per group vary from over 50 to as low as 4. Thanks for any help. Bill Shipley Subject Matter Editor, Ecology North American Editor, Annals of Botany Dpartement de biologie, Universit de Sherbrooke, Sherbrooke (Qubec) J1K 2R1 CANADA [EMAIL PROTECTED] http://callisto.si.usherb.ca:8080/bshipley/ http://callisto.si.usherb.ca:8080/bshipley/ [[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
Off topic -- large data sets. Was RE: [R] 64 Bit R Background Question
In message [EMAIL PROTECTED], r-help- [EMAIL PROTECTED] writes Can comeone give me an example (perhaps in a private response, since I'm off topic here) where one actually needs all cases in a large data set (large being 1e6, say) to do a STATISTICAL analysis? By statistical I exclude, say searching for some particular characteristic like an adverse event in a medical or customer repair database, etc. Maybe a definition of statistical is: anything that cannot be routinely done in a single pass database query. If the dimensionality of the data is large, you may need a large number of cases too. An example from my own experience would be using quadratic discriminant analysis (with regularization) for classifying symbols for an OCR program. With 200 classes and 100 features, I'd really like many millions of cases. I've been using about 20,000 per class or 4 million in total, but if I had 40 million it would probably work better. Compared to many applications in pattern recognition and data mining, I think this is a fairly small example. -- Graham Jones, author of SharpEye Music Reader http://www.visiv.co.uk 21e Balnakeil, Durness, Lairg, Sutherland, IV27 4PT, Scotland, UK __ 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 many 7th of the month is there between two dates
Omar Lakkis uofiowa at gmail.com writes: : I want to caculate how many an exact day of the month there is between : two dates. : : For example; How many 7th of the month is there between 1998/12/17 : and 2000/1/7. To make the problem simple, the day of the month (7) : is the day in the 2nd date. d1 - as.Date(1998/12/17) d2 - as.Date(2000/1/7) day - function(x) as.numeric(format(x, %d)) sum(day(seq(from = d1, to = d2, by = day)) == day(d2)) __ 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] nlsList vs nlme with lsoda function
Hello list members, I'm trying to get nlme to work with lsoda from odesolve. Currently, I can use nlsList to fit a simple logistic growth model to some simulated data, but I get the following error message with nlme: Error in model.frame(formula, rownames, variables, varnames, extras, extranames, : invalid variable type The commands I'm trying are below. Any suggestions for how to get this to work? Thanks! This works: fit.nlsList-nlsList(xvals~lsoda(x0,times,Logist, +c(r=r,K=K,x0=x0))[,2]|group, +start=list(r=1,K=500,x0=2),data=data.group) This returns the error message above: fit.nlme-nlme(xvals~(lsoda(x0,times,Logist, +c(r=r,K=K,x0=x0))[,2]), +start=list(r=1,K=500,x0=2),data=data.group, +fixed=r+K+x0~1,random=x0~1) The model function is: Logist=function(t,x,parms) { N1-x[1] with(as.list(parms),{ dN1=r*N1*(1-(N1/K)) list(c(dN1)) })} My data set looks like: Grouped Data: xvals ~ times | group xvals times group 1-0.44543969 0 1 238.86411972 3 1 3 310.77106961 6 1 4 486.78653327 9 1 5 0.40613173 0 2 634.94635643 3 2 7 307.81884870 6 2 8 486.46098417 9 2 etc... __ 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: Off topic -- large data sets. Was RE: [R] 64 Bit R Background Question
On Tue, 15 Feb 2005, Graham Jones wrote: In message [EMAIL PROTECTED], r-help- [EMAIL PROTECTED] writes [Actually quoting Bert Gunter, BTW] Can comeone give me an example (perhaps in a private response, since I'm off topic here) where one actually needs all cases in a large data set (large being 1e6, say) to do a STATISTICAL analysis? By statistical I exclude, say searching for some particular characteristic like an adverse event in a medical or customer repair database, etc. Maybe a definition of statistical is: anything that cannot be routinely done in a single pass database query. If the dimensionality of the data is large, you may need a large number of cases too. An example from my own experience would be using quadratic discriminant analysis (with regularization) for classifying symbols for an OCR program. With 200 classes and 100 features, I'd really like many millions of cases. I've been using about 20,000 per class or 4 million in total, but if I had 40 million it would probably work better. Compared to many applications in pattern recognition and data mining, I think this is a fairly small example. But Bert's caveats apply: you have 200 problems of size 20,000 since in QDA each class's distribution is estimated separately, and a single pass will give you the sufficient statistics however large the dataset is. -- 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] how many 7th of the month is there between two dates
It is really easy using R's date and datetime classes: st - as.Date(1998-12-17) en - as.Date(2000-1-7) st0 - st # Get the 7th of the month before st0 - as.POSIXlt(st); st0$mday - 7; st0$mon - st0$mon - 1 st0 - as.Date(st0) ll - seq.Date(st0, en, by=month) sum(ll as.Date(st) ll en) and use = etc depending what you mean by `between'. On Tue, 15 Feb 2005, Omar Lakkis wrote: This is a eaeir way to ask my prior question: I want to caculate how many an exact day of the month there is between two dates. For example; How many 7th of the month is there between 1998/12/17 and 2000/1/7. To make the problem simple, the day of the month (7) is the day in the 2nd date. -- 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] Problem with R CMD Rd2dvi: Rd.sty not found
Duncan Murdoch murdoch at stats.uwo.ca writes: : : On Thu, 10 Feb 2005 21:27:03 + (UTC), Gabor Grothendieck : ggrothendieck at myway.com wrote : : : Tuszynski, Jaroslaw W. JAROSLAW.W.TUSZYNSKI at saic.com writes: : : : : : Hi, : : : : I run into a problem with R CMD Rd2dvi command: it gives me File `Rd.sty' : : not found error (See the output message on the bottom). I get the same : : error when running RCMD check. My system is: Windows 2000, R version : : 2.0.1, MiKTeX version 2.4. I do have a Rd.sty file in R_HOME/share/texm : : directory. : : : ... : : There is some information on this at: : : http://www.murdoch-sutherland.com/Rtools/miktex.html : : The information there is slightly out of date: the last version of : Miktex that I've tried has broken that solution. I asked the author : about it, and he replied as below. : : I think this is pretty poor behaviour on the part of Miktex, but in : other respects I prefer it to fptex. : : For what you're doing, you need to set up your : $R_HOME/bin/helpprint.bat file so that it makes use of the TEXINPUTS : environment variable. You want the pdftex command to have something : like : : -include-directory=%TEXINPUTS% : : in it and then it will look in the right places for the includes. : : Duncan Murdoch I tried playing around with MiKTeX and got some advice on the TeX list and now am using this config. This example assumes you have put your localtexmf file in \localtexmf and are running R 2.0.1 from \Program Files\R\rw2001; otherwise, make the obvious changes. 1. Create a new tex subfolder of your \localtexmf\texmf folder: md \localtexmf\texmf\tex 2. Copy your R .sty and .fd files into it: cd \Program Files\R\rw2001\share\texmf copy *.* \localtexmf\texmf\tex 3. Go into Start | Programs | MiKTeX | MiKTeX Options | General and press Update Now and Refresh Now. (I am not certain that this step needs to be done but it can't hurt.) #2 and #3 need to be repeated each time one installs a new version of R if the *.sty or *.fd files have changed but if they have not changed then nothing at all needs to be done. It has the advantage that it leaves all MiKTeX options at their defaults. It represents a fourth alternative to the three listed at: http://www.murdoch-sutherland.com/Rtools/miktex.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] Problem with R CMD Rd2dvi: Rd.sty not found
Gabor Grothendieck ggrothendieck at myway.com writes: I tried playing around with MiKTeX and got some advice on the TeX list and now am using this config. This example assumes you have put your localtexmf file in \localtexmf and are running R 2.0.1 from \Program Files\R\rw2001; otherwise, make the obvious changes. Sorry, there were some errors in my description of configuring MiKTeX so that it finds the Rd.sty file and associated files. It should be: 1. Create a new tex subfolder of your \localtexmf folder: md \localtexmf\tex 2. Copy your R .sty and .fd files into it: cd \Program Files\R\rw2001\share\texmf copy *.* \localtexmf\tex 3. Go into Start | Programs | MiKTeX | MiKTeX Options | General and press Update Now and Refresh Now. (I am not certain that this step needs to be done but it can't hurt.) 4. You can check whether it is finding Rd.sty in the right place with the command: findtexmf Rd.sty #2 - #4 need to be repeated each time one installs a new version of R if the *.sty or *.fd files have changed but if they have not changed then nothing at all needs to be done. It has the advantage that it leaves all MiKTeX options at their defaults and does not need a custom miktex.ini file. It represents a fourth alternative to the three listed at: http://www.murdoch-sutherland.com/Rtools/miktex.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] Problem with R CMD Rd2dvi: Rd.sty not found
On Tue, 15 Feb 2005 19:44:51 + (UTC), Gabor Grothendieck [EMAIL PROTECTED] wrote : I tried playing around with MiKTeX and got some advice on the TeX list and now am using this config. This example assumes you have put your localtexmf file in \localtexmf and are running R 2.0.1 from \Program Files\R\rw2001; otherwise, make the obvious changes. Thanks, I'll put this on the web page (with the corrections from your next message). It's probably the best solution for most users, but it doesn't work for me: I need to have more than one R version on my machine at the same time (I currently have R 2.0.1, R-patched, and R-devel; usually I have a few old ones too, but I'm working on a laptop with limited space.) This solution assumes you have only one version installed. I don't know what people developing LaTeX styles do: presumably they'd have multiple versions of the style online, testable in multiple versions of LaTeX. 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] Trellis barchart, column display order issue
I have searched the archives and instructions high and low but have not found what I need. I have a dataframe named Data with columns: AdjResND0 - the numeric result Parameter - the classification (chemical name) Sort - the order I want the chemical names to appear, with leading zeros so all are the same length Client.Name - the sample name I am trying to generate a series of 1 page barcharts, one for each sample. The dataframe has been sorted by Client.name and sort I use the following command: barchart(AdjResND0 ~ Parameter | Client.Name , Data, box.ratio = 0.8, ylab= (Concentration (mg/kg)), layout = c(0,1), scales = list(x=free,rot=90,cex = 0.5, axs = i,abbreviate = TRUE ) ) The parameters are listed alphabetic order from left to right in the bar chart so I constructed a new variable Label which is the sort and parameter variables concatenated together. barchart(AdjResND0 ~ Label | Client.Name , Data, box.ratio = 0.8, ylab= (Concentration (mg/kg)), layout = c(0,1), scales = list(x=free,rot=90,cex = 0.5, axs = i,abbreviate = TRUE ) ) This has everything in the correct order but obviously I don't want the column labels to have the sort order glued on the front. Any way to get the order I want and the labels I want? I am using : R : Copyright 2004, The R Foundation for Statistical Computing Version 2.0.1 (2004-11-15), ISBN 3-900051-07-0 Under Windows XP Also, I am as yet unable to figure out how to get R to spool the graphs to a pdf file so I get all 30 graphs and I can save them for an appendix. I just want to be able to save all the graphs for future use, so I am not set on pdf or anything else. An example would be great! Thanks, Mike Michael J. Bock, PhD. ARCADIS 24 Preble St. Suite 100 Portland, ME 04101 207.828.0046 fax 207.828.0062 [[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] Making a Package
Hello. I have what I know to be a simple question, but never having done anything like this it is pretty tough. I'm trying to write an R package. I have a collection of functions that I loaded into R and then used package.skeleton(). After editing everything in the resulting folder, call it NewPackage, I tried to follow along with some instructions I found for Windows users. I installed ActivePearl to the C drive, placed the unzipped Rtools folder, tools, there also (is placing it there all that is necessary; I couldn't find anything to install tools), and I'm using R 2.00 (also installed on the C drive) on a Win-XP machine. My understaning is that (since I have no C, Fortran, etc., code) I can move the NewPackage folder (with all of the edited material) to the C:\r\rw2000\src\library folder and then open MS-Dos in C:\r\rw2000\src\gnuwin32 and type: make pkg-NewPackage. After that, a folder called NewPackage should be placed in the C:\r\rw2000\library and from there loaded into R and/or zipped and distributed. If everything I have said is correct (which it may not be) then I'm stuck. When I type 'make pkg-NewPackage' in the C:\r\rw2000\src\gnuwin32 directory I get 'make' is not recognized as an internal or external command, operable program or batch file. I've tried to use Writing R Extensions, but I could only follow part of what it was saying and got confused as to what was done on Windows and what was done on Unix machines. I know this is probably an easy question, but it has proved difficult for me to figure out how to make my own package. Thanks, Ken - [[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] Making a Package
KKThird at Yahoo.Com kkthird at yahoo.com writes: : : Hello. : I have what I know to be a simple question, but never having done anything like this it is : pretty tough. : : I'm trying to write an R package. I have a collection of functions that I loaded into R and : then used package.skeleton(). After editing everything in the resulting folder, call it : NewPackage, I tried to follow along with some instructions I found for Windows users. : : I installed ActivePearl to the C drive, placed the unzipped Rtools folder, tools, there also : (is placing it there all that is necessary; I couldn't find anything to install tools), and : I'm using R 2.00 (also installed on the C drive) on a Win-XP machine. : : My understaning is that (since I have no C, Fortran, etc., code) I can move the NewPackage : folder (with all of the edited material) to the C:\r\rw2000\src\library folder and then open : MS-Dos in C:\r\rw2000\src\gnuwin32 and type: make pkg-NewPackage. After that, a folder called : NewPackage should be placed in the C:\r\rw2000\library and from there loaded into R and/or zipped and : distributed. If everything I have said is correct (which it may not be) then I'm : stuck. When I type 'make pkg-NewPackage' in the C:\r\rw2000\src\gnuwin32 directory I get : 'make' is not recognized as an internal or external command, operable program or batch file. : : I've tried to use Writing R Extensions, but I could only follow part of what it was saying : and got confused as to what was done on Windows and what was done on Unix machines. : I know this is probably an easy question, but it has proved difficult for me to figure out how : to make my own package. : Thanks, Ken Make sure that: - the tools and the R bin folders are in your path - you have installed fptex or MiKTeX, perl and the Microsoft help compiler (MiKTeX is a bit harder to install but I otherwise prefer it. If you are writing vignettes you must use MiKTeX.) - you have read: Writing R Extensions Manual http://www.murdoch-sutherland.com/Rtools/ I am assuming here that your R installation is in \Program Files\R\rw 1. Assuming your source package tree is in \Rpkgs\mypackage then in the Windows cmd line: cd \Rpkgs Rcmd install mypackage which will install it to your \Program Files\R\rw\library\mypackage or if you want to install it to a separate library: cd \Rpkgs md library Rcmd install -l library mypackage 2. Now in R: library(mypackage) ... test it out ... or if you installed it to a separate library: library(mypackage, lib.loc = /Rpkgs/library) 3. Once it seems reasonably ok, see if it passes Rcmd check: cd \Rpkgs Rcmd check mypackage and fix it up until it does. 4. Now create versions for UNIX and Windows that you can distribute: cd \Rpkgs Rcmd build mypackage Rcmd build mypackage --binary __ 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] summary(aov(...)) into a string?
It doesn't print anything: the summary.aov (or summary.aovlist) print method does. ?summary.aov tells you the structure of the objects they return. On Tue, 15 Feb 2005, RenE J.V. Bertin wrote: I'd like to annotate a plot with the output of summary(aov(model)), ideally just with the significant effects. I don't find a means to redirect what that command prints into a string. Is this possible, and if so, how? -- 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] Trellis barchart, column display order issue
On Tuesday 15 February 2005 14:57, Bock, Michael wrote: I have searched the archives and instructions high and low but have not found what I need. I have a dataframe named Data with columns: AdjResND0 - the numeric result Parameter - the classification (chemical name) Sort - the order I want the chemical names to appear, with leading zeros so all are the same length Client.Name - the sample name Try looking at levels(Data$Parameter) I'll bet they are in alphabetical order. See ?factor for how to specify the levels explicitly. Alternatively, barring unforeseen complications, and assuming that you have a sufficiently modern version of R, you should be able to do barchart(AdjResND0 ~ reorder(Parameter, sort) | Client.Name, Data, ... and if that doesn't work as intended, barchart(AdjResND0 ~ reorder(Parameter, as.numeric(as.character(sort))) | Client.Name, Data, ... Deepayan I am trying to generate a series of 1 page barcharts, one for each sample. The dataframe has been sorted by Client.name and sort I use the following command: barchart(AdjResND0 ~ Parameter | Client.Name , Data, box.ratio = 0.8, ylab= (Concentration (mg/kg)), layout = c(0,1), scales = list(x=free,rot=90,cex = 0.5, axs = i,abbreviate = TRUE ) ) The parameters are listed alphabetic order from left to right in the bar chart so I constructed a new variable Label which is the sort and parameter variables concatenated together. barchart(AdjResND0 ~ Label | Client.Name , Data, box.ratio = 0.8, ylab= (Concentration (mg/kg)), layout = c(0,1), scales = list(x=free,rot=90,cex = 0.5, axs = i,abbreviate = TRUE ) ) This has everything in the correct order but obviously I don't want the column labels to have the sort order glued on the front. Any way to get the order I want and the labels I want? I am using : R : Copyright 2004, The R Foundation for Statistical Computing Version 2.0.1 (2004-11-15), ISBN 3-900051-07-0 Under Windows XP Also, I am as yet unable to figure out how to get R to spool the graphs to a pdf file so I get all 30 graphs and I can save them for an appendix. I just want to be able to save all the graphs for future use, so I am not set on pdf or anything else. An example would be great! Thanks, Mike Michael J. Bock, PhD. ARCADIS 24 Preble St. Suite 100 Portland, ME 04101 207.828.0046 fax 207.828.0062 __ 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] summary(aov(...)) into a string?
RenE J.V. Bertin rjvbertin at hotmail.com writes: : : Hello, : : I'd like to annotate a plot with the output of summary(aov(model)), ideally just with the significant : effects. I don't find a means to redirect what that command prints into a string. : Is this possible, and if so, how? : ?capture.output __ 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] Tcl/tk
I am new to R so I apologize if this is a simple question I have created a tkgrid with a button that, once pushed, will extract the information from the grid, but I can not figure out the command to save the entries from my grid. Any thoughts? Thanks Alex Bettinardi Transportation Analyst TDD - Planning Section - TPAU 555 13th Street NE, Suite 2 Salem, OR 97301-4178 503.986.4398 [[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] Making a Package
I follow the guide here and it never fails: http://www.biostat.jhsph.edu/~kbroman/Rintro/Rwinpack.html So if I have the MyPackage directory, that was created with package.skeleton and subsequently edited, I will cd to the directory that contains the MyPackage directory and: R CMD INSTALL --build MyPackage MyPackage_1.0.zip will be created. Then install using the menus. -Original Message- From: [EMAIL PROTECTED] on behalf of Gabor Grothendieck Sent: Tue 2/15/2005 9:59 PM To: r-help@stat.math.ethz.ch Cc: Subject: Re: [R] Making a Package KKThird at Yahoo.Com kkthird at yahoo.com writes: : : Hello. : I have what I know to be a simple question, but never having done anything like this it is : pretty tough. : : I'm trying to write an R package. I have a collection of functions that I loaded into R and : then used package.skeleton(). After editing everything in the resulting folder, call it : NewPackage, I tried to follow along with some instructions I found for Windows users. : : I installed ActivePearl to the C drive, placed the unzipped Rtools folder, tools, there also : (is placing it there all that is necessary; I couldn't find anything to install tools), and : I'm using R 2.00 (also installed on the C drive) on a Win-XP machine. : : My understaning is that (since I have no C, Fortran, etc., code) I can move the NewPackage : folder (with all of the edited material) to the C:\r\rw2000\src\library folder and then open : MS-Dos in C:\r\rw2000\src\gnuwin32 and type: make pkg-NewPackage. After that, a folder called : NewPackage should be placed in the C:\r\rw2000\library and from there loaded into R and/or zipped and : distributed. If everything I have said is correct (which it may not be) then I'm : stuck. When I type 'make pkg-NewPackage' in the C:\r\rw2000\src\gnuwin32 directory I get : 'make' is not recognized as an internal or external command, operable program or batch file. : : I've tried to use Writing R Extensions, but I could only follow part of what it was saying : and got confused as to what was done on Windows and what was done on Unix machines. : I know this is probably an easy question, but it has proved difficult for me to figure out how : to make my own package. : Thanks, Ken Make sure that: - the tools and the R bin folders are in your path - you have installed fptex or MiKTeX, perl and the Microsoft help compiler (MiKTeX is a bit harder to install but I otherwise prefer it. If you are writing vignettes you must use MiKTeX.) - you have read: Writing R Extensions Manual http://www.murdoch-sutherland.com/Rtools/ I am assuming here that your R installation is in \Program Files\R\rw 1. Assuming your source package tree is in \Rpkgs\mypackage then in the Windows cmd line: cd \Rpkgs Rcmd install mypackage which will install it to your \Program Files\R\rw\library\mypackage or if you want to install it to a separate library: cd \Rpkgs md library Rcmd install -l library mypackage 2. Now in R: library(mypackage) ... test it out ... or if you installed it to a separate library: library(mypackage, lib.loc = /Rpkgs/library) 3. Once it seems reasonably ok, see if it passes Rcmd check: cd \Rpkgs Rcmd check mypackage and fix it up until it does. 4. Now create versions for UNIX and Windows that you can distribute: cd \Rpkgs Rcmd build mypackage Rcmd build mypackage --binary __ 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] summary(aov(...)) into a string?
Prof Brian Ripley [EMAIL PROTECTED] writes: It doesn't print anything: the summary.aov (or summary.aovlist) print method does. ?summary.aov tells you the structure of the objects they return. Yes. but wouldn't capture.output() and par(family=mono) be closer to the mark? On Tue, 15 Feb 2005, RenE J.V. Bertin wrote: I'd like to annotate a plot with the output of summary(aov(model)), ideally just with the significant effects. I don't find a means to redirect what that command prints into a string. Is this possible, and if so, how? -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- 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] Tcl/tk
[EMAIL PROTECTED] writes: I am new to R so I apologize if this is a simple question I have created a tkgrid with a button that, once pushed, will extract the information from the grid, but I can not figure out the command to save the entries from my grid. Any thoughts? Several, all starting with What... What is tkgrid (not the geometry manager surely)? What information? What form is it extracted in? What entries? What do you want to save it to? -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- 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] Problem with
I am just in the middle of a conversation on the tex list but just in case you are about to post anything to your web site here is a slight simplification. It consists of updating the MiKTeX file database using initexmf instead of the GUI so you don't have to move back and forth during the instructions. The discussion there might correct or improve this further. Regards. --- 1. Create a new tex subfolder of your \localtexmf folder: md \localtexmf\tex 2. Copy your R .sty and .fd files into it, update the MiKTeX texmf file name database and check that Rd.sty is found in the correct place, viz. \localtexmf\tex: cd \Program Files\R\rw2001\share\texmf copy *.* \localtexmf\tex initexmf -u findtexmf Rd.sty #2 should be repeated each time one installs a new version of R or any time the *.sty or *.fd files are changed. (Actually nothing needs to be done if the .sty and .fd files have not changed from one version of R to another and the initexmf command does not need to be run unless the filenames themselves have changed even if their contents have; however, none of these steps will hurt and its safest to do them all.) __ 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] Could anyone answer for the following question
options(graphics.record=TRUE) turns on graphics recording. This can also be done from the History menu of the graphics window. All subsequent plots to the graphics device will be saved. -- Bert Gunter Genentech Non-Clinical Statistics South San Francisco, CA The business of the statistician is to catalyze the scientific learning process. - George E. P. Box -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Cuichang Zhao Sent: Tuesday, February 15, 2005 2:36 PM To: r-help@stat.math.ethz.ch Subject: [R] Could anyone answer for the following question Hello , could anyone answer for the following question for me: I am using R 2.0.1 under Windows XP. I want to write a function that makes four graphs and stores each of them in graphics history. When the function finishes, in other words, I want its graphical output to be stored in a way that I can look at it using PgUp and PgDn. I think I need commands I can put in a function that --clear graphics history --either turn on graphics history recording or store the current graph Can anyone explain how to do this? Thank you so much C-Ming Feb 15, 2005 - [[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] Could anyone answer for the following question
Cuichang Zhao cuiczhao at yahoo.com writes: : I am using R 2.0.1 under Windows XP. I want to write a function that : makes four graphs and stores each of them in graphics history. When the : function finishes, in other words, I want its graphical output to be : stored in a way that I can look at it using PgUp and PgDn. I think I need : commands I can put in a function that : : --clear graphics history : --either turn on graphics history recording or store the current graph windows(record = TRUE) will turn on windows graphics history and savePlot will save a windows plot in a file. You can also turn it on manually by clicking on the graphics window and making sure History | Record is checked. (Its also possible to save the low level graphics that R displays using the displaylist and you can search r-help for more info on 'displaylist' but I don't think that is what you are looking for.) __ 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] reading a 40kb csv file in R
I think you may need to check the list for posts relating to memory, large datasets, etc http://cran.r-project.org/search.html It would help those on the list, if you gave more details about the machine you are using. Given the vintage of R you areusing is considered quite old, there is the possibility that you are using a machine with 256 Mb of RAM in which case it is likely you have hit a barrier that requires you to think about alternative ways of processing your data. The reason I suggest you serach the posts yourself rather than give advice is that there are many possibilities which depend on information you have that I don't, about other software you have such as sql databases or you may understand better than I do some of the posts about sparse matrices. Tom -Original Message- From: Fairouz Makhlouf [mailto:[EMAIL PROTECTED] Sent: Wednesday, 16 February 2005 1:05 AM To: r-help@stat.math.ethz.ch Subject: [R] reading a 40kb csv file in R I am trying to read a matrix of 5000*2000 from a csv file of real numbers in R using the scan function. I am getting an error saying that R can not read a vector 32KB and it can't allocate a memory of 191 MB etc. What should I do. I am using R1.7.1. Thanks = Thanks Fairouz Makhlouf __ 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] lattice multiple plots per page
Hi Deepayan Sarkar wrote: On Monday 14 February 2005 18:12, Jeff Jorgensen wrote: Dear R-sters, I was wondering if anyone has encountered the following issues. I've figured out how to get multiple levelplots [library(lattice)] on a single plot. However, when I add text (adding axis labels for the entire four panel plot) the text is missing when I insert the *.eps file I've created into my LaTeX document (via MikTeX-WinEdt). And, I've just upgraded to R v2.0.1 from v1.8.1 (Windows), and each individual levelplot is smaller compared to the older R release. Any clues as to 1) how I can recover the lost text, and Your use is incorrect, in the sense that the return value of grid.locator() is not a valid input of ltext. You could do the following to fix this, library(grid) ltext(lapply(grid.locator(), convertX, native, TRUE), lab=x-axis label, where I click,cex=1.5) ltext(lapply(grid.locator(), convertX, native, TRUE), lab = y-axis label, where I click, cex=1.5, srt=90) but that doesn't seem to help either. So, there seems to be a bug somewhere. I think this has the same problem as the original code, which is that it attempts to use ltext() outside of a lattice panel function, so the native units assumed by ltext() could be inappropriate. Within the top-level grid viewport, native units are (at least approximately) the native units of the device (i.e., pixels). The original drawing on screen is occurring in terms of screen pixels, but when you do the dev.copy2eps(), native means something different and you end up drawing in terms of PostScript points (1/72 inches), so the locations may be different (especially if your screen is, say, 96 dpi). However, you presumably want something like grid.text(x = .05, y = .5, lab=y-axis label, default.units = npc, gp = gpar(cex=1.5), rot = 90) grid.text(x = .5, y = .05, lab=x-axis label, default.units = npc, gp = gpar(cex=1.5)) for the last two steps, which seems to work. Right. First of all, use grid.text() not ltext(), and secondly, use npc coordinates (in fact, anything other than native should do), which should transform to the same location on the dev.copy2eps(). If you want to position the labels using a mouse click ... xy - grid.locator(npc) grid.text(xy$x, xy$y, lab=y-axis label, gp = gpar(cex=1.5), rot = 90) xy - grid.locator(npc) grid.text(xy$x, xy$y, lab=x-axis label, gp = gpar(cex=1.5)) ... again using npc locations rather than native. Paul 2) increase the size of each of the levelplots? You need to change the following settings, in particular the entries that end with 'padding' (similarly for 'layout.heights'). str(trellis.par.get(layout.widths)) List of 13 $ left.padding : num 1 $ key.left : num 1 $ key.ylab.padding : num 1 $ ylab : num 1 $ ylab.axis.padding: num 1 $ axis.left: num 1 $ axis.panel : num 1 $ panel: num 1 $ between : num 1 $ axis.right : num 1 $ axis.key.padding : num 1 $ key.right: num 1 $ right.padding: num 1 Deepayan Cheers, Jeff Jorgensen ~~~ Sample code that illustrates what I'm trying to do: #create a levelplot x-seq(pi/4, 5*pi, length = 100) y-seq(pi/4, 5*pi, length = 100) r-as.vector(sqrt(outer(x^2, y^2, +))) grid-expand.grid(x=x, y=y) grid$z-cos(r^2) * exp(-r/(pi^3)) a-levelplot(z~x*y, grid, cuts = 50, xlab=, ylab=, colorkey = FALSE) #create the multiple panel plot, here using all the same levelplot trellis.par.set(list(background=list(col=white))) #white background #using position to scale the plots up and to the right ~10% #to make room for the axis labels print(a,position=c(0.1,0.1,1,1),split=c(1,1,2,2),more=T) print(a,position=c(0.1,0.1,1,1),split=c(1,2,2,2),more=T) print(a,position=c(0.1,0.1,1,1),split=c(2,1,2,2),more=T) print(a,position=c(0.1,0.1,1,1),split=c(2,2,2,2),more=F) #commands that let you click where you want the labels centered ltext(grid::grid.locator(),lab=x-axis label, where I click,cex=1.5) ltext(grid::grid.locator(),lab=y-axis label, where I click,cex=1.5,srt=90) #save device to an *.eps file, to be called later by a \includegraphics command dev.copy2eps(file=twobytwoplot.eps) __ 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 -- Dr Paul Murrell Department of Statistics The University of Auckland Private Bag 92019 Auckland New Zealand 64 9 3737599 x85392 [EMAIL PROTECTED] http://www.stat.auckland.ac.nz/~paul/ __ 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