Re: [R] Error setting rowname if rowname currently NULL
snubian wrote: snip I've noticed a behaviour when using rownames() that I think is odd, wondering if I'm doing something wrong. To illustrate, say I create a very simple matrix (called fred): fred-matrix(,4,2) It looks like this: [,1] [,2] [1,] NA NA [2,] NA NA [3,] NA NA [4,] NA NA ... and rownames(fred) # NULL If I now try and set a row name for one of the rows (say the first row) to APPLE, by doing this: rownames(fred)[1] - APPLE ... so now rownames(fred) would have to become a vector of length one, and you're trying to use a vector of length one to name rows in a matrix with four rows. (in principle, you could argue that the vector should be recycled on such an occassion, as it happens in many other situations.) hence I get an error: Error in dimnames(x) - dn : length of 'dimnames' [1] not equal to array extent ... because 1 != 4 However, I found that if I first set all the rownames to anything at all, by using say: rownames(fred) - c(1:4) ... so that length(rownames(fred)) == 4 Which gives me: [,1] [,2] 1 NA NA 2 NA NA 3 NA NA 4 NA NA Then my desired command works, and thus: rownames(fred)[1] - APPLE ... whereby you change one element in rownames(fred) and it still has 4 elements Gives me what I want: [,1] [,2] APPLE NA NA 2 NA NA 3 NA NA 4 NA NA So, what this says to me is that to set the row names INDIVIDUALLY, they first need to be set to something (anything!). not exactly true: fred = matrix(, 4, 2) rownames(fred)[4] = 'foo' here you'd set the 4th element of rownames(fred), which fill the first three with NA, and rownames(fred) has length 4, as needed. For what I am doing, I need to set the row names one at a time, as I iterate through a loop. is it possible that you don't really need it? vQ __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] [R-pkgs] package ltm -- version 0.9-0
Dear R-users, I'd like to announce the release of the new version of package 'ltm' (i.e., ltm_0.9-0 soon available from CRAN) for Item Response Theory analyses. This package provides a flexible framework for analyzing dichotomous and polytomous data under various IRT models. Furthermore, supporting functions for descriptive statistics, goodness-of-fit, ability estimation and plotting are available. New features include: * function gpcm() (along with supporting methods, for the generic functions, anova, plot, margins, factor.scores, summary, vcov) has been added for fitting the Generalized Partial Credit Model. The 'constraint' argument of gpcm() can be used to specify constrained versions of the model. Options are: - 'constraint = rasch' that fits the partial credit model (i.e., discrimination parameter fixed at one) - 'constraint = 1PL' that specifies a single discrimination parameter, equal among items, and - 'constraint = gpcm' that specifies a different discrimination parameter per item. * gpcm() also allows for mixed type items, i.e., dichotomous and polytomous; however, for data sets with only dichotomous items, functions rasch() and ltm() are more efficient. * The new function GoF.gpcm() calculates the Pearson chi-squared statistic for Generalized Partial Credit models. A p-value can be also produced either using the asymptotic chi-squared distribution or a parametric Bootstrap approach. * The new function rmvordlogis() can be used to simulate ordinal data under the Graded Response and the Generalized Partial Credit models. * for more details and other new features check the NEWS file that ships with the package. More information as well as *.R files illustrating the capabilities of the package can be found in the R-wiki page of 'ltm' available at: http://wiki.r-project.org/rwiki/doku.php?id=packages:cran:ltm. As always, any kind of feedback (e.g., questions, suggestions, bug-reports, etc.) is more than welcome. Best, Dimitris -- Dimitris Rizopoulos Assistant Professor Department of Biostatistics Erasmus University Medical Center Address: PO Box 2040, 3000 CA Rotterdam, the Netherlands Tel: +31/(0)10/7043478 Fax: +31/(0)10/7043014 ___ R-packages mailing list r-packa...@r-project.org https://stat.ethz.ch/mailman/listinfo/r-packages __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Goldbach partitions code
Hi interesting blog! not strictly relevant, but there are various number-theoretic functions implemented in the elliptic package which you might find useful. best wishes Robin murali.me...@fortisinvestments.com wrote: Folks, I put up a brief note describing my naive attempts to compute Goldbach partitions, starting with a brute-force approach and refining progressively. http://jostamon.blogspot.com/2009/02/goldbachs-comet.html I'd welcome your suggestions on improvements, alternatives, other optimisations, esp. to do with space vs time tradeoffs. Is this an example interesting enough for pedagogical purposes, do you think? Please advise. Cheers, MM [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Robin K. S. Hankin Uncertainty Analyst University of Cambridge 19 Silver Street Cambridge CB3 9EP 01223-764877 __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Help with Wilcoxon Test
Amit Patel wrote: Hi I have 2 sets of data that I want to do a Wilcoxon test on. They are of the same dimension. One has 4 zero values and the other has 5. dim(SampA) [1] 1 10 dim(SampV) [1] 1 10 I get the folowing error Error in wilcox.test.default(SampA, SampV, na.rm = TRUE, paired = FALSE, : 'x' must be numeric I am using the function wilcox.test(SampA, SampV, na.rm=TRUE, paired=FALSE, conf.level=0.95) Hi Amit, This is probably due to the fact that SampA is not numeric (SampB may be the same). When you say zero values, I suspect you may mean NAs, and SampA and SampB are read in from a data file of some sort. It's quite common for users to assume that an empty field (or even a space or a period .) will be read as NA. If this sounds like it may be the problem, check the function that you are using to read in the data for an na.strings argument, and see if it is what you think it is. If it isn't, any column of a resulting data frame will usually come in as a factor. Jim __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Characters To Matrix
Hi, I have a data set that I read in using readLines, but R classes it as a character string. The data set has multiple columns and rows, but I only want the first 2 columns in the form of a matrix. Thanks for the help. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Error loading packages at sprintf()
This is a Mac-specific issue, as you are installing binary packages. The real error message was /bin/sh: line 1: tar: command not found so the problem is that tar is not in your path. I've no idea how that would come to be (all Macs I have used had /usr/bin/tar), so please ask on R-sig-mac. As far as I can tell, the sprintf message also comes from Mac-specific code and I thnk I know how to correct that. But the real error was the one about 'tar'. On Mon, 2 Mar 2009, Andrew J. Rominger wrote: Hello all, I'm running R2.8.1 on a Mac OS 10.4.11. While trying to install the package gdata, I was presented with the following (error at end of report): R install.packages(gdata) also installing the dependency ‘gtools’ trying URL 'http://cran.stat.ucla.edu/bin/macosx/universal/contrib/2.8/gtools_2.5.0-1.tgz' Content type 'application/x-tar' length 85484 bytes (83 Kb) opened URL == downloaded 83 Kb trying URL 'http://cran.stat.ucla.edu/bin/macosx/universal/contrib/2.8/gdata_2.4.2.tgz' Content type 'application/x-tar' length 539301 bytes (526 Kb) opened URL == downloaded 526 Kb /bin/sh: line 1: tar: command not found 2009-03-02 20:42:06.081 R[357] tossing reply message sequence 3 on thread 0x1ce3ae0 Error in sprintf(gettext(fmt, domain = domain), ...) : argument is missing, with no default I can't figure out why this error is occurring [the error in sprintf(gettext(fmt, domain = domain), ...]. I've never been prompted to supply arguments to sprintf before. Upon trying to install various other packages (e.g. 'vegan', 'ade4', 'ads'...) I get the same error message regarding sprintf. I recently removed many old data objects from my workspace, could I have accidentally messed with sprintf? In a previous R-help post a similar error resulted while trying to install a package from source, and was corrected by specifying type=source, but I'm not really sure how to properly specify the argument 'lib' and so leave it as it's default value. Without specifying 'lib' is it appropriate to call type=source? Thanks in advance for any help-- Andy Rominger __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Brian D. Ripley, rip...@stats.ox.ac.uk 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@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] regular expression question
markle...@verizon.net wrote: can someone show me how to use a regular expression to break the string at the bottom up into its three components : (-0.791,-0.263] (-38,-1.24] (0.96,2.43] I tried to use strplit because of my regexpitis ( it's not curable. i've been to many doctors all over NYC. they tell me there's no cure ) but it doesn't work because there also dots inside the brackets. Thanks. (-0.791,-0.263].(-38,-1.24].(0.96,2.43] here's one way to get a matrix of numeric values: text = (-0.791,-0.263].(-38,-1.24].(0.96,2.43] values = matrix(ncol=2, byrow=TRUE, as.numeric( grep(pattern='.', value=TRUE, x=strsplit(x=text, split=']\\.\\(|\\(|]|,')[[1]]))) modify any of the steps according to your needs. vQ __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Characters To Matrix
Bob Roberts quagmire54321 at yahoo.com writes: I have a data set that I read in using readLines, but R classes it as a character string. The data set has multiple columns and rows, but I only want the first 2 columns in the form of a matrix. Thanks for the help. data(CO2) str(CO2) col2 = as.matrix(CO2[,1:2]) str(col2) Dieter __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] RWeka dataset running
ahmet kocyigit kh3l at hotmail.com writes: I'm a newbie to R and will work on it from now. I'm preparing my thesis based on rpart and RWeka. I've managed to run already existing test data but couldn't import a data set from out of R. There is a special Export/Import manual R-data.pdf in Rxxx/doc/manual. Dieter __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] regular expression question
Wacek Kusnierczyk wrote: markle...@verizon.net wrote: can someone show me how to use a regular expression to break the string at the bottom up into its three components : (-0.791,-0.263] (-38,-1.24] (0.96,2.43] I tried to use strplit because of my regexpitis ( it's not curable. i've been to many doctors all over NYC. they tell me there's no cure ) but it doesn't work because there also dots inside the brackets. Thanks. (-0.791,-0.263].(-38,-1.24].(0.96,2.43] here's one way to get a matrix of numeric values: text = (-0.791,-0.263].(-38,-1.24].(0.96,2.43] values = matrix(ncol=2, byrow=TRUE, as.numeric( grep(pattern='.', value=TRUE, x=strsplit(x=text, split=']\\.\\(|\\(|]|,')[[1]]))) modify any of the steps according to your needs. vQ Here is another way with the gsubfn package: require( gsubfn ) strapply( text, \\(.*?,.*?], perl = T )[[1]] 1] (-0.791,-0.263] (-38,-1.24] (0.96,2.43] Note that gregexpr would also help you here: g - gregexpr( \\(.*?,.*?], text, perl = T )[[1]] g [1] 1 17 29 attr(,match.length) [1] 15 11 11 But there is always the missing part of extracting the match from the result of (g)regexpr substring( text, g, g + attr(g, match.length ) - 1 ) [1] (-0.791,-0.263] (-38,-1.24] (0.96,2.43] Romain -- Romain Francois Independent R Consultant +33(0) 6 28 91 30 30 http://romainfrancois.blog.free.fr __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] New mailing list: R-help-es
Hi all, I am writing to announce the creation of a new R mailing list: R-help-es. It is going to be a list for spanish speakers users of R so I am sorry to write the rest of the email in spanish. Os escribo para anunciaros que hemos creado una nueva lista de correo llamada R-help-es para usuarios de R hispanohablantes en respuesta a la creciente comunidad de usuarios en España y Latinoamérica. La lista está orientada aquellas noticias, acontecimientos, ofertas laborales, etc. de interés para la comunidad hispanoparlante únicamente y para aquellos que necesiten ayuda y tengan dificultades con el ingles. No se trata de solapar contenidos con ninguna de las otras listas que ya existen. Somos conscientes de que existe un creciente interés por R en todo el mundo y creemos que la descentralización de alguna información como ofertas de empleo locales, y cursos en castellano puede ser muy interesante para alguno de nosotros y para el mejor funcionamiento de la comunidad de R. Se va a abrir un hilo en la nueva lista para plantear posibilidades mas ambiciosas la como creación de un portal para la comunidad de usuarios de R en español, con el objetivo de conocernos, intercambiar recursos y en definitiva dar apoyo y ser un referente de usuarios en España y Latinoamérica. Para subscribiros a la nueva lista, tan solo tenéis que ir a este enlace: https://stat.ethz.ch/mailman/listinfo/r-help-es Os esperamos ahí, un saludo a todos. _ Descubre todas las formas en que puedes estar en contacto con amigos y familiares. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Inefficiency of SAS Programming
Ajay, I think this is somewhat unnecessary! For regular SAS users PROC SUMMARY/MEANS is the data manipulation staple. A SAS user interested in R will look at Intro to R and find it recommends tapply for data manipulation jobs. This is a problem - tapply just isn't simple and obvious. A better option would be for the Intro to R to link to Bob Muenchen's original 80-page document R for SAS and SPSS users, page 52 where the summarize function is described. Gerard Ajay ohri ohri2...@gmail.c omTo Greg Snow greg.s...@imail.org 03/03/2009 04:58 cc Gerard M. Keogh gmke...@justice.ie, r-help-boun...@r-project.org r-help-boun...@r-project.org, R list r-h...@stat.math.ethz.ch Subject Re: [R] Inefficiency of SAS Programming for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. It basically is meant only for SAS users who want to learn R , without upsetting the schedule of the corporate users. Simply put , it is a wrapper on SAS language using the function command...ie procunivariate function in Anne package would call the summary function and so on... Regards, Ajay www.decisionstats.com On Tue, Mar 3, 2009 at 9:20 AM, Greg Snow greg.s...@imail.org wrote: This does not really address my point. Yes, if the few nerds who want to do funny stuff are the ones making the purchase, then there is a good chance (but still not guaranteed) that they will get IML, but do all companies that buy SAS actually think about that, or do they just see the extra price (no matter how low), or not even think to look at that piece because the person making the purchase does not really the funny things you can do with it. If you want your SAS code to be able to be run by anyone with SAS, you cannot assume that they have IML. If you want your R code to be run by anyone, you cannot make your code dependent on packages/tools that are not available for all platforms. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org 801.408.8111 -Original Message- From: Gerard M. Keogh [mailto:gmke...@justice.ie] Sent: Monday, March 02, 2009 3:22 AM To: Greg Snow Cc: Frank E Harrell Jr; R list; r-help-boun...@r-project.org Subject: Re: [R] Inefficiency of SAS Programming Yes Greg, but if you're buying SAS they'll throw in IML pretty cheaply - SAS think it's only for a few nerds out there who wan to do funny stuff. G Greg Snow greg.s...@imail. org To Sent by: Gerard M. Keogh r-help-boun...@r- gmke...@justice.ie, Frank E project.org Harrell Jr f.harr...@vanderbilt.edu cc 27/02/2009 19:05 r-help-boun...@r-project.org r-help-boun...@r-project.org, R list r-h...@stat.math.ethz.ch Subject Re: [R] Inefficiency of SAS Programming But SAS/IML is not part of base SAS, it costs extra, so there is a good chance that a user that has SAS will not be able to run code that uses SAS/IML. I have known of SAS programmers who know IML well that still write matrix/vector tools using macros or proc transpose so that a user without IML can still use the code (the fact that the code that
Re: [R] Inefficiency of SAS Programming
Ajay ohri wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. If you want to market this, Ajay, I'd suggest a name like SASsieR. Jim __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Inefficiency of SAS Programming
Ok. Basically everything that SAS can do, R can do, but vice versa isnt true. using the Anne package just renames the functions into standardized data and proc steps for user comfort. Once SAS user finds that R is productive , and useful and even more powerful for even less money, he can unload the Anne Package , and move straight away into R like intro of R' It is also a good personal exercise for me to learn how to create R packages. Jai ho !!! Ajay You can read more on this concept idea here (note it is an idea not a package now-- and so i have posted on it) http://www.decisionstats.com/2009/03/an-r-package-only-for-sas-users/ On Tue, Mar 3, 2009 at 10:28 AM, Ajay ohri ohri2...@gmail.com wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. It basically is meant only for SAS users who want to learn R , without upsetting the schedule of the corporate users. Simply put , it is a wrapper on SAS language using the function command...ie procunivariate function in Anne package would call the summary function and so on... Regards, Ajay www.decisionstats.com- Show quoted text - On Tue, Mar 3, 2009 at 9:20 AM, Greg Snow greg.s...@imail.org wrote: This does not really address my point. Yes, if the few nerds who want to do funny stuff are the ones making the purchase, then there is a good chance (but still not guaranteed) that they will get IML, but do all companies that buy SAS actually think about that, or do they just see the extra price (no matter how low), or not even think to look at that piece because the person making the purchase does not really the funny things you can do with it. If you want your SAS code to be able to be run by anyone with SAS, you cannot assume that they have IML. If you want your R code to be run by anyone, you cannot make your code dependent on packages/tools that are not available for all platforms. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org 801.408.8111 -Original Message- From: Gerard M. Keogh [mailto:gmke...@justice.ie] Sent: Monday, March 02, 2009 3:22 AM To: Greg Snow Cc: Frank E Harrell Jr; R list; r-help-boun...@r-project.org Subject: Re: [R] Inefficiency of SAS Programming Yes Greg, but if you're buying SAS they'll throw in IML pretty cheaply - SAS think it's only for a few nerds out there who wan to do funny stuff. G Greg Snow greg.s...@imail. org To Sent by: Gerard M. Keogh r-help-boun...@r- gmke...@justice.ie, Frank E project.org Harrell Jr f.harr...@vanderbilt.edu cc 27/02/2009 19:05 r-help-boun...@r-project.org r-help-boun...@r-project.org, R list r-h...@stat.math.ethz.ch Subject Re: [R] Inefficiency of SAS Programming But SAS/IML is not part of base SAS, it costs extra, so there is a good chance that a user that has SAS will not be able to run code that uses SAS/IML. I have known of SAS programmers who know IML well that still write matrix/vector tools using macros or proc transpose so that a user without IML can still use the code (the fact that the code that started this thread was found on a website, suggests that it was meant for general use rather than something only used internally where you know what add-ons will be available). Just another way that R makes life easier for both programmer and user. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org 801.408.8111 -Original Message- From: r-help-boun...@r-project.org [mailto:r-help-boun...@r- project.org] On Behalf Of Gerard M. Keogh Sent: Friday, February 27, 2009 7:19 AM To: Frank E Harrell Jr Cc: r-help-boun...@r-project.org; R list Subject: Re: [R] Inefficiency of SAS Programming Yes Frank, I accept your point but nevertheless IML is the proper place for matrix work in SAS - mixing macro-level logic and computation is another question - R is certainly more seemless in this respect. Gerard Frank E Harrell Jr f.harr...@vander To bilt.edu Gerard M. Keogh gmke...@justice.ie 27/02/2009 13:55 cc R list r- h...@stat.math.ethz.ch,
[R] select nls starting values from list was: nls does not accept start values
Hallo Here is my final solution, but i still wonder if it could not be improved somehow. fit - modely(fol, data.frame(x=x, y=y1) fit or fit - modely(fol, data.frame(x=x, y=y2) fit shows that it arrives to some results. However selection starting values as means of response variable seems to me suboptimal. Is it possible to have some list of functions which will for each item in formula list compute starting values for nls? Or is it this mean guess as good as any other? Petr petr.pi...@precheza.cz # formulas fol - list(y~1/(a-x), y~a*x^2+b*log(x)) set.seed(111) x-1:10 y1 - 1/(0.5-x) +rnorm(length(x))/10 y2 - 2*x^2 + 3*log(x) + rnorm(length(x)) # here is a function modely - function(formula, data, ...){ ll - length(formula) # no of items in formula list # initiation of vectors result2 - vector(list, ll) result1 - rep(NA, ll) for(i in 1:ll) { # some start values ***this shall be improved somehow*** n - length(all.vars(fol[[i]])) - 2 start - as.list(rep(mean(data[,2]),n)) names(start) - letters[1:n] # fit fit-try(nls(formula[[i]], data, start)) if( class(fit)==try-error) result1[i] - NA else result1[i] - sum(resid(fit)^2) if( class(fit)==try-error) result2[[i]] - NA else result2[[i]] - coef(fit) } # ordering and combining results ooo-order(result1) result - mapply(c, sq.resid = result1, result2) names(result) - as.character(formula) result[ooo] } Uwe Ligges lig...@statistik.tu-dortmund.de napsal dne 02.03.2009 10:54:16: Petr PIKAL wrote: Thank you It was simplified version of my problem. I want to elaborate a function which can take predefined list of formulas, some data and evaluate which formulas can fit the data. I was inspired by some article in Chemical engineering in which some guy used excel solver for such task. I was curious if I can do it in R too. I am not sure if nls is appropriate tool for such task but I had to start somewhere. Here is a function which takes list of formulas and data and gives a result for each formula. modely - function(formula, data, ...){ ll - length(formula) #no of items in formula list result2 - vector(list, ll) #prepare results result1 - rep(NA, ll) for(i in 1:ll) { fit-try(nls(formula[[i]], data)) if( class(fit)==try-error) result1[i] - NA else result1[i] - sum(resid(fit)^2) if( class(fit)==try-error) result2[[i]] - NA else result2[[i]] - coef(fit) } ooo-order(result1) #order results according to residual sum #combine results into one list together with functions used result - mapply(c, sq.resid = result1, result2) names(result) - as.character(formula) # output result[ooo] } # data x -1:10 y -1/(.5-x)+rnorm(10)/100 # list of formulas fol - structure(list(a = y ~ 1/(a - x), b = y ~ a * x^2 + b * log(x), c = y ~ x^a), .Names = c(a, b, c)) modely(fol, data.frame(x=x, y=y) does not use correct model because when using default start values it results in nls(fol[[1]], data.frame(x=x, y=y)) Error in numericDeriv(form[[3]], names(ind), env) : Missing value or an infinity produced when evaluating the model however nls(fol[[1]], data.frame(x=x, y=y), start=list(a=mean(y))) gives correct result. Therefore I started think about how to add a better starting value for some fits as a second part of my formula list to define structure like list(a= formula1, start.formula1, b=formula2, start.formula2, ) I wonder If you can push me to better direction. You can make up a list of lists (each containing one formula and its starting values) or specify formulas in one list and starting values in a corresponding second list. You need just the corresponding subsetting in your call to nls such as in the simple case I suggested already. Best, Uwe Thanks again Best regards Petr Uwe Ligges lig...@statistik.tu-dortmund.de napsal dne 02.03.2009 09:41:45: Petr PIKAL wrote: Hi to all OK as I did not get any response and I really need some insight I try again with different subject line I have troubles with correct evaluating/structure of nls input Here is an example # data x -1:10 y -1/(.5-x)+rnorm(10)/100 # formula list form - structure(list(a = list(quote(y ~ 1/(a - x)), list(a=mean(y, .Names = a) # This gives me an error due to not suitable default starting value fit - nls(form [[1]] [[1]], data.frame(x=x, y=y)) # This works and gives me a result fit - nls(form [[1]] [[1]], data.frame(x=x, y=y), start=list(a=mean(y))) *** How to organise list form and call to nls to enable to use other then default starting values***. I thought about something like fit - nls(form [[1]] [[1]], data.frame(x=x, y=y), start=get(form [[1]] [[2]])) ^^^ but this gives me an error so it is not correct syntax. (BTW I tried eval, assign, sustitute,
Re: [R] Inefficiency of SAS Programming
no market for R packages exists in true economic sense as there is demand and supply and utility but no price Ajay Did Tom Sawyer create the first collaborative project ever ( to paint the fence ?) On Tue, Mar 3, 2009 at 4:24 PM, Jim Lemon j...@bitwrit.com.au wrote: Ajay ohri wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. If you want to market this, Ajay, I'd suggest a name like SASsieR. Jim [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Arbirtrary column names with write.csv
Is there a way to prevent write.csv to transform the column names a la make.names? I mean, if I write the code: x - structure (NULL) x [[a+b]] - c (1,2) write.csv (x, file = foo.csv, row.names = FALSE, quote = FALSE) I get in foo.csv: a.b 1 2 but I want: a+b 1 2 Any suggestions? Cheers, -- Rafael Laboissiere __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] monotonic GAM with more than one term
There's an example at the end of the pcls help file in version 1.5-0 --- just submitted to CRAN. best, Simon On Wednesday 25 February 2009 18:48, Benjamin Tyner wrote: Hi, Does anyone know how to fit a GAM where one or more smooth terms are constrained to be monotonic, in the presence of by variables or other terms? I looked at the example in ?pcls but so far have not been able to adapt it to the case where there is more than one predictor. For example, require(mgcv) set.seed(0) n-100 # Generate data from a monotonic truth. x-runif(100)*4-1 x-sort(x) m - structure(rep(1:2,50), .Label=c(one,two), class=factor) f-as.integer(m) * exp(4*x)/(1+exp(4*x)) y-f+rnorm(100)*0.1 plot(x,y,xlim=c(min(x), max(x)*2)) dat-data.frame(x=x,m=m,y=y) # Show regular spline fit (and save fitted object) f.ug-gam(y~m+s(x,k=10,by=m,bs=cr)) bool - m==one yhat - fitted(f.ug) lines(x[bool],yhat[bool]) lines(x[!bool],yhat[!bool]) xx - seq(max(x), 2*max(x), length=100) mm - structure(rep(1:2,50), .Label=c(one,two), class=factor) yy - predict(f.ug, newdata=data.frame(m=mm,x=xx)) bool - mm==one lines(xx[bool],yy[bool], lty=2) # show effect of extrapolation lines(xx[!bool],yy[!bool], lty=2) # this is where I start running into trouble sm-smoothCon(s(x,k=10,by=m,bs=cr),dat,knots=NULL)[[1]] FF-mono.con(sm$xp); # get constraints G-list(y=y, w=rep(1, n), X=sm$X, C=matrix(0,0,0), S = sm$S, off = 0, sp=f.ug$sp, p=sm$xp, Ain = FF$A, bin = FF$b ) p-pcls(G) # fit spline (using s.p. from unconstrained fit) fv-Predict.matrix(sm,data.frame(x=x))%*%p # can we do this without calling smoothCon directly ? # also having trouble here. f.nofit-gam(y~m+s(x,k=10,by=m,bs=cr),fit=FALSE) FF2 - mono.con(f.nofit$smooth[[1]]$xp) stopifnot(identical(FF, FF2)) G2 - list(y = f.nofit$y, w = f.nofit$w, X = f.nofit$X, C = f.nofit$C, S = f.nofit$smooth[[1]]$S, off = f.nofit$off, sp = f.ug$sp, p = f.nofit$smooth[[1]]$xp, Ain = FF2$A, bin = FF2$b ) p2 - pcls(G2) fv2-Predict.matrix(f.nofit$smooth[[1]],data.frame(x=x))%*%p2 Many thanks -Ben __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Simon Wood, Mathematical Sciences, University of Bath, Bath, BA2 7AY UK +44 1225 386603 www.maths.bath.ac.uk/~sw283 __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] barplot with specific order of x axis labels
sort.df() in the reshape packge. cheers, Paul R User R User wrote: Hi all, I'd appreciate your help with this problem. I need to plot a barplot with the categories in a specific order. My data might be: hours Freq AN 10 MO 14 LU 30 I need the categories to be in the order: MO LU AN Is there some way to pass sort the dataset into this order so that it plots correctly when passed to barplot()? Thanks very much, Richie [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Drs. Paul Hiemstra Department of Physical Geography Faculty of Geosciences University of Utrecht Heidelberglaan 2 P.O. Box 80.115 3508 TC Utrecht Phone: +3130 274 3113 Mon-Tue Phone: +3130 253 5773 Wed-Fri http://intamap.geo.uu.nl/~paul __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] R - MATLAB apply like function
Dear all, I very often use the R function apply, for speedup purposes. I am now also using MATLAB, and would like to use the same kind of function. I have already asked MATLAB people, and the answer is : vectorize... but of course, this is not always possible. So, instead of using a FOR loop all the time, I tried using the bsxfun. So you R people, who might also use MATLAB, do you have any solution? Are you also nervous about this fact (maybe also because of this R-MATLAB question). Thanks for your help, Dave [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Self-Organizing Map analysis
Dear list, I read the SOM package manual but I don't understand how to perform (for example) 1) the SOM analysis on Iris data 2) with a visualization similar to that of figure 7 in http://www.cis.hut.fi/projects/somtoolbox/package/papers/techrep.pdf Any suggestion? Thanks in advance, Gianandrea -- View this message in context: http://www.nabble.com/Self-Organizing-Map-analysis-tp22306689p22306689.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] survival::survfit,plot.survfit
You are right. In my case it doesn´t make much difference since the mean of my covariates is about 0. I made some more plots and again I´m confused. I thought when forcing all covariates via the newdata argument to be zero I would get the baseline function since the linear predictor is then zero, too. As I pointed out in another post I´m trying out models with different number of covariates on the same data. For example I use a model with only one covariate. The mean of this covariate is 0.01. So the curves of surv - survfit(coxsst4_full) and attach(nino4) newS4 - data.frame(S5=0.) detach() new_surv - survfit(coxsst4, newdata=newS4) differ very slightly. Of course also the mean curves for models with more then one covariate differe slightly from the Zero-curve. BUT shouldn´t the Zero-curves for all models be the same? #Two models coxsst4 - coxph(Surv(schaeden)~ S5, data=nino4) coxsst4_full - coxph(Surv(schaeden)~ 0+S1+S2+S3+S4+S5+S6+S7+S8+S9+S10, data=nino4) #Set all covariates 0 attach(nino4) newS4 - data.frame(S0=0., S1=0., S2=0., S3=0., S4=0., S5=0., S6=0., S7=0., S8=0., S9=0., S10=0.) detach() new_surv1 - survfit(coxsst4, newdata=newS4) new_surv2 - survfit(coxsst4_full, newdata=newS4) Yields two different curves. What did I get wrong? Regards Bernhard Terry Therneau wrote: plot(survfit(fit)) should plot the survival-function for x=0 or equivalently beta'=0. This curve is independent of any covariates. This is not correct. It plots the curve for a hypothetical subject with x= mean of each covariate. This is NOT the average survival of the data set. Imagine a cohort made up of 60 year old men and their 10 year old grandsons: the expected survival of this cohort does not look that for a 35 year old male. Terry T __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Inefficiency of SAS Programming
2009/3/3 Jim Lemon j...@bitwrit.com.au: Ajay ohri wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. If you want to market this, Ajay, I'd suggest a name like SASsieR. Nice name. I'd call it 'Ursas', which works on several levels: 1. U-R-SAS. (pun on You Are SAS, or You R-SAS) 2. Ur-SAS. (Germanic prefix Ur- meaning 'proto' or 'primitive')[1] 3. Ursas (from Latin 'Ursa' meaning 'bear', because it seems very few R users can bear SAS) Barry [1] http://en.wiktionary.org/wiki/ur- __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] regular expression question
Here are two solutions using gsubfn package. strapply works by matching the what you want rather than what you don't want which may make it easier in this case. The two solutions are the same except we use \\ escapes in the first and [ ... ] in the second, i.e. \\( has the same effect as [(]. In each case we first match the ( then a sequence of characters that is not ] and finally we match the terminating ]. library(gsubfn) x - (-0.791,-0.263].(-38,-1.24].(0.96,2.43] strapply(x, \\([^]]+[]])[[1]] [1] (-0.791,-0.263] (-38,-1.24] (0.96,2.43] strapply(x, [(][^]]+[]])[[1]] [1] (-0.791,-0.263] (-38,-1.24] (0.96,2.43] On Tue, Mar 3, 2009 at 1:16 AM, markle...@verizon.net wrote: can someone show me how to use a regular expression to break the string at the bottom up into its three components : (-0.791,-0.263] (-38,-1.24] (0.96,2.43] I tried to use strplit because of my regexpitis ( it's not curable. i've been to many doctors all over NYC. they tell me there's no cure ) but it doesn't work because there also dots inside the brackets. Thanks. (-0.791,-0.263].(-38,-1.24].(0.96,2.43] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Question about multiple plots of zoo objects
Hello, everyone I have a zoo object containing several time series of daily frequency. One of these timeseries is an indicator function with value (-1) at certain times, and (+1) at the other. I do a plot of several of the timeseries in one go (a multiple plot). I wonder if I can automatically in EACH plot color the area where indicator variable is (-1)? Of course, I could do it simply with layout() and then for each timeseries do plot and a color overlay, but I wonder if with plot.zoo() simething similar is possible to do automatically. Thanks in advance for help! Best, Sergey __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] calculating standard deviation and variance
Hiii, I have some problems to plot standard deviation and variance from a texfile. Ich have the following code so far: x -read.table(file=new.txt) x$gruppe - 100*rep(1:36, each=20) png(filename = D:/Uni/new.png, width = 640, height = 480,pointsize = 12, bg = white) boxplot(V3 ~ gruppe, data = x, col = blue,boxwex=0.7,ylab =Datenrate(MBit(sec)),xlab = Zeit(sec)) dev.off() My question is, how can I plot the standard deviation and variance of x$gruppe - 100*rep(1:36, each=20) Can anybody help me please ? greetings, johnh -- View this message in context: http://www.nabble.com/calculating-standard-deviation-and-variance-tp22306851p22306851.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] xlsReadWrite package repository for Ubuntu
reverend33 wrote: I'm sorry, maybe i didn't explain clearly: i'm trying to install xlsReadWrite on a Linux-type OS (Ubuntu)... I think you didn't understand the answers: the package is only available under Windows. cheers Ben Bolker -- View this message in context: http://www.nabble.com/xlsReadWrite-package-repository-for-Ubuntu-tp22283704p22308383.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Characters To Matrix
Why aren't you using 'read.table'? On Tue, Mar 3, 2009 at 3:47 AM, Bob Roberts quagmire54...@yahoo.com wrote: Hi, I have a data set that I read in using readLines, but R classes it as a character string. The data set has multiple columns and rows, but I only want the first 2 columns in the form of a matrix. Thanks for the help. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Jim Holtman Cincinnati, OH +1 513 646 9390 What is the problem that you are trying to solve? __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Question about multiple plots of zoo objects
Are you trying to color the points themselves? This plots the first two series in frame 1 (they are the same but one is plotted as points and the other as a line) and the third series is shown in frame 2 and for the series of points it colors them green or red. The lines are all colored black: library(zoo) set.seed(1) x - zoo(rnorm(10)) s - sign(x) plot(cbind(x, x, s), screen = c(1, 1, 2), type = c(p, l, l), col = list(ifelse(s 0, green, red), 1, 1), pch = 20) On Tue, Mar 3, 2009 at 7:48 AM, Sergey Goriatchev serg...@gmail.com wrote: Hello, everyone I have a zoo object containing several time series of daily frequency. One of these timeseries is an indicator function with value (-1) at certain times, and (+1) at the other. I do a plot of several of the timeseries in one go (a multiple plot). I wonder if I can automatically in EACH plot color the area where indicator variable is (-1)? Of course, I could do it simply with layout() and then for each timeseries do plot and a color overlay, but I wonder if with plot.zoo() simething similar is possible to do automatically. Thanks in advance for help! Best, Sergey __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] xlsReadWrite package repository for Ubuntu
If your purpose is to read Excel spreadsheets on Linux read.xls in the gdata packages can do that. It uses perl software underneath which is portable across systems. On Tue, Mar 3, 2009 at 7:38 AM, reverend33 periac_r...@hotmail.com wrote: I'm sorry, maybe i didn't explain clearly: i'm trying to install xlsReadWrite on a Linux-type OS (Ubuntu)... Uwe Ligges-3 wrote: reverend33 wrote: Hi, I'm trying to install R on Ubuntu. I succeeded at installing the r-recommended package that is present in the synaptics, but i can't find the xlsReadWrite package in the repositories included in my synaptics manager. Does anybody know a liable repository in which this package is present. If you consider the CRAN master to be liable, it tells you for xlsReadWrite: OS_type: windows Moreover it tells you that the package's status for R-devel is ERROR. Uwe Ligges Thanks in advance __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- View this message in context: http://www.nabble.com/xlsReadWrite-package-repository-for-Ubuntu-tp22283704p22307883.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Question about multiple plots of zoo objects
Hi, Gabor No, what I am trying to do is similar to: abline(v=time(spread)[spread[,Indicator]==(-1)], col=yellow), where spread is the multivariate zoo object (say, 5 timeseries). That is, I want to color parts of the plots where indicator==(-1), but do the coloring without using layout() and then repeating plot() and abline() for each of the 5 timeseries. Best, Sergey On Tue, Mar 3, 2009 at 14:22, Gabor Grothendieck ggrothendi...@gmail.com wrote: Are you trying to color the points themselves? This plots the first two series in frame 1 (they are the same but one is plotted as points and the other as a line) and the third series is shown in frame 2 and for the series of points it colors them green or red. The lines are all colored black: library(zoo) set.seed(1) x - zoo(rnorm(10)) s - sign(x) plot(cbind(x, x, s), screen = c(1, 1, 2), type = c(p, l, l), col = list(ifelse(s 0, green, red), 1, 1), pch = 20) On Tue, Mar 3, 2009 at 7:48 AM, Sergey Goriatchev serg...@gmail.com wrote: Hello, everyone I have a zoo object containing several time series of daily frequency. One of these timeseries is an indicator function with value (-1) at certain times, and (+1) at the other. I do a plot of several of the timeseries in one go (a multiple plot). I wonder if I can automatically in EACH plot color the area where indicator variable is (-1)? Of course, I could do it simply with layout() and then for each timeseries do plot and a color overlay, but I wonder if with plot.zoo() simething similar is possible to do automatically. Thanks in advance for help! Best, Sergey __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- I'm not young enough to know everything. /Oscar Wilde Experience is one thing you can't get for nothing. /Oscar Wilde When you are finished changing, you're finished. /Benjamin Franklin Tell me and I forget, teach me and I remember, involve me and I learn. /Benjamin Franklin Luck is where preparation meets opportunity. /George Patten __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] xlsReadWrite package repository for Ubuntu
You perhaps missed the key point in Uwe's response, which is that the package is only available under Windows, as it depends upon Windows specific functionality (MS Office API via a third party library which is available for Windows only) to natively read and write Excel files. Thus, there is no package version available for Linux, or OSX for that matter. If you need to read Excel files under Linux, you could look at the read.xls() function in the 'gdata' CRAN package. This package requires that Perl be installed, as it calls a Perl routine (xls2csv) for converting the Excel file to a CSV file, which can then be read into R. If you need to write Excel files under Linux, you can use a Perl routine that I had posted back in 2007: https://stat.ethz.ch/pipermail/r-help/2007-July/135968.html and have updated since then to handle Unicode issues. I am attaching a 2k text file here with the updated routine. HTH, Marc Schwartz on 03/03/2009 06:38 AM reverend33 wrote: I'm sorry, maybe i didn't explain clearly: i'm trying to install xlsReadWrite on a Linux-type OS (Ubuntu)... Uwe Ligges-3 wrote: reverend33 wrote: Hi, I'm trying to install R on Ubuntu. I succeeded at installing the r-recommended package that is present in the synaptics, but i can't find the xlsReadWrite package in the repositories included in my synaptics manager. Does anybody know a liable repository in which this package is present. If you consider the CRAN master to be liable, it tells you for xlsReadWrite: OS_type: windows Moreover it tells you that the package's status for R-devel is ERROR. Uwe Ligges #!/usr/bin/perl -w # Called as: WriteXLS.pl [--CSVpath] [--CSVfiles] ExcelFileName # Spreadsheet::WriteExcel # http://search.cpan.org/~jmcnamara/Spreadsheet-WriteExcel/lib/Spreadsheet/WriteExcel.pm # Excel 2007specifications and limitations # http://office.microsoft.com/en-us/excel/HP100738491033.aspx # For unicode issues: # http://www.ahinea.com/en/tech/perl-unicode-struggle.html use strict; use Spreadsheet::WriteExcel; use Getopt::Long; use File::Glob; use File::Basename; use Text::CSV_XS; use Encode; # Initialize and get command line arguments my $CSVPath = '.'; my $CSVFiles = *.csv; GetOptions ('CSVpath=s' = \$CSVPath, 'CSVfiles=s' = \$CSVFiles); my $ExcelFileName = $ARGV[0]; # Create Excel XLS File print Creating Excel File: $ExcelFileName\n\n; my $XLSFile = Spreadsheet::WriteExcel-new($ExcelFileName); # Glob file path and names my @FileNames = $CSVPath/$CSVFiles; foreach my $FileName (@FileNames) { print Reading: $FileName\n; # Open CSV File my $csv = Text::CSV_XS-new ({ binary = 1 }); open (CSVFILE, $FileName) || die ERROR: cannot open $FileName. $!\n; # Create new sheet with filename prefix # ($base, $dir, $ext) = fileparse ($FileName, '..*'); my $FName = (fileparse ($FileName, '\..*'))[0]; # Only take the first 31 chars, which is the # limit for a worksheet name my $SheetName = substr($FName, 0, 31); print Creating New WorkSheet: $SheetName\n\n; my $WorkSheet = $XLSFile-add_worksheet($SheetName); # Rows and columns are zero indexed my $Row = 0; # Write to Sheet while (CSVFILE) { if ($csv-parse($_)) { my @Fields = $csv-fields(); my $Col = 0; foreach my $Fld (@Fields) { $WorkSheet-write($Row, $Col, decode_utf8($Fld)); $Col++; } $Row++; } } close CSVFILE; } __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] R - need more memory, or rejection sampling algorithm doesn't work?
ekwaters wrote: b12=log(.4/.6) b13=log(.85/.15) Can you point out the syntax errors in the first one? You have a double caret (^^). f(k) should be f[k] Watch out for raising numbers to such large powers (although to my surprise things didn't over/underflow when I tried your code). Ben Bolker -- View this message in context: http://www.nabble.com/R---need-more-memory%2C-or-rejection-sampling-algorithm-doesn%27t-work--tp22302800p22308778.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Inefficiency of SAS Programming
Ajay ohri wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. It basically is meant only for SAS users who want to learn R , without upsetting the schedule of the corporate users. Simply put , it is a wrapper on SAS language using the function command...ie procunivariate function in Anne package would call the summary function and so on... Go ahead and add to the confusion. You've already created some by using summary for procunivariate. I created the describe function in the Hmisc package to replace univariate. Frank Regards, Ajay www.decisionstats.com On Tue, Mar 3, 2009 at 9:20 AM, Greg Snow greg.s...@imail.org wrote: This does not really address my point. Yes, if the few nerds who want to do funny stuff are the ones making the purchase, then there is a good chance (but still not guaranteed) that they will get IML, but do all companies that buy SAS actually think about that, or do they just see the extra price (no matter how low), or not even think to look at that piece because the person making the purchase does not really the funny things you can do with it. If you want your SAS code to be able to be run by anyone with SAS, you cannot assume that they have IML. If you want your R code to be run by anyone, you cannot make your code dependent on packages/tools that are not available for all platforms. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org 801.408.8111 -Original Message- From: Gerard M. Keogh [mailto:gmke...@justice.ie] Sent: Monday, March 02, 2009 3:22 AM To: Greg Snow Cc: Frank E Harrell Jr; R list; r-help-boun...@r-project.org Subject: Re: [R] Inefficiency of SAS Programming Yes Greg, but if you're buying SAS they'll throw in IML pretty cheaply - SAS think it's only for a few nerds out there who wan to do funny stuff. G Greg Snow greg.s...@imail. org To Sent by: Gerard M. Keogh r-help-boun...@r- gmke...@justice.ie, Frank E project.org Harrell Jr f.harr...@vanderbilt.edu cc 27/02/2009 19:05 r-help-boun...@r-project.org r-help-boun...@r-project.org, R list r-h...@stat.math.ethz.ch Subject Re: [R] Inefficiency of SAS Programming But SAS/IML is not part of base SAS, it costs extra, so there is a good chance that a user that has SAS will not be able to run code that uses SAS/IML. I have known of SAS programmers who know IML well that still write matrix/vector tools using macros or proc transpose so that a user without IML can still use the code (the fact that the code that started this thread was found on a website, suggests that it was meant for general use rather than something only used internally where you know what add-ons will be available). Just another way that R makes life easier for both programmer and user. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org 801.408.8111 -Original Message- From: r-help-boun...@r-project.org [mailto:r-help-boun...@r- project.org] On Behalf Of Gerard M. Keogh Sent: Friday, February 27, 2009 7:19 AM To: Frank E Harrell Jr Cc: r-help-boun...@r-project.org; R list Subject: Re: [R] Inefficiency of SAS Programming Yes Frank, I accept your point but nevertheless IML is the proper place for matrix work in SAS - mixing macro-level logic and computation is another question - R is certainly more seemless in this respect. Gerard Frank E Harrell Jr f.harr...@vander To bilt.edu Gerard M. Keogh gmke...@justice.ie 27/02/2009 13:55 cc R list r- h...@stat.math.ethz.ch, r-help-boun...@r-project.org Subject Re: [R] Inefficiency of SAS Programming Gerard M. Keogh wrote: Frank, I can't see the code you mention - Web marshall at work - but I don't think you should be too quick to run down SAS - it's a powerful and flexible language but unfortunately very expensive. Your example mentions doing a vector product in the macro language - this only suggest to me that those people writing the code need a crash course in SAS/IML (the matrix language). SAS is designed to work on records and so is inapproprorriate for matrices - macros are only an efficient code copying device. Doing matrix
Re: [R] Question about multiple plots of zoo objects
Do you mean you want to shade a portion of the plot? There are two examples of that in the examples section of ?plot.zoo and a further example using xyplot.zoo in the examples section of ?xyplot.zoo On Tue, Mar 3, 2009 at 8:37 AM, Sergey Goriatchev serg...@gmail.com wrote: Hi, Gabor No, what I am trying to do is similar to: abline(v=time(spread)[spread[,Indicator]==(-1)], col=yellow), where spread is the multivariate zoo object (say, 5 timeseries). That is, I want to color parts of the plots where indicator==(-1), but do the coloring without using layout() and then repeating plot() and abline() for each of the 5 timeseries. Best, Sergey On Tue, Mar 3, 2009 at 14:22, Gabor Grothendieck ggrothendi...@gmail.com wrote: Are you trying to color the points themselves? This plots the first two series in frame 1 (they are the same but one is plotted as points and the other as a line) and the third series is shown in frame 2 and for the series of points it colors them green or red. The lines are all colored black: library(zoo) set.seed(1) x - zoo(rnorm(10)) s - sign(x) plot(cbind(x, x, s), screen = c(1, 1, 2), type = c(p, l, l), col = list(ifelse(s 0, green, red), 1, 1), pch = 20) On Tue, Mar 3, 2009 at 7:48 AM, Sergey Goriatchev serg...@gmail.com wrote: Hello, everyone I have a zoo object containing several time series of daily frequency. One of these timeseries is an indicator function with value (-1) at certain times, and (+1) at the other. I do a plot of several of the timeseries in one go (a multiple plot). I wonder if I can automatically in EACH plot color the area where indicator variable is (-1)? Of course, I could do it simply with layout() and then for each timeseries do plot and a color overlay, but I wonder if with plot.zoo() simething similar is possible to do automatically. Thanks in advance for help! Best, Sergey __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- I'm not young enough to know everything. /Oscar Wilde Experience is one thing you can't get for nothing. /Oscar Wilde When you are finished changing, you're finished. /Benjamin Franklin Tell me and I forget, teach me and I remember, involve me and I learn. /Benjamin Franklin Luck is where preparation meets opportunity. /George Patten __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] SPSS data import: problems work arounds for GSS surveys
Dear Paul, I encountered this problem the other day, and it went away when I updated the foreign package from version 0.8-32 to 0.8-33. I hope this helps, John -- John Fox, Professor Department of Sociology McMaster University Hamilton, Ontario, Canada web: socserv.mcmaster.ca/jfox -Original Message- From: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] On Behalf Of Paul Johnson Sent: March-02-09 10:58 PM To: R-help Subject: [R] SPSS data import: problems work arounds for GSS surveys I'm using R 2.8.1 on Ubuntu 8.10. I'm writing partly to ask what's wrong, partly to tell other users who search that there is a work around. The General Social Survey is a long standing series of surveys provided by NORC (National Opinion Research Center). I have downloaded some years of the survey data in SPSS format (here's the site: http://www.norc.org/GSS+Website/Download/SPSS+Format/). When I try to import using foreign, I get an error like so: library(foreign) dat - read.spss(gss2006.sav, to.data.frame=T, trim.factor.names=T) Error in inherits(x, factor) : object cp not found In addition: Warning messages: 1: In read.spss(gss2006.sav, to.data.frame = T, trim.factor.names = T) : gss2006.sav: File contains duplicate label for value 99.9 for variable TVRELIG 2: In read.spss(gss2006.sav, to.data.frame = T, trim.factor.names = T) : gss2006.sav: File contains duplicate label for value 99.9 for variable SEI 3: In read.spss(gss2006.sav, to.data.frame = T, trim.factor.names = T) : gss2006.sav: File contains duplicate label for value 99.9 for variable FIRSTSEI 4: In read.spss(gss2006.sav, to.data.frame = T, trim.factor.names = T) : gss2006.sav: File contains duplicate label for value 99.9 for variable PASEI 5: In read.spss(gss2006.sav, to.data.frame = T, trim.factor.names = T) : gss2006.sav: File contains duplicate label for value 99.9 for variable MASEI 6: In read.spss(gss2006.sav, to.data.frame = T, trim.factor.names = T) : gss2006.sav: File contains duplicate label for value 99.9 for variable SPSEI 7: In read.spss(gss2006.sav, to.data.frame = T, trim.factor.names = T) : gss2006.sav: File contains duplicate label for value 0.75 for variable YEARSJOB 8: In read.spss(gss2006.sav, to.data.frame = T, trim.factor.names = T) : gss2006.sav: File-indicated character representation code (1252) looks like a Windows codepage No dat object is created from this. I have found a work around. I installed PSPP version 0.6.0 and used it to open the sav file, and then re-save it in SPSS sav format. That creates an SPSS file that foreign's function can open. I still see the warnings about redundant value labels, but as far as I can see these are harmless. A working object is obtained like so: dat - read.spss(gss-pspp.sav) Warning messages: 1: In read.spss(gss-pspp.sav) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable TVRELIG 2: In read.spss(gss-pspp.sav) : gss-pspp.sav: File contains duplicate label for value 0.75 for variable YEARSJOB 3: In read.spss(gss-pspp.sav) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable SEI 4: In read.spss(gss-pspp.sav) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable FIRSTSEI 5: In read.spss(gss-pspp.sav) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable PASEI 6: In read.spss(gss-pspp.sav) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable MASEI 7: In read.spss(gss-pspp.sav) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable SPSEI There is still some trouble with the importation of this SPSS file, however. It has the symptoms of being a non-rectangular data array, I think. What do you think about these warnings: dat - read.spss(gss-pspp.sav,to.data.frame=T) There were 22 warnings (use warnings() to see them) warnings() Warning messages: 1: In read.spss(gss-pspp.sav, to.data.frame = T) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable TVRELIG 2: In read.spss(gss-pspp.sav, to.data.frame = T) : gss-pspp.sav: File contains duplicate label for value 0.75 for variable YEARSJOB 3: In read.spss(gss-pspp.sav, to.data.frame = T) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable SEI 4: In read.spss(gss-pspp.sav, to.data.frame = T) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable FIRSTSEI 5: In read.spss(gss-pspp.sav, to.data.frame = T) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable PASEI 6: In read.spss(gss-pspp.sav, to.data.frame = T) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable MASEI 7: In read.spss(gss-pspp.sav, to.data.frame = T) : gss-pspp.sav: File contains duplicate label for value 99.9 for variable SPSEI 8: In xi = z[1L]
Re: [R] FW: partial residual plots
Dear Dwayne, I didn't see your post until now, because our email system was down all day yesterday. Mark Difford has already pointed your towards the cr.plots() and ceres.plots() functions in the car package. In your description of what you want to do, you mention the adjusted explanatory variable. This sounds more like an added-variable plot than a component + residual plot (although the vertical axis of the AV is not the partial residual). In any event, AV plots are not recommended as diagnostics for nonlinearity. I hope this helps, John -- John Fox, Professor Department of Sociology McMaster University Hamilton, Ontario, Canada web: socserv.mcmaster.ca/jfox -Original Message- From: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] On Behalf Of Culp, Dwayne Sent: March-02-09 9:42 AM To: R-help@r-project.org Subject: [R] FW: partial residual plots I was advised someone might be able to help me with this. Very truly yours, Dwayne E. Culp, P.E., CFM Effective February 8, 2009, my contact information becomes: Dwayne E. Culp, P.E., CFM | JACOBS ENGINEERING GROUP : North American Infrastructure | Manager: Hydrology Hydraulics Section | 5995 Rogerdale Road- B2047B | Houston, TX 77072 | 281-776-2109 | 832.351.7766 Fax | 713.898.1977 Mobile | dwayne.c...@jacobs.com mailto:dwayne.c...@jacobs.com | From: Culp, Dwayne Sent: Friday, February 27, 2009 3:04 PM To: Ted Cleveland; r-c...@r-project.org Subject: partial residual plots Is it possible to produce partial residual plots in R? I am trying to plot the partial residual vs the adjusted explanatory variable in a multiple linear regression to determine if a transformation should be done in the explanatory variable. Very truly yours, Dwayne E. Culp, P.E., CFM | Jacobs North American Infrastructure | Senior Project Manager- Land Development | 5995 Rogerdale Road | Houston, TX 77072| 281.776.2109 | 713.898.1977 Mobile | dwayne.c...@jacobs.com NOTICE - This communication may contain confidential and privileged information that is for the sole use of the intended recipient. Any viewing, copying or distribution of, or reliance on this message by unintended recipients is strictly prohibited. If you have received this message in error, please notify us immediately by replying to the message and deleting it from your computer. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] RES: xlsReadWrite package repository for Ubuntu
I think in this case its better use the write.csv. Microsoft Excel reads csv files normally. Atenciosamente, Leandro Lins Marino Centro de Avaliação Fundação CESGRANRIO Rua Santa Alexandrina, 1011 - 2º andar Rio de Janeiro, RJ - CEP: 20261-903 R (21) 2103-9600 R.:236 ( (21) 8777-7907 ( lean...@cesgranrio.org.br Aquele que suporta o peso da sociedade é precisamente aquele que obtém as menores vantagens. (SMITH, Adam) Antes de imprimir pense em sua responsabilidade e compromisso com o MEIO AMBIENTE -Mensagem original- De: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] Em nome de Marc Schwartz Enviada em: terça-feira, 3 de março de 2009 10:38 Para: reverend33 Cc: r-help@r-project.org Assunto: Re: [R] xlsReadWrite package repository for Ubuntu You perhaps missed the key point in Uwe's response, which is that the package is only available under Windows, as it depends upon Windows specific functionality (MS Office API via a third party library which is available for Windows only) to natively read and write Excel files. Thus, there is no package version available for Linux, or OSX for that matter. If you need to read Excel files under Linux, you could look at the read.xls() function in the 'gdata' CRAN package. This package requires that Perl be installed, as it calls a Perl routine (xls2csv) for converting the Excel file to a CSV file, which can then be read into R. If you need to write Excel files under Linux, you can use a Perl routine that I had posted back in 2007: https://stat.ethz.ch/pipermail/r-help/2007-July/135968.html and have updated since then to handle Unicode issues. I am attaching a 2k text file here with the updated routine. HTH, Marc Schwartz on 03/03/2009 06:38 AM reverend33 wrote: I'm sorry, maybe i didn't explain clearly: i'm trying to install xlsReadWrite on a Linux-type OS (Ubuntu)... Uwe Ligges-3 wrote: reverend33 wrote: Hi, I'm trying to install R on Ubuntu. I succeeded at installing the r-recommended package that is present in the synaptics, but i can't find the xlsReadWrite package in the repositories included in my synaptics manager. Does anybody know a liable repository in which this package is present. If you consider the CRAN master to be liable, it tells you for xlsReadWrite: OS_type: windows Moreover it tells you that the package's status for R-devel is ERROR. Uwe Ligges __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Reshape
Hi all, I would like to transform to long format a matrix which has only information about individuals and a value for each individual. I would like to have it in ling format with more information related to groups, so ell values and individuals are repeated for each group. Let me show the example: matrix in wide format: IndividualsValue A3 B4 C5 D2 And long format would be like: IndividualsValueGroup A3Group1 B4Group1 C5Group1 D2Group1 A3Group2 B4Group2 C5Group2 D2Group2 My case is of course very large so i thought this could be possible with reshape function by adding a vector with the names of the groups or something like that. Thanks in advance. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] RES: xlsReadWrite package repository for Ubuntu
That is fine if you only have one or two R objects to write out to CSV files and then read them into Excel. It becomes rapidly tedious as the number of objects increases. If you go back and read my original post, the incentive for me to write the Perl routine was that I had to create an Excel file for clients that contained a large number of tabs, each tab containing data from an R data frame. My clients are on Windows, I have been on Linux. In many cases, the number of data frames was 20. Thus, creating that number of CSV files, importing each one into a separate tab into a single Excel and naming each tab appropriately (in my case, using OO.org's Calc) became very tedious. The Perl routing automates that process, saving a great deal of time and reducing the potential for error. Regards, Marc on 03/03/2009 07:53 AM Leandro Marino wrote: I think in this case its better use the write.csv. Microsoft Excel reads csv files normally. -Mensagem original- De: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] Em nome de Marc Schwartz Enviada em: terça-feira, 3 de março de 2009 10:38 Para: reverend33 Cc: r-help@r-project.org Assunto: Re: [R] xlsReadWrite package repository for Ubuntu You perhaps missed the key point in Uwe's response, which is that the package is only available under Windows, as it depends upon Windows specific functionality (MS Office API via a third party library which is available for Windows only) to natively read and write Excel files. Thus, there is no package version available for Linux, or OSX for that matter. If you need to read Excel files under Linux, you could look at the read.xls() function in the 'gdata' CRAN package. This package requires that Perl be installed, as it calls a Perl routine (xls2csv) for converting the Excel file to a CSV file, which can then be read into R. If you need to write Excel files under Linux, you can use a Perl routine that I had posted back in 2007: https://stat.ethz.ch/pipermail/r-help/2007-July/135968.html and have updated since then to handle Unicode issues. I am attaching a 2k text file here with the updated routine. HTH, Marc Schwartz on 03/03/2009 06:38 AM reverend33 wrote: I'm sorry, maybe i didn't explain clearly: i'm trying to install xlsReadWrite on a Linux-type OS (Ubuntu)... Uwe Ligges-3 wrote: reverend33 wrote: Hi, I'm trying to install R on Ubuntu. I succeeded at installing the r-recommended package that is present in the synaptics, but i can't find the xlsReadWrite package in the repositories included in my synaptics manager. Does anybody know a liable repository in which this package is present. If you consider the CRAN master to be liable, it tells you for xlsReadWrite: OS_type:windows Moreover it tells you that the package's status for R-devel is ERROR. Uwe Ligges __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Question about multiple plots of zoo objects
Gabor, yes, I want to color portions of EACH plot of the MULTIPLE plot done with plot.zoo() I tried to do: plot(multivariate zoo object) abline(v=...) but that does not work. I will check your suggestions of the examples. Thank you for your help, as always! Best, Sergey On Tue, Mar 3, 2009 at 14:44, Gabor Grothendieck ggrothendi...@gmail.com wrote: Do you mean you want to shade a portion of the plot? There are two examples of that in the examples section of ?plot.zoo and a further example using xyplot.zoo in the examples section of ?xyplot.zoo On Tue, Mar 3, 2009 at 8:37 AM, Sergey Goriatchev serg...@gmail.com wrote: Hi, Gabor No, what I am trying to do is similar to: abline(v=time(spread)[spread[,Indicator]==(-1)], col=yellow), where spread is the multivariate zoo object (say, 5 timeseries). That is, I want to color parts of the plots where indicator==(-1), but do the coloring without using layout() and then repeating plot() and abline() for each of the 5 timeseries. Best, Sergey On Tue, Mar 3, 2009 at 14:22, Gabor Grothendieck ggrothendi...@gmail.com wrote: Are you trying to color the points themselves? This plots the first two series in frame 1 (they are the same but one is plotted as points and the other as a line) and the third series is shown in frame 2 and for the series of points it colors them green or red. The lines are all colored black: library(zoo) set.seed(1) x - zoo(rnorm(10)) s - sign(x) plot(cbind(x, x, s), screen = c(1, 1, 2), type = c(p, l, l), col = list(ifelse(s 0, green, red), 1, 1), pch = 20) On Tue, Mar 3, 2009 at 7:48 AM, Sergey Goriatchev serg...@gmail.com wrote: Hello, everyone I have a zoo object containing several time series of daily frequency. One of these timeseries is an indicator function with value (-1) at certain times, and (+1) at the other. I do a plot of several of the timeseries in one go (a multiple plot). I wonder if I can automatically in EACH plot color the area where indicator variable is (-1)? Of course, I could do it simply with layout() and then for each timeseries do plot and a color overlay, but I wonder if with plot.zoo() simething similar is possible to do automatically. Thanks in advance for help! Best, Sergey __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- I'm not young enough to know everything. /Oscar Wilde Experience is one thing you can't get for nothing. /Oscar Wilde When you are finished changing, you're finished. /Benjamin Franklin Tell me and I forget, teach me and I remember, involve me and I learn. /Benjamin Franklin Luck is where preparation meets opportunity. /George Patten -- I'm not young enough to know everything. /Oscar Wilde Experience is one thing you can't get for nothing. /Oscar Wilde When you are finished changing, you're finished. /Benjamin Franklin Tell me and I forget, teach me and I remember, involve me and I learn. /Benjamin Franklin Luck is where preparation meets opportunity. /George Patten __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Reshape
check out the melt function in the reshape package. On Tue, Mar 3, 2009 at 8:53 AM, Usuario R r.user.sp...@gmail.com wrote: Hi all, I would like to transform to long format a matrix which has only information about individuals and a value for each individual. I would like to have it in ling format with more information related to groups, so ell values and individuals are repeated for each group. Let me show the example: matrix in wide format: Individuals Value A 3 B 4 C 5 D 2 And long format would be like: Individuals Value Group A 3 Group1 B 4 Group1 C 5 Group1 D 2 Group1 A 3 Group2 B 4 Group2 C 5 Group2 D 2 Group2 My case is of course very large so i thought this could be possible with reshape function by adding a vector with the names of the groups or something like that. Thanks in advance. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Stephen Sefick Let's not spend our time and resources thinking about things that are so little or so large that all they really do for us is puff us up and make us feel like gods. We are mammals, and have not exhausted the annoying little problems of being mammals. -K. Mullis __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] RES: xlsReadWrite package repository for Ubuntu
There is a recently updated review of the alternatives on the R Wiki: http://wiki.r-project.org/rwiki/doku.php?id=tips:data-io:ms_windows On Tue, Mar 3, 2009 at 9:00 AM, Marc Schwartz marc_schwa...@comcast.net wrote: That is fine if you only have one or two R objects to write out to CSV files and then read them into Excel. It becomes rapidly tedious as the number of objects increases. If you go back and read my original post, the incentive for me to write the Perl routine was that I had to create an Excel file for clients that contained a large number of tabs, each tab containing data from an R data frame. My clients are on Windows, I have been on Linux. In many cases, the number of data frames was 20. Thus, creating that number of CSV files, importing each one into a separate tab into a single Excel and naming each tab appropriately (in my case, using OO.org's Calc) became very tedious. The Perl routing automates that process, saving a great deal of time and reducing the potential for error. Regards, Marc on 03/03/2009 07:53 AM Leandro Marino wrote: I think in this case its better use the write.csv. Microsoft Excel reads csv files normally. -Mensagem original- De: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] Em nome de Marc Schwartz Enviada em: terça-feira, 3 de março de 2009 10:38 Para: reverend33 Cc: r-help@r-project.org Assunto: Re: [R] xlsReadWrite package repository for Ubuntu You perhaps missed the key point in Uwe's response, which is that the package is only available under Windows, as it depends upon Windows specific functionality (MS Office API via a third party library which is available for Windows only) to natively read and write Excel files. Thus, there is no package version available for Linux, or OSX for that matter. If you need to read Excel files under Linux, you could look at the read.xls() function in the 'gdata' CRAN package. This package requires that Perl be installed, as it calls a Perl routine (xls2csv) for converting the Excel file to a CSV file, which can then be read into R. If you need to write Excel files under Linux, you can use a Perl routine that I had posted back in 2007: https://stat.ethz.ch/pipermail/r-help/2007-July/135968.html and have updated since then to handle Unicode issues. I am attaching a 2k text file here with the updated routine. HTH, Marc Schwartz on 03/03/2009 06:38 AM reverend33 wrote: I'm sorry, maybe i didn't explain clearly: i'm trying to install xlsReadWrite on a Linux-type OS (Ubuntu)... Uwe Ligges-3 wrote: reverend33 wrote: Hi, I'm trying to install R on Ubuntu. I succeeded at installing the r-recommended package that is present in the synaptics, but i can't find the xlsReadWrite package in the repositories included in my synaptics manager. Does anybody know a liable repository in which this package is present. If you consider the CRAN master to be liable, it tells you for xlsReadWrite: OS_type: windows Moreover it tells you that the package's status for R-devel is ERROR. Uwe Ligges __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] xlsReadWrite package repository for Ubuntu
On 3 March 2009 at 07:38, Marc Schwartz wrote: | You perhaps missed the key point in Uwe's response, which is that the | package is only available under Windows, as it depends upon Windows | specific functionality (MS Office API via a third party library which is | available for Windows only) to natively read and write Excel files. | Thus, there is no package version available for Linux, or OSX for that | matter. | | If you need to read Excel files under Linux, you could look at the | read.xls() function in the 'gdata' CRAN package. This package requires | that Perl be installed, as it calls a Perl routine (xls2csv) for | converting the Excel file to a CSV file, which can then be read into R. Also, 'sudo apt-get install r-cran-gdata' installs that package on Ubuntu. The package contains the Perl code. Years ago, I also added the appropriate Perl modules to read and write xls files to Debian, these are still available in Ubuntu so one could write a matching write.xls and contribute it back to the gdata package. Finally, note that there is an entire R manual devoted to Data Import questions. Dirk -- Three out of two people have difficulties with fractions. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Inefficiency of SAS Programming
On Mar 3, 9:58 am, Ajay ohri ohri2...@gmail.com wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. It basically is meant only for SAS users who want to learn R , without upsetting the schedule of the corporate users. Simply put , it is a wrapper on SAS language using the function command...ie procunivariate function in Anne package would call the summary function and so on... Regards, Ajay www.decisionstats.com = Bob Muenchen's book R for SAS and SPSS users provides a systematic transition plan (if I may use that term) for SAS and SPSS users intending to work on/migrate to R. Having been a newly transformed R user myself, I'm inclined to believe that creating yet another package that just houses some SAS procedures--sounding names for data manipulation/summarization would add to fair bit of confusion. my $.02.. -Girish __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Arbirtrary column names with write.csv
* Uwe Ligges lig...@statistik.tu-dortmund.de [2009-03-03 15:27]: Rafael Laboissiere wrote: Is there a way to prevent write.csv to transform the column names a la make.names? I mean, if I write the code: x - structure (NULL) x [[a+b]] - c (1,2) write.csv (x, file = foo.csv, row.names = FALSE, quote = FALSE) your x is not a data.frame, hence write.csv() (or better, write.table()) converts your *list* x to a data.frame implicitly by using as.data.frame (and this tries to convert names to more regular names by make.names()). You can avoid it by either working on data.frames or stop the make.names() conversion by forcing the data.frame in advance as in: write.csv (as.data.frame(x, optional=TRUE), file = foo.csv, row.names = FALSE, quote = FALSE) Great, it works. Thanks! -- Rafael Laboissiere __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Reshape
If you have many such repetitions it can be annoying to type the rbind(cbind ... sequence. Perhaps this would help, m - read.table(textConnection( IndividualsValue A3 B4 C5 D2), head=T) data.frame(groups= rep(c(group1, group2, group3), each=nrow(m)), Ind=m[, 1], val=m[, 2]) You could also try Hadley's plyr package, rdply(3,m) # you might want to rename the column name and make(.n) a factor with adequate levels baptiste On 3 Mar 2009, at 14:28, David Winsemius wrote: Your data objects could not be matrices with that composition of values. Let's use the correct data type: df - data.frame(Ind = letters[1:4], val = sample(1:4, 4)) df Ind val 1 a 4 2 b 2 3 c 3 4 d 1 rbind(cbind(df,GruppeEin),cbind(df,GruppeZwei)) #v ariable recycling Ind val y 1 a 4 GruppeEin 2 b 2 GruppeEin 3 c 3 GruppeEin 4 d 1 GruppeEin 5 a 4 GruppeZwei 6 b 2 GruppeZwei 7 c 3 GruppeZwei 8 d 1 GruppeZwei is.data.frame(rbind(cbind(df,GruppeEin),cbind(df,GruppeZwei))) [1] TRUE -- David Winsemius On Mar 3, 2009, at 8:53 AM, Usuario R wrote: Hi all, I would like to transform to long format a matrix which has only information about individuals and a value for each individual. I would like to have it in ling format with more information related to groups, so ell values and individuals are repeated for each group. Let me show the example: matrix in wide format: IndividualsValue A3 B4 C5 D2 And long format would be like: IndividualsValueGroup A3Group1 B4Group1 C5Group1 D2Group1 A3Group2 B4Group2 C5Group2 D2Group2 My case is of course very large so i thought this could be possible with reshape function by adding a vector with the names of the groups or something like that. Thanks in advance. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. _ Baptiste Auguié School of Physics University of Exeter Stocker Road, Exeter, Devon, EX4 4QL, UK Phone: +44 1392 264187 http://newton.ex.ac.uk/research/emag __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Question about multiple plots of zoo objects
Supreme! Thanks Gabor! On Tue, Mar 3, 2009 at 15:20, Gabor Grothendieck ggrothendi...@gmail.com wrote: If you want to use abline on a multivariate plot it must be issued within a panel function like this: # same set up as in prior email p - function(x, y, ...) { points(x, y); lines(x, y); abline(v = time(s)[s 0], col = green) } plot(cbind(x, s), panel = p) There are further examples of panel functions in the examples section of ?plot.zoo in case what you want is a variation of the above. On Tue, Mar 3, 2009 at 9:04 AM, Sergey Goriatchev serg...@gmail.com wrote: Gabor, yes, I want to color portions of EACH plot of the MULTIPLE plot done with plot.zoo() I tried to do: plot(multivariate zoo object) abline(v=...) but that does not work. I will check your suggestions of the examples. Thank you for your help, as always! Best, Sergey On Tue, Mar 3, 2009 at 14:44, Gabor Grothendieck ggrothendi...@gmail.com wrote: Do you mean you want to shade a portion of the plot? There are two examples of that in the examples section of ?plot.zoo and a further example using xyplot.zoo in the examples section of ?xyplot.zoo On Tue, Mar 3, 2009 at 8:37 AM, Sergey Goriatchev serg...@gmail.com wrote: Hi, Gabor No, what I am trying to do is similar to: abline(v=time(spread)[spread[,Indicator]==(-1)], col=yellow), where spread is the multivariate zoo object (say, 5 timeseries). That is, I want to color parts of the plots where indicator==(-1), but do the coloring without using layout() and then repeating plot() and abline() for each of the 5 timeseries. Best, Sergey On Tue, Mar 3, 2009 at 14:22, Gabor Grothendieck ggrothendi...@gmail.com wrote: Are you trying to color the points themselves? This plots the first two series in frame 1 (they are the same but one is plotted as points and the other as a line) and the third series is shown in frame 2 and for the series of points it colors them green or red. The lines are all colored black: library(zoo) set.seed(1) x - zoo(rnorm(10)) s - sign(x) plot(cbind(x, x, s), screen = c(1, 1, 2), type = c(p, l, l), col = list(ifelse(s 0, green, red), 1, 1), pch = 20) On Tue, Mar 3, 2009 at 7:48 AM, Sergey Goriatchev serg...@gmail.com wrote: Hello, everyone I have a zoo object containing several time series of daily frequency. One of these timeseries is an indicator function with value (-1) at certain times, and (+1) at the other. I do a plot of several of the timeseries in one go (a multiple plot). I wonder if I can automatically in EACH plot color the area where indicator variable is (-1)? Of course, I could do it simply with layout() and then for each timeseries do plot and a color overlay, but I wonder if with plot.zoo() simething similar is possible to do automatically. Thanks in advance for help! Best, Sergey __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- I'm not young enough to know everything. /Oscar Wilde Experience is one thing you can't get for nothing. /Oscar Wilde When you are finished changing, you're finished. /Benjamin Franklin Tell me and I forget, teach me and I remember, involve me and I learn. /Benjamin Franklin Luck is where preparation meets opportunity. /George Patten -- I'm not young enough to know everything. /Oscar Wilde Experience is one thing you can't get for nothing. /Oscar Wilde When you are finished changing, you're finished. /Benjamin Franklin Tell me and I forget, teach me and I remember, involve me and I learn. /Benjamin Franklin Luck is where preparation meets opportunity. /George Patten -- I'm not young enough to know everything. /Oscar Wilde Experience is one thing you can't get for nothing. /Oscar Wilde When you are finished changing, you're finished. /Benjamin Franklin Tell me and I forget, teach me and I remember, involve me and I learn. /Benjamin Franklin Luck is where preparation meets opportunity. /George Patten __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] REMINDER: DSC 2009 abstract submission before March 15
Just a friendly reminder that the closing date for abstract submissions for the DSC conference in Copenhagen is next Sunday. See http://www.r-project.org/dsc-2009 Notice that the website has been revised since we decided to engage a professional convention bureau. The good news is that payment by credit card is now possible. The slightly bad news is that there are now two separate registration paths: one for participation and one for abstracts. -- 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 ~~ - (p.dalga...@biostat.ku.dk) FAX: (+45) 35327907 ___ r-annou...@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-announce __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] New R mailing list: R-SIG-insurance
Dear useRs, Over the last few years R has gained a significant momentum in the insurance industry. Dedicated packages have been written, numerous presentations given, courses held, businesses set up, and IT departments convinced. ;-) Therefore, together with Markus Gesmann of Lloyd's of London, we created a Special Interest Group mailing list to discuss R usage in actuarial science, insurance (life and non-life) and risk analysis and assessment in general. (With thanks to Diego Mazzeo for the idea and to Martin Maechler for the technical help and list creation.) To subscribe to the list please visit: https://stat.ethz.ch/mailman/listinfo/r-sig-insurance For the posting guide see: http://www.r-project.org/posting-guide.html We hope that this new list will foster discussions and the transfer of knowledge. Please pass the word and participate, your contributions are most welcome! -- Vincent Goulet, Associate Professor École d'actuariat Université Laval, Québec vincent.gou...@act.ulaval.ca http://vgoulet.act.ulaval.ca __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Inefficiency of SAS Programming
why didnt you call it procunivariate if that was exactly what you wanted to do . On Tue, Mar 3, 2009 at 7:11 PM, Frank E Harrell Jr f.harr...@vanderbilt.edu wrote: Ajay ohri wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. It basically is meant only for SAS users who want to learn R , without upsetting the schedule of the corporate users. Simply put , it is a wrapper on SAS language using the function command...ie procunivariate function in Anne package would call the summary function and so on... Go ahead and add to the confusion. You've already created some by using summary for procunivariate. I created the describe function in the Hmisc package to replace univariate. Frank Regards, Ajay www.decisionstats.com On Tue, Mar 3, 2009 at 9:20 AM, Greg Snow greg.s...@imail.org wrote: This does not really address my point. Yes, if the few nerds who want to do funny stuff are the ones making the purchase, then there is a good chance (but still not guaranteed) that they will get IML, but do all companies that buy SAS actually think about that, or do they just see the extra price (no matter how low), or not even think to look at that piece because the person making the purchase does not really the funny things you can do with it. If you want your SAS code to be able to be run by anyone with SAS, you cannot assume that they have IML. If you want your R code to be run by anyone, you cannot make your code dependent on packages/tools that are not available for all platforms. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org 801.408.8111 -Original Message- From: Gerard M. Keogh [mailto:gmke...@justice.ie] Sent: Monday, March 02, 2009 3:22 AM To: Greg Snow Cc: Frank E Harrell Jr; R list; r-help-boun...@r-project.org Subject: Re: [R] Inefficiency of SAS Programming Yes Greg, but if you're buying SAS they'll throw in IML pretty cheaply - SAS think it's only for a few nerds out there who wan to do funny stuff. G Greg Snow greg.s...@imail. org To Sent by: Gerard M. Keogh r-help-boun...@r- gmke...@justice.ie, Frank E project.org Harrell Jr f.harr...@vanderbilt.edu cc 27/02/2009 19:05 r-help-boun...@r-project.org r-help-boun...@r-project.org, R list r-h...@stat.math.ethz.ch Subject Re: [R] Inefficiency of SAS Programming But SAS/IML is not part of base SAS, it costs extra, so there is a good chance that a user that has SAS will not be able to run code that uses SAS/IML. I have known of SAS programmers who know IML well that still write matrix/vector tools using macros or proc transpose so that a user without IML can still use the code (the fact that the code that started this thread was found on a website, suggests that it was meant for general use rather than something only used internally where you know what add-ons will be available). Just another way that R makes life easier for both programmer and user. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org 801.408.8111 -Original Message- From: r-help-boun...@r-project.org [mailto:r-help-boun...@r- project.org] On Behalf Of Gerard M. Keogh Sent: Friday, February 27, 2009 7:19 AM To: Frank E Harrell Jr Cc: r-help-boun...@r-project.org; R list Subject: Re: [R] Inefficiency of SAS Programming Yes Frank, I accept your point but nevertheless IML is the proper place for matrix work in SAS - mixing macro-level logic and computation is another question - R is certainly more seemless in this respect. Gerard Frank E Harrell Jr f.harr...@vander To bilt.edu Gerard M. Keogh gmke...@justice.ie 27/02/2009 13:55 cc R list r- h...@stat.math.ethz.ch, r-help-boun...@r-project.org Subject Re: [R] Inefficiency of SAS Programming Gerard M. Keogh wrote: Frank, I can't see the code you mention - Web marshall at work - but I don't think you should be too quick to run down SAS - it's a powerful and flexible language but unfortunately very expensive. Your example mentions doing a vector product in the macro language -
[R] SAS Macros for R Users Only
I think SAS Macros has capability to call R, and execute it without it being in the picture anywhere. So you can use SAS Macros in a file called R.sas In this you can create a macro called %Describe that can call R , load Hmisc ,run the describe function Note you will need repeated use of %put in this %describe for the mapping to take place Use %INCLUDE to include that file in all SAS Programs In fact there is a 50 dollar program written by Phil Rack at www.minequest.com which does exactly that see more here http://www.decisionstats.com/2009/02/interview-phil-rack/ On Tue, Mar 3, 2009 at 8:53 PM, Ajay ohri ohri2...@gmail.com wrote: why didnt you call it procunivariate if that was exactly what you wanted to do . On Tue, Mar 3, 2009 at 7:11 PM, Frank E Harrell Jr f.harr...@vanderbilt.edu wrote: Ajay ohri wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. It basically is meant only for SAS users who want to learn R , without upsetting the schedule of the corporate users. Simply put , it is a wrapper on SAS language using the function command...ie procunivariate function in Anne package would call the summary function and so on... Go ahead and add to the confusion. You've already created some by using summary for procunivariate. I created the describe function in the Hmisc package to replace univariate. Frank Regards, Ajay www.decisionstats.com On Tue, Mar 3, 2009 at 9:20 AM, Greg Snow greg.s...@imail.org wrote: This does not really address my point. Yes, if the few nerds who want to do funny stuff are the ones making the purchase, then there is a good chance (but still not guaranteed) that they will get IML, but do all companies that buy SAS actually think about that, or do they just see the extra price (no matter how low), or not even think to look at that piece because the person making the purchase does not really the funny things you can do with it. If you want your SAS code to be able to be run by anyone with SAS, you cannot assume that they have IML. If you want your R code to be run by anyone, you cannot make your code dependent on packages/tools that are not available for all platforms. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org 801.408.8111 -Original Message- From: Gerard M. Keogh [mailto:gmke...@justice.ie] Sent: Monday, March 02, 2009 3:22 AM To: Greg Snow Cc: Frank E Harrell Jr; R list; r-help-boun...@r-project.org Subject: Re: [R] Inefficiency of SAS Programming Yes Greg, but if you're buying SAS they'll throw in IML pretty cheaply - SAS think it's only for a few nerds out there who wan to do funny stuff. G Greg Snow greg.s...@imail. org To Sent by: Gerard M. Keogh r-help-boun...@r- gmke...@justice.ie, Frank E project.org Harrell Jr f.harr...@vanderbilt.edu cc 27/02/2009 19:05 r-help-boun...@r-project.org r-help-boun...@r-project.org, R list r-h...@stat.math.ethz.ch Subject Re: [R] Inefficiency of SAS Programming But SAS/IML is not part of base SAS, it costs extra, so there is a good chance that a user that has SAS will not be able to run code that uses SAS/IML. I have known of SAS programmers who know IML well that still write matrix/vector tools using macros or proc transpose so that a user without IML can still use the code (the fact that the code that started this thread was found on a website, suggests that it was meant for general use rather than something only used internally where you know what add-ons will be available). Just another way that R makes life easier for both programmer and user. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org 801.408.8111 -Original Message- From: r-help-boun...@r-project.org [mailto:r-help-boun...@r- project.org] On Behalf Of Gerard M. Keogh Sent: Friday, February 27, 2009 7:19 AM To: Frank E Harrell Jr Cc: r-help-boun...@r-project.org; R list Subject: Re: [R] Inefficiency of SAS Programming Yes Frank, I accept your point but nevertheless IML is the proper place for matrix work in SAS - mixing macro-level logic and computation is another question - R is certainly more seemless in this respect. Gerard Frank E Harrell Jr f.harr...@vander To bilt.edu Gerard M. Keogh
[R] spatial markov chain methods
Hello, can any one point me to R-packages (if available) which include spatial Markov Chain methods? My second question is more general but hopefully not OT: Currently we are using the software TPROGS, which let people simulate property distributions in space by some Markov Chain approaches. We face some problems due to the lack of information between distances of samples along borehole path and among boreholes. Is there any way to trick it like for instants in variogram modeling were one might use nested variograms...? Thank you, Torsten __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] xlsReadWrite package repository for Ubuntu
Thanks a lot for all your answers. Indeed, i hadn't understood the previous answers. I just downloaded the gdata package and it works perfectly now. Thanks again. Gabor Grothendieck wrote: If your purpose is to read Excel spreadsheets on Linux read.xls in the gdata packages can do that. It uses perl software underneath which is portable across systems. On Tue, Mar 3, 2009 at 7:38 AM, reverend33 periac_r...@hotmail.com wrote: I'm sorry, maybe i didn't explain clearly: i'm trying to install xlsReadWrite on a Linux-type OS (Ubuntu)... Uwe Ligges-3 wrote: reverend33 wrote: Hi, I'm trying to install R on Ubuntu. I succeeded at installing the r-recommended package that is present in the synaptics, but i can't find the xlsReadWrite package in the repositories included in my synaptics manager. Does anybody know a liable repository in which this package is present. If you consider the CRAN master to be liable, it tells you for xlsReadWrite: OS_type: windows Moreover it tells you that the package's status for R-devel is ERROR. Uwe Ligges Thanks in advance __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- View this message in context: http://www.nabble.com/xlsReadWrite-package-repository-for-Ubuntu-tp22283704p22307883.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- View this message in context: http://www.nabble.com/xlsReadWrite-package-repository-for-Ubuntu-tp22283704p22309785.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] PLS regression on near infrared (NIR) spectra data
Dear collegues, I´ ve worked with near infrared (NIR) spectroscopy to assess chemical, physical, mechanical and anatomical properties of wood. I use The Unscrambler software to correlate the matrix of dependent variables (Y) with the matrix of spectral data (X) and I would like to migrate to R. The matrix of spectral variables is very large (2345 columns and n lines, where n = samples), so we used Partial Least Squares Regression to predict a variable y (content of cellulose, for instance) based on the spectral variables, which are the NIR wavelengths. I am new here (since jan2009) and up to now, I not seen anyone commenting about principal component analysis and regression PLS to analyze spectral information in R system. Sorry, I am a R starter... Anybody have any package, or trick to suggest me? Grateful for yours information! -- Paulo Ricardo Gherardi Hein PhD candidate at University of Montpellier 2 CIRAD - PERSYST Department Research unit: Production and Processing of Tropical Woods - TA B-40/16 73 rue Jean-François Breton 34398 Montpellier Cedex 5, France phone: +33 4 67 61 44 51 skype: paulo_hein email: paulo.h...@cirad.fr [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] behavior of squishplot in TeachingDemos
Hi list, I wonder if anyone has had this experience with squishplot() in the TeachingDemos package. Taking the example from the ?image help page, library(TeachingDemos) x - 10*(1:nrow(volcano)) y - 10*(1:ncol(volcano)) layout(matrix(c(1,2,3,4),ncol=2,byrow=TRUE),height=c(2,1)) ## 1st plot op - squishplot(range(x),range(y),1) image(x, y, volcano, col = terrain.colors(100)) par(op) ## 2nd plot op - squishplot(range(x),range(y),1) image(x, y, volcano, col = terrain.colors(100)) par(op) The second plot comes out looking as expected, but the first plot is not squished in the desired proportions. I tried tracking the modifications to par('pin') and par('plt') in the function but gave up midway through in desire for haste - not sure what is going on but I did find that taking advantage of the behavior above, calling plot.new(); par(new=TRUE) before the first plot makes things work as expected. So the full code would be layout(matrix(c(1,2,3,4),ncol=2,byrow=TRUE),height=c(2,1)) ## 1st plot op - squishplot(range(x),range(y),1) plot.new() par(new=TRUE) image(x, y, volcano, col = terrain.colors(100)) par(op) ## 2nd plot op - squishplot(range(x),range(y),1) image(x, y, volcano, col = terrain.colors(100)) par(op) +++ I wonder if this behavior is not surprising? It is a great function overall though - thanks for the contribution. Stephen __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Mantel test!
Dear Gavin, What is the interpretation of a simulated p-value of 1 in a mantel test and how is the p-value derived? When we run two highly negatively correlated matrices we get this result: r: -1, simulated p value: 1 I would expect the high p-value of 1 to mean not significant. Could you clarify? Thank you, Jason Jason P. Sexton Graduate Group in Ecology University of California, Davis (530) 752-1701 jpsex...@ucdavis.edu __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Comparing two matrices
Hi List, I would like to compare two (confusion) matrices. I ended up with the ade4 mantel.rtest function as my best option. However it seems that my data is non-euclidean in nature (is.euclid gives a FALSE). I attached both matrices at the end of the document. Are there any tools to transform these matrices into an euclidean form so I can use mantel.rtest with confidence. Other suggestions are also welcome. Kind regards, Koen -- matrix 1 V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 1 0 60 5 0 9 3 3 0 18 1 1 2 30 0 12 0 2 45 1 0 5 4 1 3 8 32 0 1 19 5 4 0 27 1 3 4 0 0 3 0 67 1 3 0 0 0 25 5 12 3 13 26 0 12 27 0 3 2 2 6 1 72 1 0 4 0 11 0 0 10 1 7 3 2 9 0 45 40 0 0 0 0 0 8 13 3 0 0 0 1 0 0 10 0 72 9 24 26 34 0 6 2 0 2 0 2 4 10 2 21 1 0 1 67 0 0 1 0 6 11 2 2 28 11 10 10 0 22 3 13 0 matrix 2 V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 1 0 66 0 10 9 5 0 3 7 0 0 2 43 0 0 14 2 32 0 1 5 3 0 3 11 32 4 0 19 10 1 7 13 2 0 4 0 0 0 0 0 0 0 0 0 0 0 5 11 6 4 18 0 15 18 22 4 1 0 6 2 40 3 3 12 0 1 24 1 14 0 7 4 5 0 8 32 49 2 0 0 0 0 8 0 0 0 0 0 0 0 0 0 0 0 9 19 23 2 42 8 2 0 0 0 3 1 10 1 16 3 3 3 73 0 0 2 0 0 11 2 4 0 20 15 10 15 0 10 7 17 __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Mantel test!
I'm not Gavin, but since you posted this to the R-help list I assume you are open to other respondents. Since you are asking Gavin, I assume you are using mantel() from the vegan package (it's a good idea to specify, as there are other possibilities). The help for that function explains the permutation test briefly, and gives a citation for more information. As for why you are getting p=1, I'd look into what hypothesis you are actually testing - quite possibly the null hypothesis that r = 0. The mantel() function in the ecodist package offers three different hypothesis tests: r = 0 r = 0 r == 0 Sarah On Tue, Mar 3, 2009 at 11:24 AM, Jason Sexton sexton...@gmail.com wrote: Dear Gavin, What is the interpretation of a simulated p-value of 1 in a mantel test and how is the p-value derived? When we run two highly negatively correlated matrices we get this result: r: -1, simulated p value: 1 I would expect the high p-value of 1 to mean not significant. Could you clarify? Thank you, Jason -- Sarah Goslee http://www.functionaldiversity.org __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Comparing two matrices
Hi, For the Mantel test, your matrices do not need to be Euclidean, but they do need to be symmetric. You could write a similar test that does not require a symmetric matrix, though. Sarah On Tue, Mar 3, 2009 at 11:40 AM, Koen Hufkens koen.hufk...@ua.ac.be wrote: Hi List, I would like to compare two (confusion) matrices. I ended up with the ade4 mantel.rtest function as my best option. However it seems that my data is non-euclidean in nature (is.euclid gives a FALSE). I attached both matrices at the end of the document. Are there any tools to transform these matrices into an euclidean form so I can use mantel.rtest with confidence. Other suggestions are also welcome. Kind regards, Koen -- matrix 1 V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 1 0 60 5 0 9 3 3 0 18 1 1 2 30 0 12 0 2 45 1 0 5 4 1 3 8 32 0 1 19 5 4 0 27 1 3 4 0 0 3 0 67 1 3 0 0 0 25 5 12 3 13 26 0 12 27 0 3 2 2 6 1 72 1 0 4 0 11 0 0 10 1 7 3 2 9 0 45 40 0 0 0 0 0 8 13 3 0 0 0 1 0 0 10 0 72 9 24 26 34 0 6 2 0 2 0 2 4 10 2 21 1 0 1 67 0 0 1 0 6 11 2 2 28 11 10 10 0 22 3 13 0 matrix 2 V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 1 0 66 0 10 9 5 0 3 7 0 0 2 43 0 0 14 2 32 0 1 5 3 0 3 11 32 4 0 19 10 1 7 13 2 0 4 0 0 0 0 0 0 0 0 0 0 0 5 11 6 4 18 0 15 18 22 4 1 0 6 2 40 3 3 12 0 1 24 1 14 0 7 4 5 0 8 32 49 2 0 0 0 0 8 0 0 0 0 0 0 0 0 0 0 0 9 19 23 2 42 8 2 0 0 0 3 1 10 1 16 3 3 3 73 0 0 2 0 0 11 2 4 0 20 15 10 15 0 10 7 17 -- Sarah Goslee http://www.functionaldiversity.org __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] modifying a built in function from the stats package (fixing arima)
Dear members of the list, I'm a beginner in R and I'm having some trouble with: Error in optim(init[mask], armafn, method = BFGS, hessian = TRUE, control = optim.control, : non-finite finite-difference value [8] when running arima. I've seen that some people have come accross the same problem: https://stat.ethz.ch/pipermail/r-help/2008-August/169660.html So I'd like to modify the code of arima to change the optimization function with another one that handles these problems automatically , however I don't find the way to do it and http://tolstoy.newcastle.edu.au/R/e6/help/09/01/2476.html points out a way that doesn't work for me: * If I type edit(arima) and I modify it, changes are not saved, * If I copy the code and I save it like a different function, I get the hard error: Error in Delta %+% c(1, -1) : object R_TSconv not found Anybody can give me a hint? I miss matlab's easy way of doing this (edit function.m). Thanks in advance MarC (AleaSoft) [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Inefficiency of SAS Programming
Ajay ohri wrote: why didnt you call it procunivariate if that was exactly what you wanted to do . Why would I ever do that? describe is a improvement on univariate, saving an estimated 87 +/- 87 trees per year in paper by printing what you need in much less space and concentrating on real descriptive statistics. Frank On Tue, Mar 3, 2009 at 7:11 PM, Frank E Harrell Jr f.harr...@vanderbilt.edu mailto:f.harr...@vanderbilt.edu wrote: Ajay ohri wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. It basically is meant only for SAS users who want to learn R , without upsetting the schedule of the corporate users. Simply put , it is a wrapper on SAS language using the function command...ie procunivariate function in Anne package would call the summary function and so on... Go ahead and add to the confusion. You've already created some by using summary for procunivariate. I created the describe function in the Hmisc package to replace univariate. Frank Regards, Ajay www.decisionstats.com http://www.decisionstats.com On Tue, Mar 3, 2009 at 9:20 AM, Greg Snow greg.s...@imail.org mailto:greg.s...@imail.org wrote: This does not really address my point. Yes, if the few nerds who want to do funny stuff are the ones making the purchase, then there is a good chance (but still not guaranteed) that they will get IML, but do all companies that buy SAS actually think about that, or do they just see the extra price (no matter how low), or not even think to look at that piece because the person making the purchase does not really the funny things you can do with it. If you want your SAS code to be able to be run by anyone with SAS, you cannot assume that they have IML. If you want your R code to be run by anyone, you cannot make your code dependent on packages/tools that are not available for all platforms. -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare greg.s...@imail.org mailto:greg.s...@imail.org 801.408.8111 -Original Message- From: Gerard M. Keogh [mailto:gmke...@justice.ie mailto:gmke...@justice.ie] Sent: Monday, March 02, 2009 3:22 AM To: Greg Snow Cc: Frank E Harrell Jr; R list; r-help-boun...@r-project.org mailto:r-help-boun...@r-project.org Subject: Re: [R] Inefficiency of SAS Programming Yes Greg, but if you're buying SAS they'll throw in IML pretty cheaply - SAS think it's only for a few nerds out there who wan to do funny stuff. G Greg Snow greg.s...@imail. org To Sent by: Gerard M. Keogh r-help-boun...@r- gmke...@justice.ie mailto:gmke...@justice.ie, Frank E project.org http://project.org Harrell Jr f.harr...@vanderbilt.edu mailto:f.harr...@vanderbilt.edu cc 27/02/2009 19:05 r-help-boun...@r-project.org mailto:r-help-boun...@r-project.org r-help-boun...@r-project.org mailto:r-help-boun...@r-project.org, R list r-h...@stat.math.ethz.ch mailto:r-h...@stat.math.ethz.ch Subject Re: [R] Inefficiency of SAS Programming But SAS/IML is not part of base SAS, it costs extra, so there is a good chance that a user that has SAS will not be able to run code that uses SAS/IML. I have known of SAS programmers who know IML well that still write matrix/vector tools using macros or proc transpose so
Re: [R] locfit smoothing question (package maintainer not reachable)
Dear all, I just realized that using family=qgauss restores normal-looking confidence bands... I read that using family=gaussian rather than family=qgauss fixes the dispersion parameter at 1, but without knowing the theory behind the code, I dont understand why there is such a difference between the two. If there is a simple explanation or recommendation, I am eager to hear it. Thanks, Suresh On Tue, 03 Mar 2009 16:56:43 +0100, Suresh Krishna madzient...@gmail.com wrote: Dear list members, I am trying to understand this output from the smoothing package locfit (1.5-4, running on R 2.8.1 on Windows Vista 64 bit). # sample code x-1:100 y-rnorm(100) fit-locfit(y~x,family=gaussian) #default parameters are fine plot(fit,band=global) #plot seems reasonable, confidence bands use a global estimate of variance y-1000*rnorm(100) fit-locfit(y~x,family=gaussian) plot(fit,band=global) #aren't these confidence bands too small ? am i using this function wrongly ? Using band=local gives results that seem to make more sense. Could someone offer me some guidance ? Thanks, Suresh ps. The package maintainer, Catherine Loader, is no longer reachable at her Auckland address. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] preparing data for barplot()
What is the best way to produce a barplot from my data? I would like the barplot to show each person with the values stacked val1+val2+val3, so there is one bar for each person When I use barplot(data.matrix(realdata)), it shows one bar for each value instead. To post here, I created an artificical data set, but it works fine. fakedata - as.data.frame(list(LETTERS[1:3])) colnames(fakedata) - 'people' fakedata['val1'] = abs(rnorm(3)) fakedata['val2'] = abs(rnorm(3)) fakedata['val3'] = abs(rnorm(3)) barplot(data.matrix(fakedata)) At a glance there is no substantial difference is between my fake data and my real data. realdata person val1 val2 val3 1 A 221 71 175 2 B 222 85 147 mydata people val1 val2 val3 1 A 0.75526748 0.445386 0.09186245 2 B 0.06107566 2.008815 2.50269410 3 C 0.47171611 0.592037 0.57612168 However data.matrix(realdata) person val1 val2 val3 1 NA 221 71 175 2 NA 222 85 147 Warning messages: 1: NAs introduced by coercion 2: NAs introduced by coercion So then I converted 'person' from a list to factors, which removed the coercision error, but barplot() still shows each bar as value instead of a person. My serialized() data subset is here (look at bottom half where there are no line numbers) http://pastebin.com/m6d1e1d79 Thanks in advance, Andrew __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] ggplot2: annotated horizontal lines
Hello, I'm using geom_hline to add a minimum line to my plot (representing the best solution found so far by a search algorithm). I'd like to annotate this line with it's numerical value to save trying to read it off the graph, but I can't see a clear way to do this - any ideas? (Alternatively, if this is against the spirit of the grammar of graphics, is there a better way to represent the information?) Cheers, dave -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Fit indexes in SEM with categorical data + ML estimation
Hey, Dorothee. SEM does not use the S-B chi squared index. If you want to use it, however, I have some methods written for the Satorra-Bentler Chi Square test in the sem.additions package over at r-forge. To install it directly within R type install.packages(sem.additions,repos=http://R-Forge.R-project.org;) For more details, see http://r-forge.r-project.org/projects/sem-additions/ Dorothee wrote: Hello, It has been found that SEM analysis using polychoric correlations + maximum likelihood estimator produces incorrect test statistics and standard errors (e.g., Flora, D. B., Curran, P. J. (2004). An Empirical Evaluation of Alternative Methods of Estimation for Confirmatory Factor Analysis With Ordinal Data. Psychological Methods, 9(4), 466-491). Standard errors can be dealt with by using bootstrap estimations. But, I was wondering how the fit indexes in the SEM package were estimated when the observed variables are categorical (ordinal or/and binary) and when one is using polychoric correlation (hetcor in polycor package) in combination with ML. Does SEM use the Satorra-Bentler rescaled Chi-square for instance? I look in previous posts but couldn't find what i am looking for (hope i looked well enough!). Cheers, Dorothee -- View this message in context: http://www.nabble.com/Fit-indexes-in-SEM-with-categorical-data-%2B-ML-estimation-tp21782611p22314468.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] ggplot2: annotated horizontal lines
What's wrong with geom_text? my.value = 0.65 qplot(1,1)+geom_hline(v=0)+ geom_text(mapping=aes(x=1,y=0),label=paste(my.value),vjust=-1) baptiste On 3 Mar 2009, at 18:10, Dave Murray-Rust wrote: Hello, I'm using geom_hline to add a minimum line to my plot (representing the best solution found so far by a search algorithm). I'd like to annotate this line with it's numerical value to save trying to read it off the graph, but I can't see a clear way to do this - any ideas? (Alternatively, if this is against the spirit of the grammar of graphics, is there a better way to represent the information?) Cheers, dave -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. _ Baptiste Auguié School of Physics University of Exeter Stocker Road, Exeter, Devon, EX4 4QL, UK Phone: +44 1392 264187 http://newton.ex.ac.uk/research/emag __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] nlme: problem with fitting logistic function
Dear colegues I am trying to analyze growth data on mice. To do this I attempted to fit logistic curve using nlme package. However, the dataset I use is large (in total ca 20 000 measures on ca 3 000 individuals) with relatively complicated structure (several explanatory variables with interactions + random effect where individual offspring are nested within particular litters are nested within particular parental pairs). Although I had no problems to fit models, where the complexity of random effect was reduce (which is conceptually incorrect), the fitting procedure of the full model did not reach the 1 iteration after several hours. Do you have any idea how to solve this problem? Thank you very much for your advice Best regards Jakub Kreisinger __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] locfit smoothing question (package maintainer not reachable)
I think you should read (or re-read) the locfit help page and *also* the links from that page to the help pages for locfit.raw and rv. I would have thought that since family= is not an argument to locfit per se, but rather is documented in locfit.raw that you have yet done so, but perhaps not? -- David Winsemis On Mar 3, 2009, at 12:39 PM, Suresh Krishna wrote: Dear all, I just realized that using family=qgauss restores normal-looking confidence bands... I read that using family=gaussian rather than family=qgauss fixes the dispersion parameter at 1, but without knowing the theory behind the code, I dont understand why there is such a difference between the two. If there is a simple explanation or recommendation, I am eager to hear it. Thanks, Suresh On Tue, 03 Mar 2009 16:56:43 +0100, Suresh Krishna madzient...@gmail.com wrote: Dear list members, I am trying to understand this output from the smoothing package locfit (1.5-4, running on R 2.8.1 on Windows Vista 64 bit). # sample code x-1:100 y-rnorm(100) fit-locfit(y~x,family=gaussian) #default parameters are fine plot(fit,band=global) #plot seems reasonable, confidence bands use a global estimate of variance y-1000*rnorm(100) fit-locfit(y~x,family=gaussian) plot(fit,band=global) #aren't these confidence bands too small ? am i using this function wrongly ? Using band=local gives results that seem to make more sense. Could someone offer me some guidance ? Thanks, Suresh ps. The package maintainer, Catherine Loader, is no longer reachable at her Auckland address. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] scatter plot question
Hi R Users, I have a dataframe like this: id x rho A 1 0.1 B 20 0.5 C 2 0.9 ... I want to do a scatter plot of x versus rho but for each point on the scatter plot I want the corresponding entry for id instead of points. In STATA I can do so by twoway (scatter x rho, mlabel(id)) How can I do the same in R? I am sure there is some simple way to do this. Dipankar [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] locfit smoothing question (package maintainer not reachable)
From: Suresh Krishna [...] ps. The package maintainer, Catherine Loader, is no longer reachable at her Auckland address. For the record, I'm the package maintainer for locfit, and I have not exactly vanished (yet). Please see the package description. That said, it doesn't mean I know all the details about the code. I just do enough to keep the package on CRAN. Best, Andy Notice: This e-mail message, together with any attachme...{{dropped:12}} __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] save the layout using igraph
Hi R users, I am using built-in functions in igraph package to draw networks . I need to compare several network with exactly the same structure but with edge hightlighted differently. I am wondering if there is a way to save the layout so that every graph will look the same as each other except for the colors of edges. Or is there any parameter I can set for this purpose?? Thanks in advance, Shukai -- View this message in context: http://www.nabble.com/save-the-layout-using-igraph-tp22316155p22316155.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] rgl persp3d bounding box color problem.
rgl_0.83-3 did not help Duncan Murdoch wrote: On 3/2/2009 2:48 PM, chaogai wrote: Oops, sorry R version 2.8.1 (2008-12-22) i686-pc-linux-gnu locale: LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=C;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=C attached base packages: [1] tcltk stats graphics grDevices utils datasets methods [8] base other attached packages: [1] rgl_0.81svGUI_0.9-44svSocket_0.9-43 svMisc_0.9-46 0.81 isn't the current version of rgl on CRAN. Please try that one, or even better, the development version on R-forge. Duncan Murdoch Duncan Murdoch wrote: On 3/2/2009 2:00 PM, chaogai wrote: Whatever it is, it is also happening on my Acer, Suse 11.1, 32 bits, self build R. Versions of R and rgl? Duncan Murdoch Kees Bo Zhou wrote: Thanks Duncan. I haven't got a chance to try it in windows yet. But I won't be surprised if it's my driver's fault. I had OpenGL problems on this dell laptop before. Date: Sun, 1 Mar 2009 18:29:20 -0500 From: murd...@stats.uwo.ca To: bozhou1...@hotmail.com CC: remkoduur...@gmail.com; r-help@r-project.org Subject: Re: [R] rgl persp3d bounding box color problem. On 01/03/2009 5:54 PM, Bo Zhou wrote: site down.. here is the code library(rgl) plot1: persp3d(x=1:11,y=1:210,z=matrix(rep(1,11*210),11,210),col=matrix(rep(#FF,11*210),11,210)) plot2: persp3d(x=1:11,y=1:10,z=matrix(rep(1,11*10),11,10),col=matrix(rep(#FF,11*10),11,10)) Works fine on Mac OS 10.5 too. Looks like a bug in your OpenGL driver... Duncan Murdoch Thanks! Bo Date: Mon, 2 Mar 2009 09:28:31 +1100 Subject: Re: [R] rgl persp3d bounding box color problem. From: remkoduur...@gmail.com To: bozhou1...@hotmail.com I would not mind trying, but the link no longer works... Remko - Remko Duursma Post-Doctoral Fellow Centre for Plant and Food Science University of Western Sydney Hawkesbury Campus Richmond NSW 2753 Dept of Biological Science Macquarie University North Ryde NSW 2109 Australia Mobile: +61 (0)422 096908 On Mon, Mar 2, 2009 at 8:16 AM, Bo Zhou bozhou1...@hotmail.com wrote: Thanks, Duncan. What's your platform? Could this be an issues related to my platform? I'm on Ubuntu 8.10 64bit using cran's R package. I built rgl package on my machine using install.packages(). I'm going to give it a try in windows. I would be very grateful if anyone using similar setup could give it a try. Thanks, Bo Date: Sun, 1 Mar 2009 07:03:10 -0500 From: murd...@stats.uwo.ca To: bozhou1...@hotmail.com CC: r-help@r-project.org Subject: Re: [R] rgl persp3d bounding box color problem. On 28/02/2009 4:20 PM, Bo Zhou wrote: Hi guys, I hit on a problem when I use rgl. Could you try to run the code here in this link and see why the first persp3d gives a red bounding box and the second shows black? They both show as black on my system. Duncan Murdoch http://rafb.net/p/g1i7ur33.html (sorry for not pasting the code here directly but my previous email got filtered by this list so I suspect my code looks like spam to the spam filter) I'm expecting black color to be the right result but the code is virtually same except different data size. Looks like a mem bug to me? BTW you will need to install the rgl package Thanks, Bo _ [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. _ [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. _ [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. _ Hotmail® is up to 70% faster. Now good news travels really fast. [[alternative HTML version deleted]]
Re: [R] Inefficiency of SAS Programming
On 3/03/2009, at 5:58 PM, Ajay ohri wrote: for an inefficient language , it sure has dominated the predictive analytics world for 3 plus decades. I referred once to intellectual jealousy between newton and liebnitz. i am going ahead and creating the R package called Anne. It basically is meant only for SAS users who want to learn R , without upsetting the schedule of the corporate users. Simply put , it is a wrapper on SAS language using the function command...ie procunivariate function in Anne package would call the summary function and so on... Reminds me of fortune(38). cheers, Rolf Turner ## Attention:\ This e-mail message is privileged and confid...{{dropped:9}} __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] save the layout using igraph
Shukai, the layout functions return a matrix that you can supply as the 'layout' argument of the plotting functions, e.g. g1 - erdos.renyi.game(50,2/50) lay - layout.fruchterman.reingold(g1) g2 - delete.edges(g1, sample(ecount(g1), 30)-1) plot(g1, layout=lay, vertex.size=10) plot(g2, layout=lay, vertex.size=10) Best, Gabor On Tue, Mar 3, 2009 at 8:37 PM, kevinchang shu...@seas.upenn.edu wrote: Hi R users, I am using built-in functions in igraph package to draw networks . I need to compare several network with exactly the same structure but with edge hightlighted differently. I am wondering if there is a way to save the layout so that every graph will look the same as each other except for the colors of edges. Or is there any parameter I can set for this purpose?? Thanks in advance, Shukai -- View this message in context: http://www.nabble.com/save-the-layout-using-igraph-tp22316155p22316155.html Sent from the R help mailing list archive at Nabble.com. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Gabor Csardi gabor.csa...@unil.ch UNIL DGM __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] ggplot2: annotated horizontal lines
On 3 Mar 2009, at 18:41, baptiste auguie wrote: What's wrong with geom_text? my.value = 0.65 qplot(1,1)+geom_hline(v=0)+ geom_text(mapping=aes(x=1,y=0),label=paste(my.value),vjust=-1) Well, firstly I hadn't thought to use it. Ooops. Secondly, I can't make it just do a single value - it seems to want a value for every point in the dataset, e.g.: qplot( wt, mpg, data=mtcars ) + geom_text(mapping=aes(x=1,y=0,label=paste(0.5),vjust=-1)) Error in data.frame(..., check.names = FALSE) : arguments imply differing number of rows: 1, 32 Cheers, dave baptiste On 3 Mar 2009, at 18:10, Dave Murray-Rust wrote: Hello, I'm using geom_hline to add a minimum line to my plot (representing the best solution found so far by a search algorithm). I'd like to annotate this line with it's numerical value to save trying to read it off the graph, but I can't see a clear way to do this - any ideas? (Alternatively, if this is against the spirit of the grammar of graphics, is there a better way to represent the information?) Cheers, dave -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. _ Baptiste Auguié School of Physics University of Exeter Stocker Road, Exeter, Devon, EX4 4QL, UK Phone: +44 1392 264187 http://newton.ex.ac.uk/research/emag __ -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] par and a substitute for mtext to write one time a title per page
RE: par and a substitute for mtext to write one time a title per page Hi, Q1. Is there way I can set mfg =c(2,1) which for me could have meant 2 pages and no split on the columns per page; and mfrow =c(4,1) which for me it means 4 graphs per page in the par function? So if I have a pdf/png etc file I will output two pages with 4 graphs for each, or modifiable depending on the case of data? I obtain the following: Error in par(mfg = c(2, 1), mfrow = c(4, 1)) : parameter i in mfg is out of range steps: 1. open outstream: pdf(file=some_address_and_file.pdf) 2. do the corresponding loop to prepare the graphs for specific data 3. plot 2 or more pages of graphs with 4 graphs per page 4. close the outstream. Q2: with mtext function I can write in the outter margin including a title as a header. Is there another way to write a title one time per page in these multiple graphs, so I can save the graphs space per page? Thank you in advance, Aldi -- __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] repeated measures anova, sphericity, epsilon, etc
I have 3 questions (below). Background: I am teaching an introductory statistics course in which we are covering (among other things) repeated measures anova. This time around teaching it, we are using R for all of our computations. We are starting by covering the univariate approach to repeated measures anova. Doing a basic repeated measures anova (univariate approach) using aov() seems straightforward (e.g.: + myModel-aov(myDV~myFactor+Error(Subjects/myFactor),data=myData) + summary(myModel) Where I am currently stuck is how best to deal with the issue of the assumption of homogeneity of treatment differences (in other words, the sphericity assumption) - both how to test it in R and how to compute corrected df for the F-test if the assumption is violated. Back when I taught this course using SPSS it was relatively straightforward - we would look at Mauchly's test of sphericity - if it was significant, then we would use one of the corrected F-tests (e.g. Greenhouse-Geisser or Huynh-Feldt) that were spat out automagically by SPSS. I gather from searching the r-help archives, searching google, and searching through various books on R, that the only way of using mauchly.test() in R is on a multivariate model object (e.g. mauchly.test cannot handle an aov() object). Question 1: how do you (if you do so), test for sphericity in a repeated measures anova using R, when using aov()? (or do you test the sphericity assumption using a different method)? Question 2: Can someone point me to an example (on the web, in a book, wherever) showing how to perform a repeated measures anova using the multivariate approach in R? Question 3: Are there any existing R functions for calculating adjusted df for Greenhouse-Geisser, Huynh-Feldt (or calculating epsilon), or is it up to me to write my own function? Thanks in advance for any suggestions, -- Paul L. Gribble, Ph.D. Associate Professor Dept. Psychology The University of Western Ontario London, Ontario Canada N6A 5C2 Tel. +1 519 661 2111 x82237 Fax. +1 519 661 3961 pgrib...@uwo.ca http://gribblelab.org [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] locfit smoothing question (package maintainer not reachable)
David Winsemis wrote: I think you should read (or re-read) the locfit help page and *also* the links from that page to the help pages for locfit.raw and rv. I would have thought that since family= is not an argument to locfit per se, but rather is documented in locfit.raw that you have yet done so, but perhaps not? I did read the help pages for locfit.raw, and found: Local likelihood family; gaussian; binomial; poisson; gamma and geom. Density and rate estimation families are dens, rate and hazard (hazard rate). If the family is preceded by a 'q' (for example, family=qbinomial), quasi-likelihood variance estimates are used. Otherwise, the residual variance (rv) is fixed at 1. The default family is qgauss if a response y is provided; density if no response is provided. However, since the fake data were generated from a known gaussian distribution, I did not imagine that using family=gaussian would lead to such wildly different results. This is what I was hoping to understand, without having to struggle with Catherine's Loader book in order to understand the above paragraph deeply enough that this behavior makes sense. Thanks again, Suresh __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] modifying a built in function from the stats package (fixing arima)
Hello, I do not think that is the way to go. If you believe that your algorithm is better than the existing one, talk to the author of the package and discuss the improvement. The whole community will benefit. If you want to tune the existing function and tailor it to your needs, you have several ways to go, among them: 1) Copy the existing function into a new file, edit it and load it via source. 2) Download the source package and modify it for your own purposes. Best regards, Carlos J. Gil Bellosta http://www.datanalytics.com On Tue, 2009-03-03 at 18:20 +0100, Marc Vinyes wrote: Dear members of the list, I'm a beginner in R and I'm having some trouble with: Error in optim(init[mask], armafn, method = BFGS, hessian = TRUE, control = optim.control, : non-finite finite-difference value [8] when running arima. I've seen that some people have come accross the same problem: https://stat.ethz.ch/pipermail/r-help/2008-August/169660.html So I'd like to modify the code of arima to change the optimization function with another one that handles these problems automatically , however I don't find the way to do it and http://tolstoy.newcastle.edu.au/R/e6/help/09/01/2476.html points out a way that doesn't work for me: * If I type edit(arima) and I modify it, changes are not saved, * If I copy the code and I save it like a different function, I get the hard error: Error in Delta %+% c(1, -1) : object R_TSconv not found Anybody can give me a hint? I miss matlab's easy way of doing this (edit function.m). Thanks in advance MarC (AleaSoft) [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] spatial markov chain methods
On 4/03/2009, at 4:36 AM, Torsten Lange wrote: Hello, can any one point me to R-packages (if available) which include spatial Markov Chain methods? My second question is more general but hopefully not OT: Currently we are using the software TPROGS, which let people simulate property distributions in space by some Markov Chain approaches. We face some problems due to the lack of information between distances of samples along borehole path and among boreholes. Is there any way to trick it like for instants in variogram modeling were one might use nested variograms...? The spatstat package has facilities for simulating (non-Poisson) spatial point processes vi MCMC (Metropolis-Hastings) methods. But it's probably not point patterns that you're after. cheers, Rolf Turner ## Attention:\ This e-mail message is privileged and confid...{{dropped:9}} __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] xlsReadWrite package repository for Ubuntu
2009/3/2 Prof Brian Ripley rip...@stats.ox.ac.uk: The CRAN Windows ones. It is a windows-only package, see http://cran.r-project.org/web/packages/xlsReadWrite/index.html (incidentally to you: it seems no longer maintained and does not build under R-devel on Windows, see http://cran.r-project.org/bin/windows/contrib/2.9/check/xlsReadWrite-check.log The maintainer has not responded to requests for a fix.) I didn't get a request (my email was 'gcha...@gmail instead of gcha...@gmail.com). I'll look into the issue this weekend. However there might be a problem because the package contains binary code and maybe xlsReadWrite has to be put on a place other than CRAN (I'll ask Rcore about this in a separate email). In any case I absolutely intend to maintain the package also in future. AFAIK it works well (apart from a small rowname bug which surfaced recently ;-) and provides a pragmatic alternative to the OSS R packages gdata and RODBC) 2009/3/3 Marc Schwartz marc_schwa...@comcast.net: You perhaps missed the key point in Uwe's response, which is that the package is only available under Windows, as it depends upon Windows specific functionality (MS Office API via a third party library which is available for Windows only) to natively read and write Excel files. It's correct that xlsReadWrite is currently only available for Windows. However it doesn't use the MS Office API but works directly with the file format (BIFF8), i.e. Excel is _not_ required. True is, that it uses a third party library (Flexcel, TMSSoftware). This library supports FreePascal which means that theoretically xlsReadWrite could be compiled for Linux/Mac (tried once, didn't succeed, will try again later (now also using Macs and Debian/Ubuntu)). -- Regards, Hans-Peter __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] xlsReadWrite package repository for Ubuntu
On 4/03/2009, at 2:38 AM, Marc Schwartz wrote: You perhaps missed the key point in Uwe's response, which is that the package is only available under Windows, as it depends upon Windows specific functionality (MS Office API via a third party library which is available for Windows only) to natively read and write Excel files. Thus, there is no package version available for Linux, or OSX for that matter. If you need to read Excel files under Linux, you could look at the read.xls() function in the 'gdata' CRAN package. This package requires that Perl be installed, as it calls a Perl routine (xls2csv) for converting the Excel file to a CSV file, which can then be read into R. If you need to write Excel files under Linux, you can use a Perl routine that I had posted back in 2007: https://stat.ethz.ch/pipermail/r-help/2007-July/135968.html and have updated since then to handle Unicode issues. I am attaching a 2k text file here with the updated routine. For those needing to write *.xls files as well as read them, the write.xls function in the package dataframes2xls might be useful. I have no actual experience in using it, but. cheers, Rolf Turner ## Attention:\ This e-mail message is privileged and confid...{{dropped:9}} __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] ggplot2: annotated horizontal lines
I see your problem (although label should be outside the mapping in your case, i think). A possible workaround is to provide some dummy data, as the default NULL doesn't seem to work, qplot( wt, mpg, data=mtcars ) + geom_text(data=data.frame(x=0,y=0),mapping=aes(x=1,y=0),label=test) I'm sure Hadley will come up with a better explanation. Best, baptiste On 3 Mar 2009, at 19:51, Dave Murray-Rust wrote: On 3 Mar 2009, at 18:41, baptiste auguie wrote: What's wrong with geom_text? my.value = 0.65 qplot(1,1)+geom_hline(v=0)+ geom_text(mapping=aes(x=1,y=0),label=paste(my.value),vjust=-1) Well, firstly I hadn't thought to use it. Ooops. Secondly, I can't make it just do a single value - it seems to want a value for every point in the dataset, e.g.: qplot( wt, mpg, data=mtcars ) + geom_text(mapping=aes(x=1,y=0,label=paste(0.5),vjust=-1)) Error in data.frame(..., check.names = FALSE) : arguments imply differing number of rows: 1, 32 Cheers, dave baptiste On 3 Mar 2009, at 18:10, Dave Murray-Rust wrote: Hello, I'm using geom_hline to add a minimum line to my plot (representing the best solution found so far by a search algorithm). I'd like to annotate this line with it's numerical value to save trying to read it off the graph, but I can't see a clear way to do this - any ideas? (Alternatively, if this is against the spirit of the grammar of graphics, is there a better way to represent the information?) Cheers, dave -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. _ Baptiste Auguié School of Physics University of Exeter Stocker Road, Exeter, Devon, EX4 4QL, UK Phone: +44 1392 264187 http://newton.ex.ac.uk/research/emag __ -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. _ Baptiste Auguié School of Physics University of Exeter Stocker Road, Exeter, Devon, EX4 4QL, UK Phone: +44 1392 264187 http://newton.ex.ac.uk/research/emag __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] repeated measures anova, sphericity, epsilon, etc
Paul Gribble wrote: I have 3 questions (below). Background: I am teaching an introductory statistics course in which we are covering (among other things) repeated measures anova. This time around teaching it, we are using R for all of our computations. We are starting by covering the univariate approach to repeated measures anova. Doing a basic repeated measures anova (univariate approach) using aov() seems straightforward (e.g.: + myModel-aov(myDV~myFactor+Error(Subjects/myFactor),data=myData) + summary(myModel) Where I am currently stuck is how best to deal with the issue of the assumption of homogeneity of treatment differences (in other words, the sphericity assumption) - both how to test it in R and how to compute corrected df for the F-test if the assumption is violated. Back when I taught this course using SPSS it was relatively straightforward - we would look at Mauchly's test of sphericity - if it was significant, then we would use one of the corrected F-tests (e.g. Greenhouse-Geisser or Huynh-Feldt) that were spat out automagically by SPSS. I gather from searching the r-help archives, searching google, and searching through various books on R, that the only way of using mauchly.test() in R is on a multivariate model object (e.g. mauchly.test cannot handle an aov() object). Question 1: how do you (if you do so), test for sphericity in a repeated measures anova using R, when using aov()? (or do you test the sphericity assumption using a different method)? Question 2: Can someone point me to an example (on the web, in a book, wherever) showing how to perform a repeated measures anova using the multivariate approach in R? Question 3: Are there any existing R functions for calculating adjusted df for Greenhouse-Geisser, Huynh-Feldt (or calculating epsilon), or is it up to me to write my own function? Thanks in advance for any suggestions, Have a look at http://cran.r-project.org/doc/Rnews/Rnews_2007-2.pdf Last time this came up, John Fox also pointed to some of his stuff, see http://finzi.psych.upenn.edu/R/Rhelp08/archive/151282.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 ~~ - (p.dalga...@biostat.ku.dk) FAX: (+45) 35327907 __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] locfit smoothing question (package maintainer not reachable)
That is what I thought to be the critical paragraph. The variance is assumed to be = 1 when you use family=gaussian rather than the default of family=qgauss. You give it a vector, 1000*rnorm(100), that ranges widely and a small (relative) variance is assumed and so the confidence intervals are plotted as very narrow. This does not seem surprising given the functions documented design. I have the book and do not think I even need to pull it off the shelf since the help pages appear fully informative in this instance. I get an rv of 1 with the gaussian option and an rv of nearly 1000 when the default is used. -- David Winsemius On Mar 3, 2009, at 3:26 PM, Suresh Krishna wrote: David Winsemis wrote: I think you should read (or re-read) the locfit help page and *also* the links from that page to the help pages for locfit.raw and rv. I would have thought that since family= is not an argument to locfit per se, but rather is documented in locfit.raw that you have yet done so, but perhaps not? I did read the help pages for locfit.raw, and found: Local likelihood family; gaussian; binomial; poisson; gamma and geom. Density and rate estimation families are dens, rate and hazard (hazard rate). If the family is preceded by a 'q' (for example, family=qbinomial), quasi-likelihood variance estimates are used. Otherwise, the residual variance (rv) is fixed at 1. The default family is qgauss if a response y is provided; density if no response is provided. However, since the fake data were generated from a known gaussian distribution, I did not imagine that using family=gaussian would lead to such wildly different results. This is what I was hoping to understand, without having to struggle with Catherine's Loader book in order to understand the above paragraph deeply enough that this behavior makes sense. Thanks again, Suresh __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] sm.density.compare
I am running the sm.density.compare function amd I am getting the following error: my code is sm.density.compare(LBSTRESN,COHORT,xlab=Units = umol/Lsubset = LBTEST==Creatinine) Error in if (from == to) rep.int(from, length.out) else as.vector(c(from, : missing value where TRUE/FALSE needed I do not understand the error and I have had no help when searching the WEB or the R documentation. I have tried the sample program in the R doc:y - rnorm(100) g - rep(1:2, rep(50,2)) sm.density.compare(y, g, model=equal) and it works fine. The only difference with my data is that I have some missing values that show as NA. I have checked that theseare numbered values by using the is.nan function. Any ideas? Thank You Jim L. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] spatial markov chain methods
Rolf Turner schrieb: On 4/03/2009, at 4:36 AM, Torsten Lange wrote: Hello, can any one point me to R-packages (if available) which include spatial Markov Chain methods? My second question is more general but hopefully not OT: Currently we are using the software TPROGS, which let people simulate property distributions in space by some Markov Chain approaches. We face some problems due to the lack of information between distances of samples along borehole path and among boreholes. Is there any way to trick it like for instants in variogram modeling were one might use nested variograms...? The spatstat package has facilities for simulating (non-Poisson) spatial point processes vi MCMC (Metropolis-Hastings) methods. But it's probably not point patterns that you're after. cheers, Rolf Turner Thanks Rolf! To be more specific I'm a novice to MC method. However, I don't look on point processes but on spatial auto-correlation of certain types of soils or sediments in 3D - parameterization for geoscientific applications like hydrogeology... We created the MC transiograms for soil classes in vertical and horizontal directions. Problems arise due to the observation gap between sampling density in vertical (very high along borehole) and horizontal direction (much lower because of distances among boreholes). Thank you for any hints, Torsten __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] repeated measures anova, sphericity, epsilon, etc
Have a look at http://cran.r-project.org/doc/Rnews/Rnews_2007-2.pdf Wow. I think my students would keel over. Anova() from the car package looks promising - I will check it out. Thanks On Tue, Mar 3, 2009 at 4:00 PM, Peter Dalgaard p.dalga...@biostat.ku.dkwrote: Paul Gribble wrote: I have 3 questions (below). Background: I am teaching an introductory statistics course in which we are covering (among other things) repeated measures anova. This time around teaching it, we are using R for all of our computations. We are starting by covering the univariate approach to repeated measures anova. Doing a basic repeated measures anova (univariate approach) using aov() seems straightforward (e.g.: + myModel-aov(myDV~myFactor+Error(Subjects/myFactor),data=myData) + summary(myModel) Where I am currently stuck is how best to deal with the issue of the assumption of homogeneity of treatment differences (in other words, the sphericity assumption) - both how to test it in R and how to compute corrected df for the F-test if the assumption is violated. Back when I taught this course using SPSS it was relatively straightforward - we would look at Mauchly's test of sphericity - if it was significant, then we would use one of the corrected F-tests (e.g. Greenhouse-Geisser or Huynh-Feldt) that were spat out automagically by SPSS. I gather from searching the r-help archives, searching google, and searching through various books on R, that the only way of using mauchly.test() in R is on a multivariate model object (e.g. mauchly.test cannot handle an aov() object). Question 1: how do you (if you do so), test for sphericity in a repeated measures anova using R, when using aov()? (or do you test the sphericity assumption using a different method)? Question 2: Can someone point me to an example (on the web, in a book, wherever) showing how to perform a repeated measures anova using the multivariate approach in R? Question 3: Are there any existing R functions for calculating adjusted df for Greenhouse-Geisser, Huynh-Feldt (or calculating epsilon), or is it up to me to write my own function? Thanks in advance for any suggestions, Have a look at http://cran.r-project.org/doc/Rnews/Rnews_2007-2.pdf Last time this came up, John Fox also pointed to some of his stuff, see http://finzi.psych.upenn.edu/R/Rhelp08/archive/151282.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 ~~ - (p.dalga...@biostat.ku.dk) FAX: (+45) 35327907 -- Paul L. Gribble, Ph.D. Associate Professor Dept. Psychology The University of Western Ontario London, Ontario Canada N6A 5C2 Tel. +1 519 661 2111 x82237 Fax. +1 519 661 3961 pgrib...@uwo.ca http://gribblelab.org [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] par and a substitute for mtext to write one time a title per page
Hmm, Reading some responses in the archives on mfg, the first one was simple the mfg does not control the pages. After removing mfg=c(2,1), and leaving in the program only mfrow=(4,1) although I am printing one after the other in a loop, pages are printed and do not overwrite the same page. So by defining mfrow only, one can extend to many pages with no problem. Q2. Here is a small test program: x-rnorm(1000) y-rnorm(1000) par(mfrow=c(2,1)) plot(x,y) mtext(This is my title,3,line=1) mtext(This is my title,3,line=2) mtext(This is my title,3,line=3) plot(x,y) Is there a function that can print a title in every one new page outside of the space designated for graphs, instead of me finding the first graph that starts the page and there using mtext as shown above? Thanks, Aldi Aldi Kraja wrote: RE: par and a substitute for mtext to write one time a title per page Hi, Q1. Is there way I can set mfg =c(2,1) which for me could have meant 2 pages and no split on the columns per page; and mfrow =c(4,1) which for me it means 4 graphs per page in the par function? So if I have a pdf/png etc file I will output two pages with 4 graphs for each, or modifiable depending on the case of data? I obtain the following: Error in par(mfg = c(2, 1), mfrow = c(4, 1)) : parameter i in mfg is out of range steps: 1. open outstream: pdf(file=some_address_and_file.pdf) 2. do the corresponding loop to prepare the graphs for specific data 3. plot 2 or more pages of graphs with 4 graphs per page 4. close the outstream. Q2: with mtext function I can write in the outter margin including a title as a header. Is there another way to write a title one time per page in these multiple graphs, so I can save the graphs space per page? Thank you in advance, Aldi -- __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] scatter plot question
Dipankar Basu basu.15 at gmail.com writes: Hi R Users, I have a dataframe like this: id x rho A 1 0.1 B 20 0.5 C 2 0.9 ... I want to do a scatter plot of x versus rho but for each point on the scatter plot I want the corresponding entry for id instead of points. test - data.frame(id=c(A, B, C), x=c(1, 20, 2), rho=c(0.1, 0.5, 0.9)) plot(rho~x, test, pch=as.character(id)) Mark Lyman __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] portable R editor
EditPad Pro is commercial, which makes it a nonchoice for recommending it to my students. I recently switched from tinn-R to Notepad++ with NppToR and am quite happy with it. tinn-R is quite good, but possibly the project is getting too ambitous now. It needs quite some fiddling with Rprofile.site to make it work. Thompson, David (MNR) wrote: Werner, Another alternative is EditPad Pro http://www.editpadpro.com/. DaveT. * Silviculture Data Analyst Ontario Forest Research Institute Ontario Ministry of Natural Resources david.john.thomp...@ontario.ca http://ontario.ca/ofri * -Original Message- From: Andrew Redd [mailto:ar...@stat.tamu.edu] Sent: March 2, 2009 10:29 PM To: Michael Bibo Cc: r-h...@stat.math.ethz.ch Subject: Re: [R] portable R editor Thanks for the plug on NppToR. Yes it is portable, but a few of the features don't work. I have it on my plan to have a launcher for the portable apps menu. Also the website for NppToR is now https://sourceforge.net/projects/npptor/ On my personal website I also keep, semi-up-to-date a portableapps.comcompatible launcher for R. http://www.stat.tamu.edu/~aredd/site/?q=node/2. There is a full installation and just the launcher. If you download just the launcher you can install the latest version of R into the bin directory and It will work. I think it is version 2.8.0 in the installer. I'll update that soon. Andrew Redd On Mon, Mar 2, 2009 at 8:13 PM, Michael Bibo michael_b...@health.qld.gov.au wrote: Werner Wernersen pensterfuzzer at yahoo.de writes: Hi, I have been dreaming about a complete R environment on my USB stick for a long time. Now I finally want to realize it but what I am missing is a good, portable editor for R which has tabs and syntax highlighting, can execute code, has bookmarks and a little project file management facility pretty much like Tinn-R has those. I like Tinn-R but it seems like there is only a very old version of Tinn-R which works standalone. Hi Werner, Three options: I have previously posted about using Emacs + ESS on a USB stick: http://finzi.psych.upenn.edu/R/Rhelp02/archive/107419.html Tinn-R will work portably. I have simply copied the installed Tinn-R folder from one machine to another on which I do not have administrator privileges and therefore cannot do a normal install. Another option is Notepad++ (http://notepad- plus.sourceforge.net/uk/site.htm). There is even a portable version available fom www.portableapps.com. Andrew Redd has created (and is continuing to develop) NppToR (http://www.stat.tamu.edu/~aredd/site/?q=node/37http://www.sta t.tamu.edu/%7Earedd/site/?q=node/37) which provides syntax highlighting, code folding and code passing to R. Notepad++ has a tabbed interface as well as add-ons including a file explorer, windows manager and multiclipboard manager. It also allows for the recording of macros, and the instructions found at this blog (http://aztecpassage.blogspot.com/2007/11/load-new-file-from-te mplate-in- notepad.htmlhttp://aztecpassage.blogspot.com/2007/11/load-new- file-from-template-in-%0Anotepad.html) use Notepad++'s built-in capacity to run external tools to create new files from templates. Michael Bibo Queensland Health __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.0.237 / Virus Database: 270.11.6/1980 - Release Date: 03/02/09 23:02:00 -- Erich Neuwirth, University of Vienna Faculty of Computer Science Computer Supported Didactics Working Group Visit our SunSITE at http://sunsite.univie.ac.at Phone: +43-1-4277-39464 Fax: +43-1-4277-39459 __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] detect outliers and high levarage points
Hi friends, How to detect outliers and high leverage points for GLM ? Could I use plot(model) (i) Residuals vs Fitted graph to detect the outliers ? (ii) Residuals vs Leverage graph to detect the high leverage points ? And then remove those points from the data and re-run the model ? [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] clust with hclustWard
Hello everyone I tried to use clust function using hclustWard clustering method The distance mesaured I have used is rf The data used is in the form of data frame (it is not a distance matrix) - sd.df The piece of code I have used is: c1 - clust(sd.df,k=6,method=hclustWard,distMethod=rf) But I get the following error message, Error in if (n 2) stop(must have n = 2 objects to cluster) : argument is of length zero Could any one help me with this? Also I tried to check the code of clust to find out where this error message appears in the code and I could not find it. Please let me know how can I make the clust work for hclustWard method. Thanks! Anirudh Kondaveeti [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Inference for R Spam
Dieter == Dieter Menne dieter.me...@menne-biomed.de writes: And, since my son asked me and I am basketball ignorant: Why are basketball scores mostly much too close to equality? The arguments (loose power when leading) might suggest that 2:0 might not be significant, but relevant. I tend to argue the other way round though, in medical statistics. Sports scores are not statistics, they are measurements (counts) of the number of times each team scores. There is no sampling and vanishingly small possibility of systematic error in the measurement. Mike -- Michael A. Miller mmill...@iupui.edu Department of Radiology, Indiana University School of Medicine __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] reading scanned graphs
I have a graph (a curve with a x- and y-axis) on paper. (The device from which I receive this is not able to output the data digitally unfortunately.) Is there a procedure by which I can scan the graph, save it into, e.g., a bitmap file and make it R to read this file into x and y coordinates? Thanks, Aad Termorshuizen _ Disclaimer: Dit e-mailbericht is uitsluitend bestemd voor de geadresseerde(n). Indien dit e-mailbericht niet voor u bestemd is, verzoeken wij u vriendelijk doch dringend, de afzender van het bericht op hoogte te brengen en alle informatie uit uw computer te verwijderen zonder acht te nemen van de inhoud. This E-mail has been scanned for viruses by the McAfee Appliance of Blgg B.V. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] reading scanned graphs
Check out http://www.datathief.org/ Remko - Remko Duursma Post-Doctoral Fellow Centre for Plant and Food Science University of Western Sydney Hawkesbury Campus Richmond NSW 2753 Dept of Biological Science Macquarie University North Ryde NSW 2109 Australia Mobile: +61 (0)422 096908 2009/3/4 Aad Termorshuizen aad.termorshui...@blgg.nl: I have a graph (a curve with a x- and y-axis) on paper. (The device from which I receive this is not able to output the data digitally unfortunately.) Is there a procedure by which I can scan the graph, save it into, e.g., a bitmap file and make it R to read this file into x and y coordinates? Thanks, Aad Termorshuizen _ Disclaimer: Dit e-mailbericht is uitsluitend bestemd voor de geadresseerde(n). Indien dit e-mailbericht niet voor u bestemd is, verzoeken wij u vriendelijk doch dringend, de afzender van het bericht op hoogte te brengen en alle informatie uit uw computer te verwijderen zonder acht te nemen van de inhoud. This E-mail has been scanned for viruses by the McAfee Appliance of Blgg B.V. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] periodogram smoothing question
Hello - I am currently simulating bivariate AR(1) time series data and have the following line in my code: Px=spec.pgram(ts.union(X,XX),spans=c(?,?)) The spans option is where I enter in the vector containing the Daniell smoother numbers, but I don't know what a Daniell smoother is (hence the question marks). Can somebody please tell me? Is there another option where I can simply enter in the smoothing window width? Thanks, Ferebee __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] reading scanned graphs
Remko Duursma wrote: Check out http://www.datathief.org/ Remko For Linux I recommend the engauge-digitizer package. Frank - Remko Duursma Post-Doctoral Fellow Centre for Plant and Food Science University of Western Sydney Hawkesbury Campus Richmond NSW 2753 Dept of Biological Science Macquarie University North Ryde NSW 2109 Australia Mobile: +61 (0)422 096908 2009/3/4 Aad Termorshuizen aad.termorshui...@blgg.nl: I have a graph (a curve with a x- and y-axis) on paper. (The device from which I receive this is not able to output the data digitally unfortunately.) Is there a procedure by which I can scan the graph, save it into, e.g., a bitmap file and make it R to read this file into x and y coordinates? Thanks, Aad Termorshuizen _ Disclaimer: Dit e-mailbericht is uitsluitend bestemd voor de geadresseerde(n). Indien dit e-mailbericht niet voor u bestemd is, verzoeken wij u vriendelijk doch dringend, de afzender van het bericht op hoogte te brengen en alle informatie uit uw computer te verwijderen zonder acht te nemen van de inhoud. This E-mail has been scanned for viruses by the McAfee Appliance of Blgg B.V. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Frank E Harrell Jr Professor and Chair School of Medicine Department of Biostatistics Vanderbilt University __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] score test statistic in logistic regression
re post from bkelcey at umich.edu bkelcey at umich.edu Wed Feb 27 15:09:48 CET 2008 If the response y is given as the proportion of successes out of n trials, and y, n, p, x, and z are vectors of length M, and the model is logit(p) = b0 + b1*x + b2*z then for the score test for the null hypothesis b1=0 use des-array(c(rep(1,M),x,z),dim=c(M,3)) m0-glm(y~z,binomial,weights=n) f-fitted(m0) efscor-1:3 for (i in 1:3) { efscor[i]-sum((y-f)*n*des[,i]) } fim-outer(1:3,1:3) for (i in 1:3) { for (j in 1:3) { fim[i,j]-sum(des[,i]*des[,j]*n*f*(1-f)) } } score-crossprod(efscor,crossprod(solve(fim),efscor)) score ok? greg __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.