Re: [R] converting character strings to numbers
There is no support for 'thousands separators' in R's input/conversion routines, mainly because C has no support either (not even for output in the C standard). We could of course add our own layer, but as far as I am aware this is the first time this has come up. On Thu, 14 Jun 2007, Andrew J Tyre wrote: I have a comma delimited text file in which many columns of numbers are also quoted and have commas as well as decimals. I was surprised to find read.csv() didn't import this seamlessly, even after messing around with the colClasses argument. I did find a solution to convert the character strings after reading them in, but wonder if there isn't a better one I overlooked. test = c(10,522.5,11,768.9,11,354.3) as.numeric(test) # fails as.numeric(gsub(,,,test)) # works Any suggestions? Or is this as good as it gets? I'm not complaining ... just curious! Drew Tyre School of Natural Resources University of Nebraska-Lincoln 416 Hardin Hall, East Campus Lincoln, NE 68583-0974 phone: +1 402 472 4054 fax: +1 402 472 2946 email: [EMAIL PROTECTED] http://snr.unl.edu/tyre [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] problem with hist()
Mario Dejung wrote: Hey everybody, I try to make a graph with two different plots. First I make a boxplot of my data. It is a collection off correlation values of different pictures. For example: 0.23445 pica 0.34456 pica 0.45663 pica 0.98822 picb 0.12223 picc 0.34443 picc etc. Ok, I make this boxplot and I get for every picture the boxes. After this I want to know, how many correlations per picture exist. So I make a new vector y - as.numeric(data$picture) So I get for my example something like this: y [1] 1 1 1 1 1 1 1 1 1 1 [11] 1 1 1 1 1 1 1 1 2 2 ... [16881] 59 59 59 60 60 60 60 60 60 60 After this I make something like this boxplot(cor ~ pic) par(new = TRUE) hist(y, nclass = 60) But there is my problem. I have 60 pictures, so I get 60 different boxplots, and I want the hist behind the boxes. But it makes only 59 histbars. What can I do? I tried also hist(y, 1:60) # same effect and hist(y, 1:61) this give me 60 places, but only 59 bars. the last bar is 0. In fact, I am pretty sure you really want to have a barplot() rather than a histogram. If you really want to use hist(), then perhaps hist(y, 0:60). Uwe Ligges Ok, it seems you are right, but I'm a beginner in R, so maybe you can help me a little bit more. When I use the hist function, it automatically uses the frequency of the different numbers, so I will get normally 60 bars. When I use the barplot function, then I have to count the frequency of the numbers. Is there a function who can do this, or should I write a function on my own. Sorry, I'm sure it is a stupid question, but maybe someone can give me a good reference for answers, because I am a beginner :-) Thanks to everyone Mario I hope anyone can help me. Regards Mario __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] problem with hist()
On 6/14/07, Mario Dejung [EMAIL PROTECTED] wrote: Hey everybody, I try to make a graph with two different plots. First I make a boxplot of my data. It is a collection off correlation values of different pictures. For example: 0.23445 pica 0.34456 pica 0.45663 pica 0.98822 picb 0.12223 picc 0.34443 picc etc. Ok, I make this boxplot and I get for every picture the boxes. After this I want to know, how many correlations per picture exist. So I make a new vector y - as.numeric(data$picture) So I get for my example something like this: y [1] 1 1 1 1 1 1 1 1 1 1 [11] 1 1 1 1 1 1 1 1 2 2 ... [16881] 59 59 59 60 60 60 60 60 60 60 After this I make something like this boxplot(cor ~ pic) par(new = TRUE) hist(y, nclass = 60) But there is my problem. I have 60 pictures, so I get 60 different boxplots, and I want the hist behind the boxes. But it makes only 59 histbars. What can I do? I tried also hist(y, 1:60) # same effect and hist(y, 1:61) this give me 60 places, but only 59 bars. the last bar is 0. I hope anyone can help me. What does the y axis represent? It will be counts for the histogram, and correlations for the boxplots. These aren't comparable, so you're probably better off making two separate graphics. Hadley __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] problem with hist()
Mario Dejung wrote: Mario Dejung wrote: Hey everybody, I try to make a graph with two different plots. First I make a boxplot of my data. It is a collection off correlation values of different pictures. For example: 0.23445 pica 0.34456 pica 0.45663 pica 0.98822 picb 0.12223 picc 0.34443 picc etc. Ok, I make this boxplot and I get for every picture the boxes. After this I want to know, how many correlations per picture exist. So I make a new vector y - as.numeric(data$picture) So I get for my example something like this: y [1] 1 1 1 1 1 1 1 1 1 1 [11] 1 1 1 1 1 1 1 1 2 2 ... [16881] 59 59 59 60 60 60 60 60 60 60 After this I make something like this boxplot(cor ~ pic) par(new = TRUE) hist(y, nclass = 60) But there is my problem. I have 60 pictures, so I get 60 different boxplots, and I want the hist behind the boxes. But it makes only 59 histbars. What can I do? I tried also hist(y, 1:60) # same effect and hist(y, 1:61) this give me 60 places, but only 59 bars. the last bar is 0. In fact, I am pretty sure you really want to have a barplot() rather than a histogram. If you really want to use hist(), then perhaps hist(y, 0:60). Uwe Ligges Ok, it seems you are right, but I'm a beginner in R, so maybe you can help me a little bit more. When I use the hist function, it automatically uses the frequency of the different numbers, so I will get normally 60 bars. When I use the barplot function, then I have to count the frequency of the numbers. Is there a function who can do this, or should I write a function on my own. You are looking for table(). Sorry, I'm sure it is a stupid question, but maybe someone can give me a good reference for answers, because I am a beginner :-) There are several manuals, books and the mailing list archives. Uwe Ligges Thanks to everyone Mario I hope anyone can help me. Regards Mario __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] interpretation of F-statistics in GAMs
dear listers, I use gam (from mgcv) for evaluation of shape and strength of relationships between a response variable and several predictors. How can I interpret the 'F' values viven in the GAM summary? Is it appropriate to treat them in a similar manner as the T-statistics in a linear model, i.e. larger values mean that this variable has a stronger impact than a variable with smaller F? When I run my analysis for two different response varables (but identical predictors), is there a way to compare the F values among tests (like to standardize them by teh sum of F within each test?) I append two summaries below. Thanks in advance, Robert ### example 1 ### Family: gaussian Link function: identity Formula: dep[sel, i] ~ s(date, k = 3) + s(depth, k = kn) + s(temp, k = kn) + s(light, k = kn) + s(PO4, k = kn) + s(DIN, k = kn) + s(prop.agpla, k = kn) Parametric coefficients: Estimate Std. Error t value Pr(|t|) (Intercept) 5.1048 0.0384 132.9 2e-16 *** --- Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 Approximate significance of smooth terms: edf Est.rank F p-value s(date) 1.6692 12.161 1.07e-05 *** s(depth) 1.6712 36.125 4.85e-14 *** s(temp) 1.9272 6.686 0.00156 ** s(light) 1.8862 12.604 7.20e-06 *** s(PO4)1.6762 3.237 0.04143 * s(DIN)1.0001 38.428 3.41e-09 *** s(prop.agpla) 1.4052 15.987 3.79e-07 *** --- Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 R-sq.(adj) = 0.687 Deviance explained = 70.5% GCV score = 0.31995 Scale est. = 0.30076 n = 204 ### example 2 ### Family: gaussian Link function: identity Formula: dep[sel, i] ~ s(date, k = 3) + s(depth, k = kn) + s(temp, k = kn) + s(light, k = kn) + s(PO4, k = kn) + s(DIN, k = kn) + s(prop.agpla, k = kn) Parametric coefficients: Estimate Std. Error t value Pr(|t|) (Intercept) 7.135880.05549 128.6 2e-16 *** --- Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 Approximate significance of smooth terms: edf Est.rank F p-value s(date) 1.9442 15.997 3.67e-07 *** s(depth) 1.8762 25.427 1.52e-10 *** s(temp) 1.0001 2.866 0.0921 . s(light) 1.7512 4.212 0.0162 * s(PO4)1.9502 10.632 4.14e-05 *** s(DIN)1.8052 10.745 3.73e-05 *** s(prop.agpla) 1.7152 2.674 0.0715 . --- Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 R-sq.(adj) = 0.479 Deviance explained = 50.9% GCV score = 0.6863 Scale est. = 0.64348 n = 209 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] arturo invites you to join Zorpia
Hi r-help! Your friend arturo from , just invited you to online photo albums and journals at Zorpia.com. So what is Zorpia? It is an online community that allows you to upload unlimited amount of photos, write journals and make friends. We also have a variety of skins in store for you so that you can customize your homepage freely. Join now for free! Please click the following link to join Zorpia: http://signup2.zorpia.com/signup?invitation_key=20070657c3fd0889d8d1e73b6d8a00c8referral=arturitocoral This message was delivered with the arturo's initiation. If you wish to discontinue receiving invitations from us, please click the following link: http://signup2.zorpia.com/email/optout/r-help@stat.math.ethz.ch [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] problem with hist()
On 6/15/07, Mario Dejung [EMAIL PROTECTED] wrote: On 6/14/07, Mario Dejung [EMAIL PROTECTED] wrote: Hey everybody, I try to make a graph with two different plots. First I make a boxplot of my data. It is a collection off correlation values of different pictures. For example: 0.23445 pica 0.34456 pica 0.45663 pica 0.98822 picb 0.12223 picc 0.34443 picc etc. Ok, I make this boxplot and I get for every picture the boxes. After this I want to know, how many correlations per picture exist. So I make a new vector y - as.numeric(data$picture) So I get for my example something like this: y [1] 1 1 1 1 1 1 1 1 1 1 [11] 1 1 1 1 1 1 1 1 2 2 ... [16881] 59 59 59 60 60 60 60 60 60 60 After this I make something like this boxplot(cor ~ pic) par(new = TRUE) hist(y, nclass = 60) But there is my problem. I have 60 pictures, so I get 60 different boxplots, and I want the hist behind the boxes. But it makes only 59 histbars. What can I do? I tried also hist(y, 1:60) # same effect and hist(y, 1:61) this give me 60 places, but only 59 bars. the last bar is 0. I hope anyone can help me. What does the y axis represent? It will be counts for the histogram, and correlations for the boxplots. These aren't comparable, so you're probably better off making two separate graphics. Hadley The boxplots show only the median, min, max, etc of the different pictures, but I want to know, how many entry's are in this plot. Now I have done this by the hist function, and when I use different colors, you can see, for the first picture there are about 130 entry, but for the 8th picture, there are only 40 entry's... Doesn't make this sense? I think your plot would be more clear if you used two graphics - one showing the spread, and one showing the number of points (you might also want to look at notched boxplots). In the graphic you attached the bars of the barchart (not histogram! - that's for continuous data) distract the eye from the boxplots. You might also want to try ordering the x axis by mean or number of observations as this will make it easier to see trends in the data. The confusion with the barchart arises because there are really two quite different types of barcharts. One type is basically the same as a dotchart, but you draw bars instead of dots - this is the default in R. The other type is the categorical analog of the histogram, and this is the default in ggplot2 (http://had.co.nz/ggplot2/geom_bar.html), allow the next version will automatically work out which version you want. Hadley __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] package with fortran 90 subroutines under windows xp
Hello, I work under windows xp and I am trying to build a R package with a subroutine written in fortran 90. I have installed all the updated tools and I am working with R-2.4.0 or R-2.5.0. When I check a package with a subroutine in fortran 77 (and extension f) everything is ok. When I try to build the same package with a subroutine in fortran 90 (with extension f90) the following warning appears: Subdirectory 'src' contains no source files and the package can not be built. The funny thing is that I have successfully built the same package with fortran 90 last March and everything was good. I can not imagine what is the problem, could do you help me? Cinzia Viroli Dipartimento di Scienze Statistiche Paolo Fortunati Via delle Belle Arti 41 40126 Bologna Italy Ph. +39 051 2094628 Fax +39 051 232153 home: www2.stat.unibo.it/viroli __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] problems with matrix, list and other..
hi can anyone help me to solve these problems? i have: 1) d matrix with 1096 rows; for example, d[2]= [,1] [,2] [,3] [1,] 0.1192566 0.000 0.000 [2,] 0.000 0.1065938 0.000 [3,] 0.000 0.000 0.103 if I class (d[2]) = list solve(d[2]) = error!!! 2) e list with 1096 rows; for example e[2]2= [[1]] [1] -1.0892216 -0.7304947 -1.2883680 d[2]%*%t(e[2]) this is the error: requires numeric matrix/vector arguments i've tried to coerce e to a matrix, but it's doesn't work... in the end.. i'd like this: for (i in (1:1096)) {solve(d[i])*t(e[i])} help me, please :) Vincenzo -- View this message in context: http://www.nabble.com/problems-with-matrix%2C-list-and-other..-tf3926701.html#a11135888 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] gpclib problem
Hi all, I am trying to test the new adehabitat package but I have a problem with a linked library, gpclib. When I try do install it I have this output: install.packages(gpclib, dependencies=TRUE, repos=http://cran.cnr.berkeley.edu/;) Avviso in install.packages(gpclib, dependencies = TRUE, repos = http://cran.cnr.berkeley.edu/;) : argument 'lib' is missing: using /usr/local/lib/R/site-library dependency ''gpclib'' is not available I have the same problem in other repository. Leonardo __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] back-transform predictors for x-axis in plot -- mgcv package
Hi Suzan, You can do sort of backtransformation inside of ggplot2 (http://had.co.nz/ggplot2). library(ggplot2) # Create the base scatterplot with y and x axes transformed by logging, # and then back transformed by exponentiating (base - qplot(carat, price, data=diamonds) + scale_x_log10() + scale_y_log10() + coord_trans(y=pow10, x=pow10)) base + geom_smooth(method=lm) library(mgcv) base + geom_smooth(method=gam, formula = y ~ s(x, bs=cr)) base + geom_smooth(method=gam, formula = y ~ s(x, bs=cr), fill=grey50) # cf. qplot(carat, price, data=diamonds) + geom_smooth(method=lm) qplot(carat, price, data=diamonds) + geom_smooth(method=gam, formula = y ~ s(x, bs=cr), fill=grey50) Regards, Hadley On 6/14/07, Suzan Pool [EMAIL PROTECTED] wrote: My question is related to plot( ) in the mgcv package. Before modelling the data, a few predictors were transformed to normalize them. Therefore, the x-axes in the plots show transformed predictor values. How do I back-transform the predictors so that the plots are easier to interpret? Thanks in advance, Suzan -- Suzan Pool Oregon State University Cooperative Institute for Marine Resources Studies c/o NOAA Fisheries 520 Heceta Place P.O. Box 155 Hammond, OR 97121 [EMAIL PROTECTED] [EMAIL PROTECTED] Phone: 503-861-1818 x36 TTY Voice to TTY: 711 Fax: 503-861-2589 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] JGR, Java and Kubuntu 7.04 ...
You could download the latest R (2.5.0) directly from CRAN, this just as a side remark, You find instructions here: http://cran.au.r-project.org/bin/linux/ubuntu/README With java you should make sure that you have the Java 5 JDK installed. I dont know how it is with Ubuntu but on fedora I have to set the java alternative (sun java installation does not change the default java there). Which java is installed you can check with java -version there you see it it is properly installed and a R CMD javareconf also indicates if your java is configured correctly. Additionally on Fedora I had to set JAVA_HOME to the JDK directory manually. Stefan Original Message Subject: [R] JGR, Java and Kubuntu 7.04 ... From: John Logsdon [EMAIL PROTECTED] To: r-help@stat.math.ethz.ch Date: 14.06.2007 19:16 R-ists Yet again Java rears its ugly head. I have Kubuntu 7.04 running the Kubuntu-repository version of R 2.4.1-1. Yes it isn't the very latest version but this is not the issue here. I want a Windows-like environment and everyone is talking about JGR. I downloaded it and installed it along with rJava. Both compile and install satisfactorily. But when I come to run it: library('JGR') Loading required package: rJava Error in dyn.load(x, as.logical(local), as.logical(now)) : unable to load shared library '/usr/local/lib/R/site-library/rJava/libs/rJava.so': /usr/local/lib/R/site-library/rJava/libs/rJava.so: undefined symbol: JNI_GetCreatedJavaVMs Error: .onLoad failed in 'loadNamespace' for 'rJava' Error: package 'rJava' could not be loaded When I first tried to run it without Java being installed, I got a message saying that JDK wasn't installed but mentioned 1.4.2. The version of Java actually installed as the latest from the Ubuntu repository is Sun 1.5.0.11. I don't see the point in installing old versions of Java just for one application because the language, or at least the writing, should be backwards compatible. In all aspects I have seen Kubuntu is a very impressive in checking compatibility. Unfortunately this is frequently not the case with Java. I steer clear of Java as much as possible. Can anyone suggest what I should do? Use Windows perhaps? Run Windows in a kvm virtual machine just to run R? Put my head in a bucket of cold water? Is there an alternative IDE? Is there a later JGR somewhere that is not yet on CRAN? TIA __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] gpclib problem
On Fri, 15 Jun 2007, Leonardo Lami wrote: Hi all, I am trying to test the new adehabitat package but I have a problem with a linked library, gpclib. When I try do install it I have this output: install.packages(gpclib, dependencies=TRUE, repos=http://cran.cnr.berkeley.edu/;) Avviso in install.packages(gpclib, dependencies = TRUE, repos = http://cran.cnr.berkeley.edu/;) : argument 'lib' is missing: using /usr/local/lib/R/site-library dependency ''gpclib'' is not available I have the same problem in other repository. You haven't told us your OS or R version (or what the 'other repository' was). The files are there http://cran.cnr.berkeley.edu/src/contrib/Descriptions/gpclib.html so a guess is that you have failed to update your R. Leonardo __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. Please do as this asks. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] package with fortran 90 subroutines under windows xp
On 12/06/2007 9:41 AM, Cinzia Viroli wrote: Hello, I work under windows xp and I am trying to build a R package with a subroutine written in fortran 90. I have installed all the updated tools and I am working with R-2.4.0 or R-2.5.0. When I check a package with a subroutine in fortran 77 (and extension f) everything is ok. When I try to build the same package with a subroutine in fortran 90 (with extension f90) the following warning appears: Subdirectory 'src' contains no source files and the package can not be built. The funny thing is that I have successfully built the same package with fortran 90 last March and everything was good. I can not imagine what is the problem, could do you help me? The current Windows tools don't support Fortran 9x, just Fortran 77. If you have your own Fortran 90 compiler, you can enter it into MkRules (either by enabling the GCC4 macro, or entering it directly in the F95 macro). But you should have got an error about the compiler not existing, not an error about no source files. Are you sure you named the file *.f90? I get a check warning about no source files if I name the file test.F90, but not for test.f90. Duncan Murdoch __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Loop for test statistic
I would like to obtain the statistic of A2 for ycf between the value 0.0032 and 0.09, and I am using the following codes. while (0.0032 ycf 0.09) { A2 - A2_GOFlaio(ycf, dist=GEV)[1] print(A2) } Could anyone give me some advice as it does not work. Many thanks. -- View this message in context: http://www.nabble.com/Loop-for-test-statistic-tf3927249.html#a11137549 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Grahpics problem
To the group: A marvelous thing is combining graphs with the par function, but there remains an issue. What if you wish to put a title on top of the set of graphs or a general x or y axis label? How does one do this? With kindest regards. Mitchell S. Wachtel, MD [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] connecting to DB2 database
Hi, On 6/14/07, Prof Brian Ripley [EMAIL PROTECTED] wrote: On Thu, 14 Jun 2007, Aydemir, Zava (FID) wrote: i am trying to connect to a DB2 server using the DBI library. The DBI *package* does not allow you to connect to anything by itself. For that you need a driver package, currently available for MySQL, ORACLE and SQLite (only, AFAIK). There are ODBC drivers for DB2 (on several platforms) so perhaps you could use RODBC: perhaps also RJDBC. RJDBC also uses the DBI interface. Regards, -- David getData - function() { driver - dbDriver(DB2) conn - dbConnect(driver,server,uname,pword) data - dbSendquery(conn, select etc.) } When I run the function, i get the error data - getData() Error in do.call(as.character(drvName), list(...)) : could not find function DB2 Can anyone help me here? Thank you Zava This is not an offer (or solicitation of an offer) to buy/se...{{dropped}} __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] JGR, Java and Kubuntu 7.04 ...
On 15 June 2007 at 12:32, Stefan Grosse wrote: | You could download the latest R (2.5.0) directly from CRAN, this just as | a side remark, You find instructions here: | http://cran.au.r-project.org/bin/linux/ubuntu/README Yes indeed, thanks for our fearless backporters. | With java you should make sure that you have the Java 5 JDK installed. | I dont know how it is with Ubuntu but on fedora I have to set the java | alternative (sun java installation does not change the default java | there). Which java is installed you can check with java -version there | you see it it is properly installed and a R CMD javareconf also | indicates if your java is configured correctly. Additionally on Fedora I | had to set JAVA_HOME to the JDK directory manually. It all worked -- I don't run Java much myself, but I think starting with the previous Ubuntu release and the actual Sun packages directly acessible via apt-get, it worked. However, with the current packages in my Kubuntu setup at work, I do get a segfault once JGR is up and loaded. I suspect, as Stefan said here, that this is due to the Java 6 packages in Ubuntu. As I said, sun-java5-jdk seems to have work along with 'sudo R CMD javareconf', Simon's powerful Java parameter setter for R. In any event, we should carry this over to r-sig-debian. Dirk | | Stefan | | Original Message | Subject: [R] JGR, Java and Kubuntu 7.04 ... | From: John Logsdon [EMAIL PROTECTED] | To: r-help@stat.math.ethz.ch | Date: 14.06.2007 19:16 | R-ists | | Yet again Java rears its ugly head. | | I have Kubuntu 7.04 running the Kubuntu-repository version of R 2.4.1-1. | Yes it isn't the very latest version but this is not the issue here. | | I want a Windows-like environment and everyone is talking about JGR. | | I downloaded it and installed it along with rJava. Both compile and install | satisfactorily. | | But when I come to run it: | | | library('JGR') | | Loading required package: rJava | Error in dyn.load(x, as.logical(local), as.logical(now)) : | unable to load shared library | '/usr/local/lib/R/site-library/rJava/libs/rJava.so': | /usr/local/lib/R/site-library/rJava/libs/rJava.so: undefined symbol: | JNI_GetCreatedJavaVMs Error: .onLoad failed in 'loadNamespace' for 'rJava' | Error: package 'rJava' could not be loaded | | | | When I first tried to run it without Java being installed, I got a message | saying that JDK wasn't installed but mentioned 1.4.2. The version of Java | actually installed as the latest from the Ubuntu repository is Sun 1.5.0.11. | I don't see the point in installing old versions of Java just for one | application because the language, or at least the writing, should be | backwards compatible. | | In all aspects I have seen Kubuntu is a very impressive in checking | compatibility. Unfortunately this is frequently not the case with Java. I | steer clear of Java as much as possible. | | Can anyone suggest what I should do? Use Windows perhaps? Run Windows in a | kvm virtual machine just to run R? Put my head in a bucket of cold water? | Is there an alternative IDE? Is there a later JGR somewhere that is not yet | on CRAN? | | TIA | | | | __ | R-help@stat.math.ethz.ch mailing list | https://stat.ethz.ch/mailman/listinfo/r-help | PLEASE do read the posting guide http://www.R-project.org/posting-guide.html | and provide commented, minimal, self-contained, reproducible code. -- Hell, there are no rules here - we're trying to accomplish something. -- Thomas A. Edison __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] problems with matrix, list and other..
You only got what you deserved when not reading the manual... R-Intro, Chapters 5 and 6, page 26 in particular. http://cran.r-project.org/doc/manuals/R-intro.pdf Petr billycorg napsal(a): hi can anyone help me to solve these problems? i have: 1) d matrix with 1096 rows; for example, d[2]= [,1] [,2] [,3] [1,] 0.1192566 0.000 0.000 [2,] 0.000 0.1065938 0.000 [3,] 0.000 0.000 0.103 if I class (d[2]) = list solve(d[2]) = error!!! 2) e list with 1096 rows; for example e[2]2= [[1]] [1] -1.0892216 -0.7304947 -1.2883680 d[2]%*%t(e[2]) this is the error: requires numeric matrix/vector arguments i've tried to coerce e to a matrix, but it's doesn't work... in the end.. i'd like this: for (i in (1:1096)) {solve(d[i])*t(e[i])} help me, please :) Vincenzo -- Petr Klasterecky Dept. of Probability and Statistics Charles University in Prague Czech Republic __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Loop for test statistic
Please read the posting guide. How shall we know what ycf, and A2 are? In general, the condition should be something like (.0032ycf ycf.09) Also make sure that the initial value of ycf actually fits between the limits (or assign something to A2 outside the loop) and that ycf is changed inside the loop (to prevent looping forever). Petr livia napsal(a): I would like to obtain the statistic of A2 for ycf between the value 0.0032 and 0.09, and I am using the following codes. while (0.0032 ycf 0.09) { A2 - A2_GOFlaio(ycf, dist=GEV)[1] print(A2) } Could anyone give me some advice as it does not work. Many thanks. -- Petr Klasterecky Dept. of Probability and Statistics Charles University in Prague Czech Republic __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Grahpics problem
On 6/15/2007 8:13 AM, Wachtel, Mitchell wrote: To the group: A marvelous thing is combining graphs with the par function, but there remains an issue. What if you wish to put a title on top of the set of graphs or a general x or y axis label? How does one do this? Use mtext() to write to the outer margins, but remember first to create some space there. For example, par(oma=c(0,0,2,0),mfrow=c(2,2)) plot(1) plot(2) plot(3) plot(4) mtext(Main title, side=3, outer=TRUE) Duncan Murdoch __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Grahpics problem
RSiteSearch(common title) Petr Wachtel, Mitchell napsal(a): To the group: A marvelous thing is combining graphs with the par function, but there remains an issue. What if you wish to put a title on top of the set of graphs or a general x or y axis label? How does one do this? With kindest regards. Mitchell S. Wachtel, MD [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Petr Klasterecky Dept. of Probability and Statistics Charles University in Prague Czech Republic __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Problem with workspace loading after languageR use
Hello R, To analyze multi-level data, I started learning and using lmer. So far so wonderful. I then found some useful functions in package languageR. But then the following problem ocurred: Whenever I load and use the languageR package, then save the workspace - or quit R with saving the workspace - I am unable to reload that workspace in a later session. That is, R doesn't start at all when I try to start it by clicking the workspace file. Loading languageR before loading the workspace doesn't help, but yields the message: Error in load(D:\\statistics\\MultilevelAnalysis\\.RData) : could not find function findPackageEnv Thus, the saved workspace remains inaccessible. I not 100% certain that languageR is the scapegoat, but my trial-and-error experiments indicate it is. My system is Win XP Home/Professional: sessionInfo() R version 2.5.0 Patched (2007-04-24 r41305) i386-pc-mingw32 locale: LC_COLLATE=German_Germany.1252;LC_CTYPE=German_Germany.1252;LC_MONETARY=German_Germany.1252;LC_NUMERIC=C;LC_TIME=German_Germany.1252 attached base packages: [1] splines stats graphics grDevices utils [6] datasets methods base other attached packages: languageR rpartMASS Designsurvival 0.23.1-367.2-342.0-12 2.31 Hmisc e1071 class cluster zipfR 3.3-21.5-167.2-341.11.7 0.6-0 lme4coda Matrix lattice 0.99875-10.11-2 0.99875-20.15-8 thanks for any helpful suggestions! best Rüdiger -- Hans-Rüdiger Pfister Professor of Business Psychology University of Lüneburg Faculty for Business Administration, Behavioral Sciences and Law Institute of Experimental Industrial Psychology (LueneLab) Wilschenbrucher Weg 84 D-21335 Lüneburg, Germany +49-(0)4131 677 7759 [EMAIL PROTECTED] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] method of rpart when response variable is binary?
Dear all, I would like to model the relationship between y and x. y is binary variable, and x is a count variable which may be possion-distribution. I think it is better to divide x into intervals and change it to a factor before calling glm(y~x,data=dat,family=binomail). I try to use rpart. As y is binary, I use class method and get the following result. rpart(y~x,data=dat,method=class) n=778 (22 observations deleted due to missingness) node), split, n, loss, yval, (yprob) * denotes terminal node 1) root 778 67 0 (0.91388175 0.08611825) * If with the default method, I get such a result. rpart(y~x,data=dat) n=778 (22 observations deleted due to missingness) node), split, n, deviance, yval * denotes terminal node 1) root 778 61.230080 0.08611825 2) x 19.5 750 53.514670 0.0773 4) x 1.25 390 17.169230 0.04615385 * 5) x=1.25 360 35.60 0.1110 * 3) x=19.5 28 6.107143 0.32142860 * If I use 1.25 and 19.5 as the cutting points, change x into factor by x2 - cut(q34b,breaks=c(0,1.25,19.5,200),right=F) The coef in y~x2 is significant and makes sense. My problem is: is it OK use the default method in rpart when response varibale is binary one? Thanks. -- Ronggui Huang Department of Sociology Fudan University, Shanghai, China __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Need Help with Dendrogram and DataFrame Leaf names
I having problem with dendrogram leaf names when I read a tab delimited file into dataframe; I have a text file, tab delimited, using read.table into a data frame as follows: test1-read.table(c:\\R\\data\\Tremont4.txt, header=TRUE, sep=\t) When I do this the test1 data frame is picking up my first column names as part of the data and not the case names, the leafs are the numbers on the left 1-13 As opposed to the text names to the right. Example Output from displaying dataframe: test1 row.names X1.31.2004 X2.29.2004 X3.31.2004 X4.30.2004 X5.31.2004 X6.30.2004 X7.31.2004 X8.31.2004 X9.30.2004 X10.31.2004 1 ConvertibleArbitrage 0.014 0.003 0.004 0.005 -0.013 -0.008 -0.002 0.003 -0.001 -0.003 2 DedicatedShortBias -0.017 0.003 -0.026 0.042 0.008 -0.013 0.081 0.013 -0.019 -0.018 3 EmergingMarkets 0.025 0.014 0.018 -0.033 -0.018 0.009 -0.001 0.018 0.023 0.024 4MarketNeutral 0.008 0.008 -0.001 -0.003 0.002 0.008 0.003 0.021 0.005 0.000 5 EventDriven 0.022 0.010 0.005 0.005 0.001 0.010 0.000 0.005 0.013 0.012 6 Distressed 0.024 0.009 0.006 0.007 0.003 0.011 0.005 0.006 0.012 0.019 7 EventdriveMultiStrategy 0.020 0.011 0.003 0.005 -0.001 0.009 -0.003 0.004 0.014 0.007 8RiskArbitrage 0.008 0.005 0.007 -0.006 0.004 0.003 -0.015 0.002 0.006 0.009 9 FixedIncomeArbitrage 0.012 0.009 -0.005 0.013 0.006 0.007 0.007 -0.004 -0.008 0.011 10 GlobalMacro 0.015 0.012 0.010 0.001 0.001 0.005 0.008 -0.008 -0.005 0.012 11 LongShortEquity 0.020 0.018 0.002 -0.014 -0.004 0.007 -0.014 0.001 0.024 0.014 12 ManagedFutures 0.011 0.069 -0.009 -0.065 -0.011 -0.028 -0.020 -0.015 0.020 0.048 13 Multi-Strategy 0.016 0.004 0.004 0.003 -0.001 0.001 -0.003 0.004 0.006 0.006 Input file looks like this: row.names 1/31/2004 2/29/2004 3/31/2004 4/30/2004 5/31/2004 6/30/2004 7/31/2004 8/31/2004 ConvertibleArbitrage0.0140.003 0.004 0.005-0.013 -0.008 -0.002 0.003 DedicatedShortBias -0.017 0.003 -0.026 0.042 0.008 -0.013 0.081 0.013 EmergingMarkets 0.0250.0140.018-0.033 -0.0180.009 -0.001 0.018 MarketNeutral 0.0080.008-0.001 -0.003 0.002 0.008 0.003 0.021 Etc... Would appreciate why the read.table into dataframe sees the text as part of the data oand Not the observation names and is making the numbers the leaf names and observation names. Thanks for any help, Neil Gottlieb This information is being sent at the recipient's request or...{{dropped}} __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] method of rpart when response variable is binary?
On Fri, 15 Jun 2007, ronggui wrote: Dear all, I would like to model the relationship between y and x. y is binary variable, and x is a count variable which may be possion-distribution. I think it is better to divide x into intervals and change it to a factor before calling glm(y~x,data=dat,family=binomail). I try to use rpart. As y is binary, I use class method and get the following result. rpart(y~x,data=dat,method=class) n=778 (22 observations deleted due to missingness) node), split, n, loss, yval, (yprob) * denotes terminal node 1) root 778 67 0 (0.91388175 0.08611825) * If with the default method, I get such a result. rpart(y~x,data=dat) n=778 (22 observations deleted due to missingness) node), split, n, deviance, yval * denotes terminal node 1) root 778 61.230080 0.08611825 2) x 19.5 750 53.514670 0.0773 4) x 1.25 390 17.169230 0.04615385 * 5) x=1.25 360 35.60 0.1110 * 3) x=19.5 28 6.107143 0.32142860 * If I use 1.25 and 19.5 as the cutting points, change x into factor by x2 - cut(q34b,breaks=c(0,1.25,19.5,200),right=F) The coef in y~x2 is significant and makes sense. My problem is: is it OK use the default method in rpart when response varibale is binary one? Thanks. Not unless you want a least-squares fit. Note that you have only 8.6% of one class, and for such an unbalanced classification problem you are unlikely to do better than declaring class 1 for all examples. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] importing .dta files
On Fri, 15 Jun 2007, Chris Linton wrote: I'm trying to read in a Stata file but I've never used this function ( read.dta). It's the only one that seems to come close to working, but I keep getting this error: data-read.dta(C:/Documents and Settings/Chris/Desktop/S4412/catestscores.dta) Error in read.dta(C:/Documents and Settings/Chris/Desktop/S4412/catestscores.dta, : a binary read error occurred There's little chance the data is corrupt considering it came from my professor and he used the data earlier. So, either I'm doing something wrong or R just doesn't like to read in Stata files. If it's a problem with R, how can I easily convert the file without purchasing Stata? R does read Stata files -- I use this facility frequently. It's hard to tell why it isn't working in your case, since we don't know anything about the file, your version of R, version of Stata, etc (we can guess you are on windows from the file name). The error message implies that the file was found, and that it started with the right sequence of bytes to be a Stata .dta file, but that something (probably the end of the file) prevented R from reading what it was expecting to read. This is why (in the absence of any further information) the natural suspicion is that the file is corrupt. It is possible that we have misunderstood some unusual possibility in the Stata file format -- this has happened once before -- but it is fairly well documented. In any case, there is not much that can be done without more information. -thomas __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Coefficients and Covariances in MNP
Hi everyone, I use the MNP package to fit my data. I can interpret the coefficients and covariance. I cannot put them into the original model which is Wi=Sum(bXij) over j, Yi(Wi)=0 if max(Wi)0, j if max(Wi)=Wij0 j=3 here, 3 choice alternatives, Clio, Punto, Polo I don't know how to put the coefficients and covariances into W, to construct W for each j. The coefficients and covariances are below. I f anyone help me on this, I would appreciate. ozlem. Call: mnp(formula = choice ~ 1, data = rowdata3, choiceX = list(Punto = PricePunto, Clio = PriceClio, Polo = PricePolo), cXnames = price, base = Polo, n.draws = 100, verbose = TRUE) Coefficients: mean std.dev.2.5%97.5% (Intercept):Clio -1.45590.6889-2.7017 -0.387 (Intercept):Punto -0.19890.7180-2.2192 0.893 price 10.0865 5.1486 0.8549 23.030 Covariances: mean std.dev.2.5%97.5% Clio:Clio1.0.1. 1.000 Clio:Punto -0.9458 0.7644 -2.1816 0.449 Punto:Punto 2.59802.53120.4339 9.281 Base category: Polo Number of alternatives: 3 Number of observations: 12 Number of estimated parameters: 5 Number of stored MCMC draws: 100 [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Problem with workspace loading after languageR use
The problem would appear to be something missing in R. From what I can see you have saved a reference to a package environment in your workspace. When load() tries to resolve this, it calls findPackageEnv and that does not exist in current R (or any recent version I looked at). I think adding the following to your new session before load() will help findPackageEnv - function(info) as.environment(paste(package, info, sep=:)) will work, but if not try findPackageEnv - function(info) .GlobalEnv If you send me the problematic workspace (or reproduction instructions) I can take a closer look. On Fri, 15 Jun 2007, Pfister wrote: Hello R, To analyze multi-level data, I started learning and using lmer. So far so wonderful. I then found some useful functions in package languageR. But then the following problem ocurred: Whenever I load and use the languageR package, then save the workspace - or quit R with saving the workspace - I am unable to reload that workspace in a later session. That is, R doesn't start at all when I try to start it by clicking the workspace file. Loading languageR before loading the workspace doesn't help, but yields the message: Error in load(D:\\statistics\\MultilevelAnalysis\\.RData) : could not find function findPackageEnv Thus, the saved workspace remains inaccessible. I not 100% certain that languageR is the scapegoat, but my trial-and-error experiments indicate it is. My system is Win XP Home/Professional: sessionInfo() R version 2.5.0 Patched (2007-04-24 r41305) i386-pc-mingw32 locale: LC_COLLATE=German_Germany.1252;LC_CTYPE=German_Germany.1252;LC_MONETARY=German_Germany.1252;LC_NUMERIC=C;LC_TIME=German_Germany.1252 attached base packages: [1] splines stats graphics grDevices utils [6] datasets methods base other attached packages: languageR rpartMASS Designsurvival 0.23.1-367.2-342.0-12 2.31 Hmisc e1071 class cluster zipfR 3.3-21.5-167.2-341.11.7 0.6-0 lme4coda Matrix lattice 0.99875-10.11-2 0.99875-20.15-8 thanks for any helpful suggestions! best Rüdiger -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595__ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] mixed model for analysing microarray data
Dear R users, We are trying to analyse microarray data. The experiment compares two conditions (light vs. dark). we would like to use a mixed-model ANOVA but we don't know much about it. could you please help? the experiment test the effect of light on a tissue and compare tissues maintained in light or dark (fixed effect). In each condition, Two RNA samples were used (random effect?) for each condition, and each sample was measured twice (random effect? nested in sample?), or maybe the replicate factor is not required? Any help will be appreciated (particularly on how to assign the random factor(s) ) for a given gene the data look like that: light RNA replicate reading y 1 1 1.67 y 1 2 1.56 y 2 1 1.60 y 2 2 1.34 n 1 1 1.67 n 1 2 1.56 n 2 1 1.60 n 2 2 1.34 *** This email was sent via http://www.AnonymousSpeech.com, the worlds leading a href=http://www.anonymousspeech.comanonymous email/a provider. Paid memberships do not show this footer message. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] method of rpart when response variable is binary?
you might use default setting if you use as.factor(y)~x in rpart(), I think. On 6/15/07, ronggui [EMAIL PROTECTED] wrote: Dear all, I would like to model the relationship between y and x. y is binary variable, and x is a count variable which may be possion-distribution. I think it is better to divide x into intervals and change it to a factor before calling glm(y~x,data=dat,family=binomail). I try to use rpart. As y is binary, I use class method and get the following result. rpart(y~x,data=dat,method=class) n=778 (22 observations deleted due to missingness) node), split, n, loss, yval, (yprob) * denotes terminal node 1) root 778 67 0 (0.91388175 0.08611825) * If with the default method, I get such a result. rpart(y~x,data=dat) n=778 (22 observations deleted due to missingness) node), split, n, deviance, yval * denotes terminal node 1) root 778 61.230080 0.08611825 2) x 19.5 750 53.514670 0.0773 4) x 1.25 390 17.169230 0.04615385 * 5) x=1.25 360 35.60 0.1110 * 3) x=19.5 28 6.107143 0.32142860 * If I use 1.25 and 19.5 as the cutting points, change x into factor by x2 - cut(q34b,breaks=c(0,1.25,19.5,200),right=F) The coef in y~x2 is significant and makes sense. My problem is: is it OK use the default method in rpart when response varibale is binary one? Thanks. -- Ronggui Huang Department of Sociology Fudan University, Shanghai, China __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- WenSui Liu A lousy statistician who happens to know a little programming (http://spaces.msn.com/statcompute/blog) __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Error: bad value ? what is that?
Peter Dalgaard p.dalgaard at biostat.ku.dk writes: Jose Quesada wrote: Hi, I'm finding a very strange error. For no good reason my R console (Rgui.exe, R 2.5.0, under win XP) stops producing anything meaningful, and just returns: Error: bad value to _whatever_ I enter. It starts doing this after a while, not immediately when launched. I have to restart R when this happens. No idea why. I didn't change anything in the R config that I remenber. Any thoughts? Thanks. Hmm that message comes from deep down inside SETCAR() and friends. I can't see other reasons for it than memory corruption. Are you running some rogue C code? Is the machine flaky in other respects? __ R-help at stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. Thanks Peter, Interesting; I upgraded my laptop to 2 gigs. and from that point on, it sometimes fauls to hibernate (just hangs). Maybe this is related? I installed an XP patch to solve hibernation problems with laptops over 2 gigs... Maybe it didn't work? -Jose Re: Rogue c code: I was running library(Matrix), which is a pretty complicated piece of code and surely has c code... __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] [OT] 'gv' and fractional points
Hi Folks, This is off-topic R-wise, but it may be close to the heart of many R-users, so I think it may be the best place to ask! Users of 'gv' (the front end to ghostscript) will be aware of the little window which gives you the x-y coordinates (in points = 1/72 inch) of the position of the cross-hair mouse cursor. These coordinates are those of the corresponding position on the printed page, relative to some origin. I have often used this to extract numerical values for data from graphs in Postscript files (also PDF files, after you have converted them to PS). Then (veering back on topic ... ) you can submit the numerical data to R and try your own analyses on these data, and compare with what the article does. However, this little window only gives the numbers in whole points. Say a smallish graphic may print out 3 inches wide or high. Then you get precision of 1/216 per 3 inches or 0.4% of full scale. This can be adequate on many occasions, but can be on the coarse side on other occasions. Even for a 6-inch-wide/high graph, you only get down to 0.2% of full scale. If it were possible to induce 'gv' to display these coordinates in tenths of a point, then much greater precision (as adequate as one can expect to hope for when, in effect, measuring off the graph) could be obtained. Does anyone know: a) Whether it is possible to persuade 'gv' to give this display in fractional points (my own search of the documentation has not revealed anything); b) Of any alternative to 'gv' as PS viewer which would provide this capability? With thanks, and best wishes to all, Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 16:13:21 -- XFMail -- __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] registration density profiles : FDA
Hello, I would like to use fda library to compare sampled curves that are density profiles. But I encounter some problems to register my data : I know that I must use a registration function to align functions in order to have important features found in each curve occur at roughly the same argument value, prior to do subsequent analysis ( PCA and cluster analysis). But I do not know which function (if there exists one) allows me to synchronize density profiles : could help me please ? Thank you by advance Best regards Lise Bellanger -- Lise Bellanger, Université de Nantes Département de Mathématiques, Laboratoire Jean Leray UMR CNRS 6629 2, Rue de la Houssinière BP 92208 - F-44322 Nantes Cedex 03 Tél. : (33|0) 2 51 12 59 00 (ou 43) - Fax : (33|0) 2 51 12 59 12 E-Mail : [EMAIL PROTECTED] URL : http://www.math.sciences.univ-nantes.fr/ __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Problem with workspace loading after languageR use
Dear Brian, thanks a lot for your advice. Prof Brian Ripley wrote: On Fri, 15 Jun 2007, Prof Brian Ripley wrote: I think adding the following to your new session before load() will help findPackageEnv - function(info) as.environment(paste(package, info, sep=:)) this does not work. more likely findPackageEnv - function(info) as.environment(info) this works, if I load the languageR library before. is correct. will work, but if not try findPackageEnv - function(info) .GlobalEnv this works unconditionally. Including that line in a local .RProfile file, basically solves the problem. If you send me the problematic workspace (or reproduction instructions) I can take a closer look. A typical session would look like this (using example data from Hox(2002)): # from Hox (2002) hoxpop - read.table(http://www.ruediger-pfister.de/download/popular.dat;, header=TRUE) hoxpop$PUPIL - factor(hoxpop$PUPIL) hoxpop$SCHOOL - factor(hoxpop$SCHOOL) hoxpop$SEX - factor(hoxpop$SEX) # load languageR library(languageR) # do some analyses ... # save the workspace save.image(D:\\statistics\\MultilevelAnalysis\\lm.RData) After quitting R, the workspace lm.RData will not reload. (this workspace can be downloaded here: http://www.ruediger-pfister.de/Downloads/lmRData.zip) best Rüdiger On Fri, 15 Jun 2007, Pfister wrote: Hello R, To analyze multi-level data, I started learning and using lmer. So far so wonderful. I then found some useful functions in package languageR. But then the following problem ocurred: Whenever I load and use the languageR package, then save the workspace - or quit R with saving the workspace - I am unable to reload that workspace in a later session. That is, R doesn't start at all when I try to start it by clicking the workspace file. Loading languageR before loading the workspace doesn't help, but yields the message: Error in load(D:\\statistics\\MultilevelAnalysis\\.RData) : could not find function findPackageEnv Thus, the saved workspace remains inaccessible. I not 100% certain that languageR is the scapegoat, but my trial-and-error experiments indicate it is. My system is Win XP Home/Professional: sessionInfo() R version 2.5.0 Patched (2007-04-24 r41305) i386-pc-mingw32 locale: LC_COLLATE=German_Germany.1252;LC_CTYPE=German_Germany.1252;LC_MONETARY=German_Germany.1252;LC_NUMERIC=C;LC_TIME=German_Germany.1252 attached base packages: [1] splines stats graphics grDevices utils [6] datasets methods base other attached packages: languageR rpartMASS Designsurvival 0.23.1-367.2-342.0-12 2.31 Hmisc e1071 class cluster zipfR 3.3-21.5-167.2-341.11.7 0.6-0 lme4coda Matrix lattice 0.99875-10.11-2 0.99875-20.15-8 thanks for any helpful suggestions! best Rüdiger -- Hans-Rüdiger Pfister Professor of Business Psychology University of Lüneburg Faculty for Business Administration, Behavioral Sciences and Law Institute of Experimental Industrial Psychology (LueneLab) Wilschenbrucher Weg 84 D-21335 Lüneburg, Germany +49-(0)4131 677 7759 [EMAIL PROTECTED] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] how to compute a garch model with t innovations ?
Hi, dear R users, I'm a new user of R and especially in time series modelling. I would like to know how to compute arch/garch model using innovations that are not normally distributed (Student / swew-Student distriibuted). The *garch* function in the time series fits a Generalized Autoregressive Conditional Heteroscedastic GARCH(p, q) time series model to the data by computing the maximum-likelihood estimates of the conditionally normal model. What function/package should I use ? Thanks ! [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] R vs. Splus in Pharma/Devices Industry
I am happy to say nice things about odfWeave. Before it was released I was looking at the RTF spec to see if I could write an RTF driver for sweave. But the task was a bit daunting and I doubt that I would have had time for it. So, I was (and still am) exited when odfWeave came out. It is a big time saver for me and I have shown it to several other people as well. I have looked at ooconvert, but unfortunately the current version is limited to *nix, and I am currently in a MS windows world. I tried getting it to run under cygwin at one point, but did not succeed at the time. Once it is available more generally, I will just set up Makefiles so that I can automatically produce either .doc or .pdf documents from my templates and data. Thanks for the great package, -- Gregory (Greg) L. Snow Ph.D. Statistical Data Center Intermountain Healthcare [EMAIL PROTECTED] (801) 408-8111 -Original Message- From: Kuhn, Max [mailto:[EMAIL PROTECTED] Sent: Thursday, June 14, 2007 3:11 PM To: Greg Snow; [EMAIL PROTECTED]; r-help@stat.math.ethz.ch Subject: RE: [R] R vs. Splus in Pharma/Devices Industry Greg, Thanks for the kind words about odfWeave. These reports are usually put out as internal webpages for various people in the organization to look at, so we could either go the odfWeave approach and generate pdf files (not as automated as I would like) I agree that automating the conversion should be easier. My wish would be that the OO binary had a flag to convert from one format to another. On Linux, there is a bash script that uses the open office binaries to do the conversion at the command line by Nathan Coulter at http://sourceforge.net/projects/ooconvert/ Also, there is a Java class called jooconvert out there (if memory serves) that has similar functionality. Max -- LEGAL NOTICE Unless expressly stated otherwise, this message is confidential and may be privileged. It is intended for the addressee(s) only. Access to this E-mail by anyone else is unauthorized. If you are not an addressee, any disclosure or copying of the contents of this E-mail or any action taken (or not taken) in reliance on it is unauthorized and may be unlawful. If you are not an addressee, please inform the sender immediately. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] [OT] 'gv' and fractional points
This doesn't answer your original question, and isn't much help unless you're on a mac, but there's a nice looking program that makes this kind of graph scraping really easy: http://www.arizona-software.ch/applications/graphclick/en/ Hadley On 6/15/07, Ted Harding [EMAIL PROTECTED] wrote: Hi Folks, This is off-topic R-wise, but it may be close to the heart of many R-users, so I think it may be the best place to ask! Users of 'gv' (the front end to ghostscript) will be aware of the little window which gives you the x-y coordinates (in points = 1/72 inch) of the position of the cross-hair mouse cursor. These coordinates are those of the corresponding position on the printed page, relative to some origin. I have often used this to extract numerical values for data from graphs in Postscript files (also PDF files, after you have converted them to PS). Then (veering back on topic ... ) you can submit the numerical data to R and try your own analyses on these data, and compare with what the article does. However, this little window only gives the numbers in whole points. Say a smallish graphic may print out 3 inches wide or high. Then you get precision of 1/216 per 3 inches or 0.4% of full scale. This can be adequate on many occasions, but can be on the coarse side on other occasions. Even for a 6-inch-wide/high graph, you only get down to 0.2% of full scale. If it were possible to induce 'gv' to display these coordinates in tenths of a point, then much greater precision (as adequate as one can expect to hope for when, in effect, measuring off the graph) could be obtained. Does anyone know: a) Whether it is possible to persuade 'gv' to give this display in fractional points (my own search of the documentation has not revealed anything); b) Of any alternative to 'gv' as PS viewer which would provide this capability? With thanks, and best wishes to all, Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 16:13:21 -- XFMail -- __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] model.frame: how does one use it?
Philipp Benner reported a Debian bug report against r-cran-rpart aka rpart. In short, the issue has to do with how rpart evaluates a formula and supporting arguments, in particular 'weights'. A simple contrived example is - library(rpart) ## using data from help(rpart), set up simple example myformula - formula(Kyphosis ~ Age + Number + Start) mydata - kyphosis myweight - abs(rnorm(nrow(mydata))) goodFunction - function(mydata, myformula, myweight) { hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } goodFunction(mydata, myformula, myweight) cat(Ok\n) ## now remove myweight and try to compute it inside a function rm(myweight) badFunction - function(mydata, myformula) { myweight - abs(rnorm(nrow(mydata))) mf - model.frame(myformula, mydata, myweight) print(head(df)) hyp - rpart(myformula, data=mf, weights=myweight, method=class) prev - hyp } badFunction(mydata, myformula) cat(Done\n) - Here goodFunction works, but only because myweight (with useless random weights, but that is not the point here) is found from the calling environment. badFunction fails after we remove myweight from there: :~ cat /tmp/philipp.R | R --slave Ok Error in eval(expr, envir, enclos) : object myweight not found Execution halted :~ As I was able to replicate it, I reported this to the package maintainer. It turns out that seemingly all is well as this is supposed to work this way, and I got a friendly pointer to study model.frame and its help page. Now I am stuck as I can't make sense of model.frame -- see badFunction above. I would greatly appreciate any help in making rpart work with a local argument weights so that I can tell Philipp that there is no bug. :) Regards, Dirk -- Hell, there are no rules here - we're trying to accomplish something. -- Thomas A. Edison __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] subscript out of bounds error in lda
I work with Windows, R version 2.4.1 I'm trying to do a discriminant analysis and, in trying to figure out how to do it following the example from R help, I'm getting an error that says 'subscript out of bounds'. I don't know what this means and how to solve it (I'm very new with R) I'm doing everything in this made-up test matrix: group var1 var2 var3 1 13 556 2 14 667 3 15 558 4 14 667 5 13 446 6 14 555 7 25 889 8 24 998 9 28 889 10 29 768 11 28 669 12 29 99 10 13 2 10 1009 I write: data.tb-read.table('locationHere.txt', header=T) data.df-as.data.frame (data.tb) train-sample (1:63, 30) table (data.df$group[train]) data.disc-lda(group~., data.tb, subset = train) data.disc predict (data.disc, data.df[-train,])$class This is where I get the message: Error in `[.data.frame`(data.df, -train, ) : subscript out of bounds Can anyone, please help me figure out what this is about? Thanks! Silvia. -- View this message in context: http://www.nabble.com/subscript-out-of-bounds-error-in-lda-tf3928773.html#a11142304 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] A question about logical controls and function arguements
Dear R-help subscribers, I'm trying to write a function to generate data simulating the image created by a point radiation source in a plane on a screen where there is filter with a single circular aperture in it between the source and the screen. Following some guides (including Intro to R and some I found online) and examples I have specified the function (full code below question) with several arguments with the form: option=c(option1,option2) For instance, I want filter to either be FALSE to tell the function there is no filter or an ordered triplet describing the location and radius of the area radiation is not blocked by the filter. There are several others along similar lines. When I source the function into R, it parses fine, but when attempting to run it with data.spect-spect.data(source.p=r,filter=c(0,0,1),file.out=FALSE) the following warning is returned: Warning messages: 1: the condition has length 1 and only the first element will be used in: if (filter == FALSE) { 2: the condition has length 1 and only the first element will be used in: if (filter == FALSE) { The code this is referencing is about 1/3 from the bottom of the function. I'm not sure how to correct this. I tried ifelse in one case and it doesn't work at all. Searching the archives for function arguments didn't yield anything about the kind of arguments that are causing the trouble. I also want to get the matrix of generated data out, and have tried data.spect$final.sample (following an example I found online), but it returns null. I also attempted to use data.spect$initial.sample, but this returned null as well. I'm still very new to writing my own functions, and any and all help would be appreciated. There are notes about what different options are supposed to do at the end of the appended code. Thanks in advance, Jason Q McClintic -- Jason Q McClintic [EMAIL PROTECTED] [EMAIL PROTECTED] spect.data-function(num.points=50,fixed=FALSE,source.p=c(r,c(0,0)), source.mean=0,source.sd=1,filter=c(FALSE,c(0,0,1)), heights=c(0.5,0.5), file.out=c(FALSE,/home/jqmcclintic/Desktop/spect.data)){ ##Determine Start Point if (source.p==r) {source-c(rnorm(1,source.mean,source.sd),rnorm(1,source.mean,source.sd))} else {source-source.p} cat(The location of the source is: ,source,\n) ##Generate the data remainder-num.points initial.sample-c(1,1) ##finds intersection points with the screen intersect.screen.at-function(x,h){ t-h[1]/(2*cos(x)) x.intercept-t*sin(x[,2])*cos(x[,1]) y.intercept-t*sin(x[,2])*sin(x[,1]) } ##finds intersection points with the collecting plate intersect.plate.at-function(x,h){ t-h[2]/(2*cos(x)) x.intercept-t*sin(x[,2])*cos(x[,1]) y.intercept-t*sin(x[,2])*sin(x[,1]) } ##determines if the intersection point is inside or outside the hole in the screen. x is the matrix of intersection points and s is the location and radius of the hole in the screen. 1 for yes, 0 for no. passes.through-function(x,s){ distance-sqrt(((x[,1]-s[1])^2)+((x[,2]-s[2])^2)) through-ifelse(distances[3],1,0) } ##Build the sample while (remainder0){ ##Generate n random vectors uniformly distributed over S2 theta-runif(remainder,0,6.2831853) phi-runif(remainder,0,1.5707963) theta.phi-cbind(theta,phi) initial.sample-rbind(initial.sample,theta.phi) ##Call intersect.screen.at intersects.screen-intersect.screen.at(initial.sample,heights) ##Call intersect.plate.at intersects.plate-if(filter==FALSE) {intersect.screen.at(initial.sample,heights)} else { intersect.plate.at(initial.sample,heights) } ##Does it intersect inside or outside the hole? intersect.hole-if(filter==FALSE){array(1,dim=length(initial.sample))} else{passes.through(intersects.screen,filter)} ##Remove points that do not pass throught the hole. By design, if there is no filter, all pass through the hole. initial.sample-cbind(initial.sample,intersect.hole) initial.sample-subset(initial.sample,initial.sample[,3]==1) ##Reset remainder remainder-if(fixed==FALSE) {0} else { num.points-length(initial.sample) } } write(initial.sample) ##remove the top row of the initial sample since it is non-random. final.sample-initial.sample[-1,] ##print the final sample to a csv file for archival purposes
Re: [R] Problem with workspace loading after languageR use
Thank you for the example: it has enabled me to test a more comprehensive fix for R-patched (soon to be 2.5.1). You should have got a warning when saving along the lines of Warning message: 'package:languageR' may not be available when loading in: save(list = ls(envir = .GlobalEnv, all.names = TRUE), file = outfile, The problem is indeed in package languageR, which is creating classes and methods in your workspace rather than in its package. It really should not be doing so (it uses setMethod in .First.lib), and I am Cc:ing the maintainer in the hope that he will stop doing so. I also find it hard to believe that Depends: R (= 2.4.0), methods, lattice, Matrix, coda, lme4, zipfR, cluster, e1071, Design, Hmisc, MASS, rpart is totally necessary, and suggest that Suggests: is used. On Fri, 15 Jun 2007, Pfister wrote: Dear Brian, thanks a lot for your advice. Prof Brian Ripley wrote: On Fri, 15 Jun 2007, Prof Brian Ripley wrote: I think adding the following to your new session before load() will help findPackageEnv - function(info) as.environment(paste(package, info, sep=:)) this does not work. more likely findPackageEnv - function(info) as.environment(info) this works, if I load the languageR library before. is correct. will work, but if not try findPackageEnv - function(info) .GlobalEnv this works unconditionally. Including that line in a local .RProfile file, basically solves the problem. If you send me the problematic workspace (or reproduction instructions) I can take a closer look. A typical session would look like this (using example data from Hox(2002)): # from Hox (2002) hoxpop - read.table(http://www.ruediger-pfister.de/download/popular.dat;, header=TRUE) hoxpop$PUPIL - factor(hoxpop$PUPIL) hoxpop$SCHOOL - factor(hoxpop$SCHOOL) hoxpop$SEX - factor(hoxpop$SEX) # load languageR library(languageR) # do some analyses ... # save the workspace save.image(D:\\statistics\\MultilevelAnalysis\\lm.RData) After quitting R, the workspace lm.RData will not reload. (this workspace can be downloaded here: http://www.ruediger-pfister.de/Downloads/lmRData.zip) best Rüdiger On Fri, 15 Jun 2007, Pfister wrote: Hello R, To analyze multi-level data, I started learning and using lmer. So far so wonderful. I then found some useful functions in package languageR. But then the following problem ocurred: Whenever I load and use the languageR package, then save the workspace - or quit R with saving the workspace - I am unable to reload that workspace in a later session. That is, R doesn't start at all when I try to start it by clicking the workspace file. Loading languageR before loading the workspace doesn't help, but yields the message: Error in load(D:\\statistics\\MultilevelAnalysis\\.RData) : could not find function findPackageEnv Thus, the saved workspace remains inaccessible. I not 100% certain that languageR is the scapegoat, but my trial-and-error experiments indicate it is. My system is Win XP Home/Professional: sessionInfo() R version 2.5.0 Patched (2007-04-24 r41305) i386-pc-mingw32 locale: LC_COLLATE=German_Germany.1252;LC_CTYPE=German_Germany.1252;LC_MONETARY=German_Germany.1252;LC_NUMERIC=C;LC_TIME=German_Germany.1252 attached base packages: [1] splines stats graphics grDevices utils [6] datasets methods base other attached packages: languageR rpartMASS Designsurvival 0.23.1-367.2-342.0-12 2.31 Hmisc e1071 class cluster zipfR 3.3-21.5-167.2-341.11.7 0.6-0 lme4coda Matrix lattice 0.99875-10.11-2 0.99875-20.15-8 thanks for any helpful suggestions! best Rüdiger -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595__ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] A question about logical controls and function arguements
Sir, I freely admit my ignorance as to the subtleties of specifying arguments. Let me make sure I understand your suggestion: create a variable filter which takes 1 or 0 (filter or not) and another called, say, filter.location which is the ordered triplet. It does add to the number of options, but would seem to simplify the underlying code. It appears I may have attempted to code above my skill level. Thanks for the assistance, Jason Q McClintic -- Jason Q McClintic [EMAIL PROTECTED] [EMAIL PROTECTED] jim holtman wrote: You are trying to use 'filter' in two ways. Your code is testing for a single value (FALSE), and that is all that == can do (single value), but you are pass in a vector (c(0,0,1)) which has three values, on the first of which can be tested by the ==. So you might want to consider having another parameter which says whether or not to use filter. On 6/15/07, *Jason Q McClintic* [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: Dear R-help subscribers, I'm trying to write a function to generate data simulating the image created by a point radiation source in a plane on a screen where there is filter with a single circular aperture in it between the source and the screen. Following some guides (including Intro to R and some I found online) and examples I have specified the function (full code below question) with several arguments with the form: option=c(option1,option2) For instance, I want filter to either be FALSE to tell the function there is no filter or an ordered triplet describing the location and radius of the area radiation is not blocked by the filter. There are several others along similar lines. When I source the function into R, it parses fine, but when attempting to run it with data.spect-spect.data(source.p=r,filter=c(0,0,1),file.out=FALSE) the following warning is returned: Warning messages: 1: the condition has length 1 and only the first element will be used . in: if (filter == FALSE) { . 2: the condition has length 1 and only the first element will be used in: if (filter == FALSE) { . The code this is referencing is about 1/3 from the bottom of the function. I'm not sure how to correct this. I tried ifelse in one case and it doesn't work at all. Searching the archives for function arguments didn't yield anything about the kind of arguments that are causing the trouble. I also want to get the matrix of generated data out, and have tried data.spect$final.sample (following an example I found online), but it returns null. I also attempted to use data.spect$initial.sample , but this returned null as well. I'm still very new to writing my own functions, and any and all help would be appreciated. There are notes about what different options are supposed to do at the . end of the appended code. Thanks in advance, Jason Q McClintic -- Jason Q McClintic [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] spect.data-function(num.points=50,fixed=FALSE,source.p=c(r,c(0,0)), source.mean=0,source.sd=1,filter=c(FALSE,c(0,0,1)), heights=c(0.5,0.5), file.out=c (FALSE,/home/jqmcclintic/Desktop/spect.data)){ ##Determine Start Point if (source.p==r) {source-c(rnorm(1,source.mean,source.sd http://source.sd ),rnorm(1,source.mean,source.sd http://source.sd))} else {source-source.p} cat(The location of the source is: ,source,\n) ##Generate the data remainder- num.points initial.sample-c(1,1) ##finds intersection points with the screen intersect.screen.at http://intersect.screen.at-function(x,h){ t-h[1]/(2*cos(x)) x.intercept-t*sin(x[,2])*cos(x[,1]) y.intercept-t*sin(x[,2])*sin(x[,1]) } ##finds intersection points with the collecting plate intersect.plate.at http://intersect.plate.at-function(x,h){ t-h[2]/(2*cos(x)) x.intercept-t*sin(x[,2])*cos(x[,1]) y.intercept -t*sin(x[,2])*sin(x[,1]) .} ##determines if the intersection point is inside or outside the hole in the screen. x is the matrix of intersection points and s is the location and radius of the hole in the screen. 1 for yes, 0 for no. passes.through-function(x,s){ distance-sqrt(((x[,1]-s[1])^2)+((x[,2]-s[2])^2)) through-ifelse(distances[3],1,0) } ##Build the sample while (remainder0){ ##Generate n
[R] Registration density profiles using FDA (Functional Data Analysis)
Hello, I would like to use fda (Functional data analysis) library to compare sampled curves that are density profiles. But I encounter some problems to register my data as functional data. I know that I must use a registration function to align functions in order to have important features found in each curve occur at roughly the same argument value, prior to do subsequent analysis ( PCA and cluster analysis). But I do not know which function (if there exists one) allows me to synchronize density profiles : could help me please ? Thank you by advance Best regards Lise Bellanger -- Lise Bellanger, Université de Nantes Département de Mathématiques, Laboratoire Jean Leray UMR CNRS 6629 2, Rue de la Houssinière BP 92208 - F-44322 Nantes Cedex 03 Tél. : (33|0) 2 51 12 59 00 (ou 43) - Fax : (33|0) 2 51 12 59 12 E-Mail : [EMAIL PROTECTED] URL : http://www.math.sciences.univ-nantes.fr/ __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] subscript out of bounds error in lda
On Fri, 15 Jun 2007, Silvia Lomascolo wrote: I work with Windows, R version 2.4.1 I'm trying to do a discriminant analysis and, in trying to figure out how to do it following the example from R help, I'm getting an error that says 'subscript out of bounds'. I don't know what this means and how to solve it (I'm very new with R) I'm doing everything in this made-up test matrix: group var1 var2 var3 1 13 556 2 14 667 3 15 558 4 14 667 5 13 446 6 14 555 7 25 889 8 24 998 9 28 889 10 29 768 11 28 669 12 29 99 10 13 2 10 1009 I write: data.tb-read.table('locationHere.txt', header=T) data.df-as.data.frame (data.tb) Wny call as.data.frame on a data frame? train-sample (1:63, 30) Why sample from 1:63 with 13 rows? table (data.df$group[train]) data.disc-lda(group~., data.tb, subset = train) data.disc predict (data.disc, data.df[-train,])$class This is where I get the message: Error in `[.data.frame`(data.df, -train, ) : subscript out of bounds traceback() is your friend: traceback() 8: `[.data.frame`(data.df, -train, ) 7: data.df[-train, ] 6: inherits(x, data.frame) 5: is.data.frame(data) 4: model.frame.default(Terms, newdata, na.action = na.pass, xlev = object$xlevels) 3: model.frame(Terms, newdata, na.action = na.pass, xlev = object$xlevels) 2: predict.lda(data.disc, data.df[-train, ]) 1: predict(data.disc, data.df[-train, ]) So, let's take a look at the top lines. train [1] 46 42 30 13 27 63 19 47 3 52 62 16 26 4 61 23 59 44 40 38 25 55 50 10 43 [26] 2 8 31 7 11 nrow(data.df) [1] 13 So, you are asking for number 46 out of 13 rows. Now perhaps you didn't show us all the problem, but hopefully this helps you find the error. If not, the bottom of every R-help message says 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, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] problems with matrix, list and other..
thank you! it works with: E=list() length(E)=1096 for(i in (2:1096)){E[[i]]=crossprod(solve(d[[i]]),cbind(e[[i]]))} Petr Klasterecky wrote: You only got what you deserved when not reading the manual... R-Intro, Chapters 5 and 6, page 26 in particular. http://cran.r-project.org/doc/manuals/R-intro.pdf Petr billycorg napsal(a): hi can anyone help me to solve these problems? i have: 1) d matrix with 1096 rows; for example, d[2]= [,1] [,2] [,3] [1,] 0.1192566 0.000 0.000 [2,] 0.000 0.1065938 0.000 [3,] 0.000 0.000 0.103 if I class (d[2]) = list solve(d[2]) = error!!! 2) e list with 1096 rows; for example e[2]2= [[1]] [1] -1.0892216 -0.7304947 -1.2883680 d[2]%*%t(e[2]) this is the error: requires numeric matrix/vector arguments i've tried to coerce e to a matrix, but it's doesn't work... in the end.. i'd like this: for (i in (1:1096)) {solve(d[i])*t(e[i])} help me, please :) Vincenzo -- Petr Klasterecky Dept. of Probability and Statistics Charles University in Prague Czech Republic __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- View this message in context: http://www.nabble.com/problems-with-matrix%2C-list-and-other..-tf3926701.html#a11144182 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] text display using expression or not
Hello, I imagine that I'm missing something straightforward, but a run thru the help files didn't turn up an answer. I noticed while formatting some figures for publication that text enclosed in expression() and used for a title displays differently than a string, regardless of the par options. On both postscript() and x11() devices, the regular text is heavier than the expression text. Here's an example. The real thing used expression() to produce superscripts, but I boiled this down to the simplest possible case. par(mfrow=c(2,1)) plot(1:10, 1:10, main=Figure A) plot(1:10, 1:10, main=expression(Figure B)) Is there some straightforward way to make them match, other than putting expression() around all strings? I'm currently using R 2.5.0 on Fedora core 5. Thanks, Sarah -- Sarah Goslee http://www.functionaldiversity.org __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] complex contrasts and logistic regression
Hi, I am doing a retrospective analysis on a cohort from a designed trial, and I am fitting the model fit-glmD(survived ~ Covariate*Therapy + confounder,myDat,X=TRUE, Y=TRUE, family=binomial()) My covariate has three levels (A,B and C) and therapy has two (treated and control), confounder is a continuous variable. Also patients were randomized to treatment in the trial, but Covariate is something that is measured posthoc and can vary in the population. I am trying to wrap my head around how to calculate a few quantities from the model and get reasonable confidence intervals for them, namely I would like to test H0: gamma=0, where gamma is the regression coefficient of the odds ratios of surviving under treatment vs control at each level of Covariate (adjusted for the confounder) and I would like to get the odds of surviving at each level of Covariate under treatment and control for each level of covariate adjusted for the confounder. I have looked at contrast in the Design library but I don't think it gives me the right quantity, for instance contrast(fit,list(covariate=A, Therapy=Treated, confounder=median(myDat$confounder), X=TRUE) ( A is the baseline level of Covariate) gives me beta0 + beta_Treated + beta_confounder*68 Is this correctly interpreted as the conditional odds of dying? As to the 1st contrast I am not sure how to get it, would it be using type = 'average' with some weights in contrast? The answers are probably staring me in the face, i am just not seeing them today. Nicholas __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] R vs. Splus in Pharma/Devices Industry
Greg, I have looked at ooconvert, but unfortunately the current version is limited to *nix, and I am currently in a MS windows world. I tried getting it to run under cygwin at one point, but did not succeed at the time. Once it is available more generally, I will just set up Makefiles so that I can automatically produce either .doc or .pdf documents from my templates and data. I found the Java tool at http://www.artofsolving.com/opensource/jodconverter (they changed the name). I just tried it on my Windows machine and it worked. You have to start a headless soffice process (see the readme file) and then it is a simple command line away. Max -- LEGAL NOTICE\ Unless expressly stated otherwise, this messag...{{dropped}} __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] removing values from a vector, where both the value and its name are the same?
I have an array such as: x=c(sum=77, min=4,max=9, count=5, min=4,max=9, count=8 , test=77) I wish to remove values where both the name and the value are identical. eg. i wish to end up with: x2=c(sum=77, min=4,max=9, count=5, count=8, test=77) What is the best way to do this? - Park yourself in front of a world of choices in alternative vehicles. [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] [OT] 'gv' and fractional points
On 6/15/07, Ted Harding [EMAIL PROTECTED] wrote: Hi Folks, This is off-topic R-wise, but it may be close to the heart of many R-users, so I think it may be the best place to ask! Users of 'gv' (the front end to ghostscript) will be aware of the little window which gives you the x-y coordinates (in points = 1/72 inch) of the position of the cross-hair mouse cursor. These coordinates are those of the corresponding position on the printed page, relative to some origin. I have often used this to extract numerical values for data from graphs in Postscript files (also PDF files, after you have converted them to PS). Then (veering back on topic ... ) you can submit the numerical data to R and try your own analyses on these data, and compare with what the article does. However, this little window only gives the numbers in whole points. Say a smallish graphic may print out 3 inches wide or high. Then you get precision of 1/216 per 3 inches or 0.4% of full scale. This can be adequate on many occasions, but can be on the coarse side on other occasions. If you are mostly concerned about small figures, one possibility is 1. zoom out to a level where you're happy with the pixel resolution 2. do a screen capture using 'import' 3. use gimp (which has the same feature, with more units) gimp can also load PS files directly, with a user supplied zoom factor at load time, but only one page at a time, AFAICT. -Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] removing values from a vector, where both the value and its name are the same?
try this: x[!(duplicated(names(x))duplicated(x))] sum min max count count test 77 4 9 5 877 On 6/15/07, new ruser [EMAIL PROTECTED] wrote: I have an array such as: x=c(sum=77, min=4,max=9, count=5, min=4,max=9, count=8 , test=77) I wish to remove values where both the name and the value are identical. eg. i wish to end up with: x2=c(sum=77, min=4,max=9, count=5, count=8, test=77) What is the best way to do this? - Park yourself in front of a world of choices in alternative vehicles. [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Jim Holtman Cincinnati, OH +1 513 646 9390 What is the problem you are trying to solve? [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] removing values from a vector, where both the value and its name are the same?
On Fri, 2007-06-15 at 11:19 -0700, new ruser wrote: I have an array such as: x=c(sum=77, min=4,max=9, count=5, min=4,max=9, count=8 , test=77) I wish to remove values where both the name and the value are identical. eg. i wish to end up with: x2=c(sum=77, min=4,max=9, count=5, count=8, test=77) What is the best way to do this? Not sure if this is the best way, but since you need to compare both the values and their name attributes for 'uniqueness': x[!(duplicated(x) duplicated(names(x)))] sum min max count count test 77 4 9 5 877 What is being done is to first compare the values for duplicates. Note that the second value is identified as the duplicate, not the first: duplicated(x) [1] FALSE FALSE FALSE FALSE TRUE TRUE FALSE TRUE and then compare the names for duplicates: duplicated(names(x)) [1] FALSE FALSE FALSE FALSE TRUE TRUE TRUE FALSE Then compare the two logical vectors and get the indices where BOTH are TRUE: (duplicated(x) duplicated(names(x))) [1] FALSE FALSE FALSE FALSE TRUE TRUE FALSE FALSE Now, negate the relation: !(duplicated(x) duplicated(names(x))) [1] TRUE TRUE TRUE TRUE FALSE FALSE TRUE TRUE and return the values in 'x' that satisfy the logical indices: x[!(duplicated(x) duplicated(names(x)))] sum min max count count test 77 4 9 5 877 See ?duplicated. You might also want to look at ?unique, ?identical, ?all.equal and ?isTRUE. Note that the above example will likely fail if any of the values are floats: duplicated(c(0.5 - 0.4, 0.1)) [1] FALSE FALSE in which case, you would need to use a looping structure where the value comparisons use isTrue(all.equal(...)) instead. HTH, Marc Schwartz __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] R vs. Splus in Pharma/Devices Industry
Hi, I just saw this thread. This issue, and the larger scale issue of open source in industry is being addressed. One has to realize that the behemoth that is the clinical aperatus of the pharma industry is very conservative and very slow to change. In many cases switching to R would meean changing a great many processes all based on legacy code. One of the big issues is that the industry demands consistency not necessarily correctness. All that said there are a great many areas where R could be used that would not impact regulatory submission, data security etc. Many in pharma are quietly working on this, but steps are small and incremental. Development takes time in industry because of the amount of documentation necessary. All this impacts the free nature of R and cost and risk (From the industries perspective) need to be justified. Probably when the statistical community is using Z big pharma will be ready to use R. %P Nicholas __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] how to plot two graphics in one window
Hello , Maybe this question you answered before, but i couldnt find something indicated in the mailing list. I wish to plot two graphics in one window, for example y=sinx and y=exp(x) in the same windows, (the same interval for x). Thanks . Miguel. -- View this message in context: http://www.nabble.com/how-to-plot-two-graphics-in-one-window-tf3929594.html#a11145186 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] model.frame: how does one use it?
On 6/15/07, Dirk Eddelbuettel [EMAIL PROTECTED] wrote: Philipp Benner reported a Debian bug report against r-cran-rpart aka rpart. In short, the issue has to do with how rpart evaluates a formula and supporting arguments, in particular 'weights'. A simple contrived example is - library(rpart) ## using data from help(rpart), set up simple example myformula - formula(Kyphosis ~ Age + Number + Start) mydata - kyphosis myweight - abs(rnorm(nrow(mydata))) goodFunction - function(mydata, myformula, myweight) { hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } goodFunction(mydata, myformula, myweight) cat(Ok\n) ## now remove myweight and try to compute it inside a function rm(myweight) badFunction - function(mydata, myformula) { myweight - abs(rnorm(nrow(mydata))) mf - model.frame(myformula, mydata, myweight) print(head(df)) hyp - rpart(myformula, data=mf, weights=myweight, method=class) prev - hyp } badFunction(mydata, myformula) cat(Done\n) - Here goodFunction works, but only because myweight (with useless random weights, but that is not the point here) is found from the calling environment. badFunction fails after we remove myweight from there: :~ cat /tmp/philipp.R | R --slave Ok Error in eval(expr, envir, enclos) : object myweight not found Execution halted :~ As I was able to replicate it, I reported this to the package maintainer. It turns out that seemingly all is well as this is supposed to work this way, and I got a friendly pointer to study model.frame and its help page. Now I am stuck as I can't make sense of model.frame -- see badFunction above. I would greatly appreciate any help in making rpart work with a local argument weights so that I can tell Philipp that there is no bug. :) I don't know if ?model.frame is the best place page to look. There's a more detailed description at http://developer.r-project.org/nonstandard-eval.pdf but here are the non-standard evaluation rules as I understand them: given a name in either (1) the formula or (2) ``special'' arguments like 'weights' in this case, or 'subset', try to find the name 1. in 'data' 2. failing that, in environment(formula) 3. failing that, in the enclosing environment, and so on. By 'name', I mean a symbol, such as 'Age' or 'myweight'. So basically, everything is as you would expect if the name is visible in data, but if not, the search starts in the environment of the formula, not the environment where the function call is being made (which is the standard evaulation behaviour). This is a feature, not a bug (things would be a lot more confusing if it were the other way round). With this in mind, either of the following might do what you want: badFunction - function(mydata, myformula) { mydata$myweight - abs(rnorm(nrow(mydata))) hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } badFunction - function(mydata, myformula) { myweight - abs(rnorm(nrow(mydata))) environment(myformula) - environment() hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } -Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] R vs. Splus in Pharma/Devices Industry
Nicholas, You're making me depressed! Cody Hamilton, PhD Edwards Lifesciences Hi, I just saw this thread. This issue, and the larger scale issue of open source in industry is being addressed. One has to realize that the behemoth that is the clinical aperatus of the pharma industry is very conservative and very slow to change. In many cases switching to R would meean changing a great many processes all based on legacy code. One of the big issues is that the industry demands consistency not necessarily correctness. All that said there are a great many areas where R could be used that would not impact regulatory submission, data security etc. Many in pharma are quietly working on this, but steps are small and incremental. Development takes time in industry because of the amount of documentation necessary. All this impacts the free nature of R and cost and risk (From the industries perspective) need to be justified. Probably when the statistical community is using Z big pharma will be ready to use R. %P Nicholas __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] removing values from a vector, where both the value and its name are the same?
In case it matters, the given solution has a problem if the data look like: x - c(sum=77, test=99, sum=99) By the description all three elements should be kept, but the duplicated solution throws out the last element. A more complicated solution is: unique(data.frame(x, names(x))) (and then put the vector back together again). Patrick Burns [EMAIL PROTECTED] +44 (0)20 8525 0696 http://www.burns-stat.com (home of S Poetry and A Guide for the Unwilling S User) jim holtman wrote: try this: x[!(duplicated(names(x))duplicated(x))] sum min max count count test 77 4 9 5 877 On 6/15/07, new ruser [EMAIL PROTECTED] wrote: I have an array such as: x=c(sum=77, min=4,max=9, count=5, min=4,max=9, count=8 , test=77) I wish to remove values where both the name and the value are identical. eg. i wish to end up with: x2=c(sum=77, min=4,max=9, count=5, count=8, test=77) What is the best way to do this? - Park yourself in front of a world of choices in alternative vehicles. [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] R vs. Splus in Pharma/Devices Industry
On 6/15/07, Nicholas Lewin-Koh [EMAIL PROTECTED] wrote: Hi, I just saw this thread. This issue, and the larger scale issue of open source in industry is being addressed. One has to realize that the behemoth that is the clinical aperatus of the pharma industry is very conservative and very slow to change. In many cases switching to R would meean changing a great many processes all based on legacy code. One of the big issues is that the industry demands consistency not necessarily correctness. All that said there are a great many areas where R could be used that would not impact regulatory submission, data security etc. Many in pharma are quietly working on this, but steps are small and incremental. Development takes time in industry because of the amount of documentation necessary. All this impacts the free nature of R and cost and risk (From the industries perspective) need to be justified. Probably when the statistical community is using Z big pharma will be ready to use R. %P I think you mean A, not Z. First there was S, then there was R. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] model.frame: how does one use it?
On Fri, 2007-06-15 at 10:47 -0500, Dirk Eddelbuettel wrote: Philipp Benner reported a Debian bug report against r-cran-rpart aka rpart. In short, the issue has to do with how rpart evaluates a formula and supporting arguments, in particular 'weights'. A simple contrived example is - library(rpart) ## using data from help(rpart), set up simple example myformula - formula(Kyphosis ~ Age + Number + Start) mydata - kyphosis myweight - abs(rnorm(nrow(mydata))) goodFunction - function(mydata, myformula, myweight) { hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } goodFunction(mydata, myformula, myweight) cat(Ok\n) ## now remove myweight and try to compute it inside a function rm(myweight) badFunction - function(mydata, myformula) { myweight - abs(rnorm(nrow(mydata))) mf - model.frame(myformula, mydata, myweight) print(head(df)) hyp - rpart(myformula, data=mf, weights=myweight, method=class) prev - hyp } badFunction(mydata, myformula) cat(Done\n) - Here goodFunction works, but only because myweight (with useless random weights, but that is not the point here) is found from the calling environment. badFunction fails after we remove myweight from there: :~ cat /tmp/philipp.R | R --slave Ok Error in eval(expr, envir, enclos) : object myweight not found Execution halted :~ As I was able to replicate it, I reported this to the package maintainer. It turns out that seemingly all is well as this is supposed to work this way, and I got a friendly pointer to study model.frame and its help page. Now I am stuck as I can't make sense of model.frame -- see badFunction above. I would greatly appreciate any help in making rpart work with a local argument weights so that I can tell Philipp that there is no bug. :) Regards, Dirk Dirk, As you note, the issue is the non-standard evaluation of the arguments in model.frame() The key section of the Details in ?model.frame is: All the variables in formula, subset and in ... are looked for first in data and then in the environment of formula (see the help for formula() for further details) and collected into a data frame. Then the subset expression is evaluated, and it is is used as a row index to the data frame. Then the na.action function is applied to the data frame (and may well add attributes). The levels of any factors in the data frame are adjusted according to the drop.unused.levels and xlev arguments. Note that even with your goodFunction(), if 'myweight' is created within the environment of the function and not in the global environment, it still fails: library(rpart) myformula - formula(Kyphosis ~ Age + Number + Start) mydata - kyphosis goodFunction - function(mydata, myformula) { myweight - abs(rnorm(nrow(mydata))) hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } goodFunction(mydata, myformula) Error in eval(expr, envir, enclos) : object myweight not found However, now let's do this: library(rpart) myformula - formula(Kyphosis ~ Age + Number + Start) mydata - kyphosis myweight - abs(rnorm(nrow(mydata))) goodFunction - function(mydata, myformula) { hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } goodFunction(mydata, myformula) It works, because 'myweight' is found in the global environment, which is where the formula is created. Now, final example, try this: library(rpart) goodFunction - function() { myformula - formula(Kyphosis ~ Age + Number + Start) mydata - kyphosis myweight - abs(rnorm(nrow(mydata))) hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } goodFunction() It works because the formula is created within the environment of the function and hence, 'myweight', which is created there as well, is found. There was a (non) bug filed on a related matter dealing with the evaluation of 'subset': http://bugs.r-project.org/cgi-bin/R/feature%26FAQ?id=3671 and you might find this document on Non-Standard Evaluation helpful: http://developer.r-project.org/nonstandard-eval.pdf HTH, Marc __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide
Re: [R] [OT] 'gv' and fractional points
On 15-Jun-07 16:29:53, Ted Harding wrote: [...] However, as a follow-up, I've since found that one can (somewhat tediously) do what I was asking with the GIMP. As well as the awkwardness of doing it the GIMP way, I've discovered another disadvantage. I'd previously tried it on a rather small image (175x70 points, = 2.43x0.97 inches). I then tried it on a full A4 page. Even at a GIMP Scale factor of 300, this leads to a 50MB temporary file being created. At 1000, this would rise to some 550MB, as I found out after this attempt filled up the limited spare space I have on the disk drive in question ... No doubt Scale=300 (as opposed to the default of 100) may be ample for most purposes, but the overhead is still unpleasant! Hence I'm once again hankering after something which will display a PS file as efficiently as 'gv', but will report the cursor position in fractions of a point! Best wishes to all, Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 19:18:48 -- XFMail -- E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 20:33:19 -- XFMail -- __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] how to plot two graphics in one window
?points x - 1:10 plot(exp(x),col=red, type =o) points(sin(x), col=blue) --- Miguel Caro [EMAIL PROTECTED] wrote: Hello , Maybe this question you answered before, but i couldnt find something indicated in the mailing list. I wish to plot two graphics in one window, for example y=sinx and y=exp(x) in the same windows, (the same interval for x). Thanks . Miguel. -- View this message in context: http://www.nabble.com/how-to-plot-two-graphics-in-one-window-tf3929594.html#a11145186 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] how to plot two graphics in one window
Hi, I answer myself plot(x,y,type=l) par(new=TRUE) plot(x,yy,type=l) Thanks all! Miguel Caro wrote: Hello , Maybe this question you answered before, but i couldnt find something indicated in the mailing list. I wish to plot two graphics in one window, for example y=sinx and y=exp(x) in the same windows, (the same interval for x). Thanks . Miguel. -- View this message in context: http://www.nabble.com/how-to-plot-two-graphics-in-one-window-tf3929594.html#a11146479 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] model.frame: how does one use it?
On 6/15/07, Deepayan Sarkar [EMAIL PROTECTED] wrote: On 6/15/07, Dirk Eddelbuettel [EMAIL PROTECTED] wrote: Philipp Benner reported a Debian bug report against r-cran-rpart aka rpart. In short, the issue has to do with how rpart evaluates a formula and supporting arguments, in particular 'weights'. A simple contrived example is - library(rpart) ## using data from help(rpart), set up simple example myformula - formula(Kyphosis ~ Age + Number + Start) mydata - kyphosis myweight - abs(rnorm(nrow(mydata))) goodFunction - function(mydata, myformula, myweight) { hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } goodFunction(mydata, myformula, myweight) cat(Ok\n) ## now remove myweight and try to compute it inside a function rm(myweight) badFunction - function(mydata, myformula) { myweight - abs(rnorm(nrow(mydata))) mf - model.frame(myformula, mydata, myweight) print(head(df)) hyp - rpart(myformula, data=mf, weights=myweight, method=class) prev - hyp } badFunction(mydata, myformula) cat(Done\n) - Here goodFunction works, but only because myweight (with useless random weights, but that is not the point here) is found from the calling environment. badFunction fails after we remove myweight from there: :~ cat /tmp/philipp.R | R --slave Ok Error in eval(expr, envir, enclos) : object myweight not found Execution halted :~ As I was able to replicate it, I reported this to the package maintainer. It turns out that seemingly all is well as this is supposed to work this way, and I got a friendly pointer to study model.frame and its help page. Now I am stuck as I can't make sense of model.frame -- see badFunction above. I would greatly appreciate any help in making rpart work with a local argument weights so that I can tell Philipp that there is no bug. :) I don't know if ?model.frame is the best place page to look. There's a more detailed description at http://developer.r-project.org/nonstandard-eval.pdf but here are the non-standard evaluation rules as I understand them: given a name in either (1) the formula or (2) ``special'' arguments like 'weights' in this case, or 'subset', try to find the name 1. in 'data' 2. failing that, in environment(formula) 3. failing that, in the enclosing environment, and so on. By 'name', I mean a symbol, such as 'Age' or 'myweight'. So basically, everything is as you would expect if the name is visible in data, but if not, the search starts in the environment of the formula, not the environment where the function call is being made (which is the standard evaulation behaviour). This is a feature, not a bug (things would be a lot more confusing if it were the other way round). Could you give an example? It's always seemed confusing to me and I don't see why looking in the environment of the formula helps. Hadley __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] [OT] 'gv' and fractional points
On 15-Jun-07 15:56:58, hadley wickham wrote: This doesn't answer your original question, and isn't much help unless you're on a mac, but there's a nice looking program that makes this kind of graph scraping really easy: http://www.arizona-software.ch/applications/graphclick/en/ Hadley Thanks, Hadley! But (as you implicitly surmise) I don't use a Mac (just non-psychedelic Linux). However, as a follow-up, I've since found that one can (somewhat tediously) do what I was asking with the GIMP. If you start the GIMP on a PostScript file, you initially get a Load PostScript window which asks you to choose (amongst other things the resolution, initially 100. If you wind this up to say 1000, then you get 10 times the positional precision in the numbers shown as below. When the image is displayed, there is again a little window which gives you the GIMP coordinates of the mouse position. With increased Resolution, the numerical values vary correspondingly more rapidly with position. The tedious aspect (compared with 'gv') is that to get better visual resolution you need to zoom the image. This enlarges the whole image, with the result that you have to pan around to locate different parts, and you can get lost in a graph with lots of widely-spread points. With 'gv', on the other hand, you can select any small part to view in zoom in a sub-window, which is much easier to cope with; and you also get a rubber-band rectangle which enables you to readily align points here with points there -- e.g. if you want to read off a curve the y-value corresponding to say x=5.0. However, this is at least a partial answer to my own question! Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 17:29:49 -- XFMail -- E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 20:32:59 -- XFMail -- __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] how to plot two graphics in one window
I hope this could help: http://tolstoy.newcastle.edu.au/R/e2/help/07/02/11127.html Alex- On 6/15/07, Miguel Caro [EMAIL PROTECTED] wrote: Hello , Maybe this question you answered before, but i couldnt find something indicated in the mailing list. I wish to plot two graphics in one window, for example y=sinx and y=exp(x) in the same windows, (the same interval for x). Thanks . Miguel. -- View this message in context: http://www.nabble.com/how-to-plot-two-graphics-in-one-window-tf3929594.html#a11145186 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Medical University of South Carolina [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] text display using expression or not
Hi Sarah, You will kick yourself (perhaps?). Try: ?par ## sub: font.main ( font, just above) ## Example par(mfrow=c(2,1)) plot(1:10, 1:10, main=Figure A, font.main=1) plot(1:10, 1:10, main=expression(Figure B)) Regards, Mark. Sarah Goslee wrote: Hello, I imagine that I'm missing something straightforward, but a run thru the help files didn't turn up an answer. I noticed while formatting some figures for publication that text enclosed in expression() and used for a title displays differently than a string, regardless of the par options. On both postscript() and x11() devices, the regular text is heavier than the expression text. Here's an example. The real thing used expression() to produce superscripts, but I boiled this down to the simplest possible case. par(mfrow=c(2,1)) plot(1:10, 1:10, main=Figure A) plot(1:10, 1:10, main=expression(Figure B)) Is there some straightforward way to make them match, other than putting expression() around all strings? I'm currently using R 2.5.0 on Fedora core 5. Thanks, Sarah -- Sarah Goslee http://www.functionaldiversity.org __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- View this message in context: http://www.nabble.com/text-display-using-expression-or-not-tf3929372.html#a11146488 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] removing values from a vector, where both the value and its name are the same?
Patrick Burns wrote: In case it matters, the given solution has a problem if the data look like: x - c(sum=77, test=99, sum=99) By the description all three elements should be kept, but the duplicated solution throws out the last element. A more complicated solution is: unique(data.frame(x, names(x))) (and then put the vector back together again). Yes, I was about to say the same. x[!duplicated(cbind(x,names(x)))] looks like it might cut the mustard. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] removing values from a vector, where both the value and its name are the same?
Good catch Patrick. I am guilty of the same error. Regards, Marc On Fri, 2007-06-15 at 20:25 +0100, Patrick Burns wrote: In case it matters, the given solution has a problem if the data look like: x - c(sum=77, test=99, sum=99) By the description all three elements should be kept, but the duplicated solution throws out the last element. A more complicated solution is: unique(data.frame(x, names(x))) (and then put the vector back together again). Patrick Burns [EMAIL PROTECTED] +44 (0)20 8525 0696 http://www.burns-stat.com (home of S Poetry and A Guide for the Unwilling S User) jim holtman wrote: try this: x[!(duplicated(names(x))duplicated(x))] sum min max count count test 77 4 9 5 877 On 6/15/07, new ruser [EMAIL PROTECTED] wrote: I have an array such as: x=c(sum=77, min=4,max=9, count=5, min=4,max=9, count=8 , test=77) I wish to remove values where both the name and the value are identical. eg. i wish to end up with: x2=c(sum=77, min=4,max=9, count=5, count=8, test=77) What is the best way to do this? - Park yourself in front of a world of choices in alternative vehicles. [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] R vs. Splus in Pharma/Devices Industry
--- Douglas Bates [EMAIL PROTECTED] wrote: On 6/15/07, Nicholas Lewin-Koh [EMAIL PROTECTED] wrote: Hi, I just saw this thread. This issue, and the larger scale issue of open source in industry is being addressed. One has to realize that the behemoth that is the clinical aperatus of the pharma industry is very conservative and very slow to change. In many cases switching to R would meean changing a great many processes all based on legacy code. One of the big issues is that the industry demands consistency not necessarily correctness. All that said there are a great many areas where R could be used that would not impact regulatory submission, data security etc. Many in pharma are quietly working on this, but steps are small and incremental. Development takes time in industry because of the amount of documentation necessary. All this impacts the free nature of R and cost and risk (From the industries perspective) need to be justified. Probably when the statistical community is using Z big pharma will be ready to use R. %P I think you mean A, not Z. First there was S, then there was R. We're regressing __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] [OT] 'gv' and fractional points
Check out the engauge digitizer: http://digitizer.sourceforge.net/ On 6/15/07, Ted Harding [EMAIL PROTECTED] wrote: Hi Folks, This is off-topic R-wise, but it may be close to the heart of many R-users, so I think it may be the best place to ask! Users of 'gv' (the front end to ghostscript) will be aware of the little window which gives you the x-y coordinates (in points = 1/72 inch) of the position of the cross-hair mouse cursor. These coordinates are those of the corresponding position on the printed page, relative to some origin. I have often used this to extract numerical values for data from graphs in Postscript files (also PDF files, after you have converted them to PS). Then (veering back on topic ... ) you can submit the numerical data to R and try your own analyses on these data, and compare with what the article does. However, this little window only gives the numbers in whole points. Say a smallish graphic may print out 3 inches wide or high. Then you get precision of 1/216 per 3 inches or 0.4% of full scale. This can be adequate on many occasions, but can be on the coarse side on other occasions. Even for a 6-inch-wide/high graph, you only get down to 0.2% of full scale. If it were possible to induce 'gv' to display these coordinates in tenths of a point, then much greater precision (as adequate as one can expect to hope for when, in effect, measuring off the graph) could be obtained. Does anyone know: a) Whether it is possible to persuade 'gv' to give this display in fractional points (my own search of the documentation has not revealed anything); b) Of any alternative to 'gv' as PS viewer which would provide this capability? With thanks, and best wishes to all, Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 16:13:21 -- XFMail -- __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] model.frame: how does one use it?
On 6/15/07, hadley wickham [EMAIL PROTECTED] wrote: On 6/15/07, Deepayan Sarkar [EMAIL PROTECTED] wrote: [...] By 'name', I mean a symbol, such as 'Age' or 'myweight'. So basically, everything is as you would expect if the name is visible in data, but if not, the search starts in the environment of the formula, not the environment where the function call is being made (which is the standard evaulation behaviour). This is a feature, not a bug (things would be a lot more confusing if it were the other way round). Could you give an example? It's always seemed confusing to me and I don't see why looking in the environment of the formula helps. Good question. I remember being convinced that it was a good idea, but no longer remember why. This is the best I can come up with right now: ## stupid function that fits a model on a random subsample lmsub - function(formula, data, p = 0.7) { n - nrow(data) newdata - data[sample(round(n * p)), ] lm(formula, newdata) } mydata - data.frame(x = 1:100, y = rnorm(100)) n - 2 lmsub(y ~ poly(x, n), data = mydata) I don't think the use of the name 'n' is unusual in either case, and we definitely wouldn't want the one inside 'lmsub' to be used for the formula. -Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] model.frame: how does one use it?
Hi Mark, Thanks for the reply. On 15 June 2007 at 14:33, Marc Schwartz wrote: | On Fri, 2007-06-15 at 10:47 -0500, Dirk Eddelbuettel wrote: | Philipp Benner reported a Debian bug report against r-cran-rpart aka rpart. | In short, the issue has to do with how rpart evaluates a formula and | supporting arguments, in particular 'weights'. | | A simple contrived example is | | - | library(rpart) | | ## using data from help(rpart), set up simple example | myformula - formula(Kyphosis ~ Age + Number + Start) | mydata - kyphosis | myweight - abs(rnorm(nrow(mydata))) | | goodFunction - function(mydata, myformula, myweight) { |hyp - rpart(myformula, data=mydata, weights=myweight, method=class) |prev - hyp | } | goodFunction(mydata, myformula, myweight) | cat(Ok\n) | | ## now remove myweight and try to compute it inside a function | rm(myweight) | | badFunction - function(mydata, myformula) { |myweight - abs(rnorm(nrow(mydata))) |mf - model.frame(myformula, mydata, myweight) |print(head(df)) |hyp - rpart(myformula, | data=mf, | weights=myweight, | method=class) |prev - hyp | } | badFunction(mydata, myformula) | cat(Done\n) | - | | Here goodFunction works, but only because myweight (with useless random | weights, but that is not the point here) is found from the calling | environment. | | badFunction fails after we remove myweight from there: | | :~ cat /tmp/philipp.R | R --slave | Ok | Error in eval(expr, envir, enclos) : object myweight not found | Execution halted | :~ | | As I was able to replicate it, I reported this to the package maintainer. It | turns out that seemingly all is well as this is supposed to work this way, | and I got a friendly pointer to study model.frame and its help page. | | Now I am stuck as I can't make sense of model.frame -- see badFunction | above. I would greatly appreciate any help in making rpart work with a local | argument weights so that I can tell Philipp that there is no bug. :) | | Regards, Dirk | | | Dirk, | | As you note, the issue is the non-standard evaluation of the arguments | in model.frame() The key section of the Details in ?model.frame is: | | | All the variables in formula, subset and in ... are looked for first in | data and then in the environment of formula (see the help for formula() | for further details) and collected into a data frame. Then the subset | expression is evaluated, and it is is used as a row index to the data | frame. Then the na.action function is applied to the data frame (and may | well add attributes). The levels of any factors in the data frame are | adjusted according to the drop.unused.levels and xlev arguments. | | | Note that even with your goodFunction(), if 'myweight' is created within | the environment of the function and not in the global environment, it | still fails: | | library(rpart) | myformula - formula(Kyphosis ~ Age + Number + Start) | mydata - kyphosis | | goodFunction - function(mydata, myformula) { | myweight - abs(rnorm(nrow(mydata))) | hyp - rpart(myformula, data=mydata, | weights=myweight, method=class) | prev - hyp | } | | | goodFunction(mydata, myformula) | Error in eval(expr, envir, enclos) : object myweight not found | | | However, now let's do this: | | | library(rpart) | myformula - formula(Kyphosis ~ Age + Number + Start) | mydata - kyphosis | myweight - abs(rnorm(nrow(mydata))) | | goodFunction - function(mydata, myformula) { | hyp - rpart(myformula, data=mydata, | weights=myweight, method=class) | prev - hyp | } | | goodFunction(mydata, myformula) | | | It works, because 'myweight' is found in the global environment, which | is where the formula is created. Well,yes, but doesn't this just recreate the working example I showed above? It works 'because we get lucky' with the data in the outer global env. | Now, final example, try this: | | | library(rpart) | goodFunction - function() { | myformula - formula(Kyphosis ~ Age + Number + | Start) | mydata - kyphosis | myweight - abs(rnorm(nrow(mydata))) | | hyp - rpart(myformula, data=mydata, | weights=myweight, method=class) | prev - hyp | } | | goodFunction() | | | It works because the formula is created within the environment of the | function and hence,
Re: [R] [OT] 'gv' and fractional points
Hi If I understand correctly, this is something that the 'grImport' package might be very useful for. You can import the PostScript image into R, which means that you can draw the image, but you also have the locations of everything that is drawn as numeric values so you should be able (probably after a bit of transformation) to extract the values to very good accuracy. If you can provide me with an example file, I'd be happy to play around and see if I could get this to work. Paul (Ted Harding) wrote: On 15-Jun-07 16:29:53, Ted Harding wrote: [...] However, as a follow-up, I've since found that one can (somewhat tediously) do what I was asking with the GIMP. As well as the awkwardness of doing it the GIMP way, I've discovered another disadvantage. I'd previously tried it on a rather small image (175x70 points, = 2.43x0.97 inches). I then tried it on a full A4 page. Even at a GIMP Scale factor of 300, this leads to a 50MB temporary file being created. At 1000, this would rise to some 550MB, as I found out after this attempt filled up the limited spare space I have on the disk drive in question ... No doubt Scale=300 (as opposed to the default of 100) may be ample for most purposes, but the overhead is still unpleasant! Hence I'm once again hankering after something which will display a PS file as efficiently as 'gv', but will report the cursor position in fractions of a point! Best wishes to all, Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 19:18:48 -- XFMail -- E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 20:33:19 -- XFMail -- __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- Dr Paul Murrell Department of Statistics The University of Auckland Private Bag 92019 Auckland New Zealand 64 9 3737599 x85392 [EMAIL PROTECTED] http://www.stat.auckland.ac.nz/~paul/ __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Upgraded FC4 to FC5 - unable to start device X11 in R
Hi, I upgraded my system from FC4 to FC5. So now I have R version 2.5.0 (2007-04-23). But I ran into a problem with starting device X11 (message below). X11() can't find X11 font Error in X11(display, width, height, pointsize, if (is.null(gamma)) 1 else gamma, : unable to start device X11 Any idea? Thanks, L. -- Li-Jung Liang, Ph.D. Department of Biostatistics UCLA School of Public Health Los Angeles, CA 90095-1772 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] model.frame: how does one use it?
Thanks for your explanation! With this in mind, either of the following might do what you want: badFunction - function(mydata, myformula) { mydata$myweight - abs(rnorm(nrow(mydata))) hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } badFunction - function(mydata, myformula) { myweight - abs(rnorm(nrow(mydata))) environment(myformula) - environment() hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } OK, this is what I have now: adaboostBad - function(formula, data) { ## local definition of the weight vector (won't work because pima.formula is not defined within this function) w - abs(rnorm(nrow(data))) rpart(formula, data=data, weights=w) } adaboostGood - function(formula, data) { ## create weight vector in the data object data$w - abs(rnorm(nrow(data))) rpart(formula, data=data, weights=w) } adaboostBest - function(formula, data) { ## associate the current environment (this function's one) with the object `formula' environment(formula) - environment() w - abs(rnorm(nrow(data))) rpart(formula, data=data, weights=w) } As far as I understand this non-standard evaluation stuff, adaboostGood() and adaboostBest() are the only two possibilities to call rpart() with weight vectors. Now suppose that I don't know what `data' contains and suppose further that it already contains a column called `w'. adaboostGood() would overwrite that column with new data which is then used as weight vector and as training data for rpart(). adaboostBest() would just use the wrong data as weight vector as it finds data$w before the real weight vector. So, in both cases I have to check for `names(data) == w` and stop if TRUE? Or is there a better way? Regards -- Philipp Benner __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] model.frame: how does one use it?
On Fri, 2007-06-15 at 15:34 -0500, Dirk Eddelbuettel wrote: Hi Mark, Thanks for the reply. On 15 June 2007 at 14:33, Marc Schwartz wrote: | On Fri, 2007-06-15 at 10:47 -0500, Dirk Eddelbuettel wrote: | Philipp Benner reported a Debian bug report against r-cran-rpart aka rpart. | In short, the issue has to do with how rpart evaluates a formula and | supporting arguments, in particular 'weights'. | | A simple contrived example is | | - | library(rpart) | | ## using data from help(rpart), set up simple example | myformula - formula(Kyphosis ~ Age + Number + Start) | mydata - kyphosis | myweight - abs(rnorm(nrow(mydata))) | | goodFunction - function(mydata, myformula, myweight) { |hyp - rpart(myformula, data=mydata, weights=myweight, method=class) |prev - hyp | } | goodFunction(mydata, myformula, myweight) | cat(Ok\n) | snip | | However, now let's do this: | | | library(rpart) | myformula - formula(Kyphosis ~ Age + Number + Start) | mydata - kyphosis | myweight - abs(rnorm(nrow(mydata))) | | goodFunction - function(mydata, myformula) { | hyp - rpart(myformula, data=mydata, | weights=myweight, method=class) | prev - hyp | } | | goodFunction(mydata, myformula) | | | It works, because 'myweight' is found in the global environment, which | is where the formula is created. Well,yes, but doesn't this just recreate the working example I showed above? It works 'because we get lucky' with the data in the outer global env. Technically, it is not the same, as I was trying to emphasize that there was no need to pass 'myweight' as an argument to the function to facilitate successful location/evaluation within the function. We don't get lucky here. The behavior is by design and consistent with the documentation, which is that 'myweight' in the call to rpart() is evaluated within the environment of the formula in this case. The formula is created in the global environment, so 'myweight' is found there. Hence, no need to pass it as an argument. A review of the code for rpart() will reveal code similar to that which is used in most R modeling functions, relative to the evaluation of the formula, associated args and the creation of the model frame. One exception to the above, is that in other modeling functions, one could forgo passing the formula and just pass the entire data frame, where the presumption is that the first column is the response variable and the remaining columns would be the independent terms. I don't see that supported in rpart(). | Now, final example, try this: | | | library(rpart) | goodFunction - function() { | myformula - formula(Kyphosis ~ Age + Number + | Start) | mydata - kyphosis | myweight - abs(rnorm(nrow(mydata))) | | hyp - rpart(myformula, data=mydata, | weights=myweight, method=class) | prev - hyp | } | | goodFunction() | | | It works because the formula is created within the environment of the | function and hence, 'myweight', which is created there as well, is | found. That works because we force it to be local. BDR claims that my 'badFunction' (derived from Philipp's original bug report) above can be made to work provide you use model.frame. I asked about model.frame -- and you were kind enough do answer, but you dodged the question. So let me try again: How can rpart be called inside a function using a local weight variable as I do above ? Either it can, and the BDR is right and there is no bug, or one cannot, and then mere mortals like myself must consider rpart to be buggy as it does not support all its argument in at least some conceivable calling situations. Is that a fair question? Regards, Dirk Yep, entirely fair. Without knowing what specific approach Prof. Ripley had in mind, I am envisioning a couple of possibilities, but here is one: library(rpart) myformula - formula(Kyphosis ~ Age + Number + Start) mydata - kyphosis badFunction - function(mydata, myformula) { mydata$myweight - abs(rnorm(nrow(mydata))) rpart(myformula, data = mydata, weights = myweight, method = class) } badFunction(mydata, myformula) Basically, there are 3 places in which 'myweights' could be found: 1. Formula environment 2. Data frame environment 3. Global environment In this case, we add the weights as a new column within the function to the 'mydata' data frame, so that it will be found in the call to rpart(), based upon location number 2 above. Does that help? Regards, Marc
[R] rect() does not work
I have typed the following commands: x - c(1:10) y - x plot(x,y) rect(2,2,4,4) I am expecting a rectangular box, however, nothing occurs. - Be a PS3 game guru. [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] R vs. Splus in Pharma/Devices Industry
On 15-Jun-07 19:30:45, Douglas Bates wrote: On 6/15/07, Nicholas Lewin-Koh [EMAIL PROTECTED] wrote: Hi, Probably when the statistical community is using Z big pharma will be ready to use R. %P I think you mean A, not Z. First there was S, then there was R. And, furthermore, A is further from R than Z is, which seems fitting. Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 15-Jun-07 Time: 23:57:13 -- XFMail -- __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] model.frame: how does one use it?
On 6/15/07, Philipp Benner [EMAIL PROTECTED] wrote: Thanks for your explanation! With this in mind, either of the following might do what you want: badFunction - function(mydata, myformula) { mydata$myweight - abs(rnorm(nrow(mydata))) hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } badFunction - function(mydata, myformula) { myweight - abs(rnorm(nrow(mydata))) environment(myformula) - environment() hyp - rpart(myformula, data=mydata, weights=myweight, method=class) prev - hyp } OK, this is what I have now: adaboostBad - function(formula, data) { ## local definition of the weight vector (won't work because pima.formula is not defined within this function) w - abs(rnorm(nrow(data))) rpart(formula, data=data, weights=w) } adaboostGood - function(formula, data) { ## create weight vector in the data object data$w - abs(rnorm(nrow(data))) rpart(formula, data=data, weights=w) } adaboostBest - function(formula, data) { ## associate the current environment (this function's one) with the object `formula' environment(formula) - environment() w - abs(rnorm(nrow(data))) rpart(formula, data=data, weights=w) } As far as I understand this non-standard evaluation stuff, adaboostGood() and adaboostBest() are the only two possibilities to call rpart() with weight vectors. Now suppose that I don't know what `data' contains and suppose further that it already contains a column called `w'. adaboostGood() would overwrite that column with new data which is then used as weight vector and as training data for rpart(). adaboostBest() would just use the wrong data as weight vector as it finds data$w before the real weight vector. So, in both cases I have to check for `names(data) == w` and stop if TRUE? Or is there a better way? Well, that depends on what you want to happen when there is a column called 'w' in data. I don't see a situation where it makes sense to use data$w as weights ('w' is just a name you happen to choose inside adaboostBest), so I would just go with adaboostGood. In case you are worried about overwriting the original data, that may not be happening in the sense you are thinking. When you say data$w - abs(rnorm(nrow(data))) inside adaboostGood, that modifies a local copy of the data argument, not the original (R argument semantics are call by value, not call by reference). You are losing data$w in the local copy in your function, but why would you care if you are not using it anyway. Of course, if your formula contains a reference to 'w' then you will get wrong results, so checking for a unique name is always safer. In addition, use an obfuscated name like '.__myWeights' instead of 'w', and the check will be almost always irrelevant. -Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] rect() does not work
On Fri, 2007-06-15 at 16:22 -0700, scott flemming wrote: I have typed the following commands: x - c(1:10) y - x plot(x,y) rect(2,2,4,4) I am expecting a rectangular box, however, nothing occurs. What version of R? What OS? No problems here using R version 2.5.0 Patched (2007-06-05 r41831) on F7. You might want to check and see what par(lty) and par(lwd) return, in case they have been set to values resulting in lines that you would not be able to see. BTW, the c() is not required in creating 'x': x - 1:10 x [1] 1 2 3 4 5 6 7 8 9 10 Marc Schwartz __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Stacked barchart color
Hi Hadley, I tried your suggestion, using ggplot2, but I am still having a problem. The final plot lacks the figure legend -- which it had before I added the scale_fill_identity() bit. Can you see what I am doing wrong? (By the way, all I am trying to do is make the figure monochrome friendly. Is there an easy way to prepare ggplot graphics for a monochrom device?) Thanks,Owen y$color = factor(y$Fnd) y$color = c(black,darkgray,lightgray,white) y Fnd locusFreq color 1 signeg A 0.087248322 black 2 neg A 0.711409396 darkgray 3 pos A 0.201342282 lightgray 4 sigpos A 0.0 white 5 signeg C 0.320754717 black 6 neg C 0.603773585 darkgray 7 pos C 0.075471698 lightgray 8 sigpos C 0.0 white 9 signeg B 0.157534247 black 10neg B 0.732876712 darkgray 11pos B 0.109589041 lightgray 12 sigpos B 0.0 white p = ggplot(y, aes(x=locus, y=Freq, fill=color)) + geom_bar(position=fill) + scale_fill_identity() p hadley wrote: Hi Dieter, You can do this with ggplot2 (http://had.co.nz/ggplot2) as follows: library(ggplot2) barley1 - subset(barley, site==Grand Rapids variety %in% c(Velvet,Peatland)) barley1[] - lapply(barley1, [, drop=TRUE) qplot(variety, yield, data=barley1, geom=bar, stat=identity, fill=factor(year)) barley1$fill - c(red,green,blue,gray) qplot(variety, yield, data=barley1, geom=bar, stat=identity, fill=fill) + scale_fill_identity() See http://had.co.nz/ggplot2/scale_identity.html and http://had.co.nz/ggplot2/position_stack.html for more details. Hadley -- View this message in context: http://www.nabble.com/Stacked-barchart-color-tf3909162.html#a11149419 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] R vs. Splus in Pharma/Devices Industry
On 15-Jun-07 19:43:55, John Kane wrote: --- Douglas Bates [EMAIL PROTECTED] wrote: On 6/15/07, Nicholas Lewin-Koh [EMAIL PROTECTED] wrote: [...] I think you mean A, not Z. First there was S, then there was R. We're regressing But not to mediocrity! [1] Ted. [1] Practical Regression and Anova using R Julian J. Faraway (July 2002), page 15. http://cran.r-project.org/doc/contrib/Faraway-PRA.pdf E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 16-Jun-07 Time: 00:43:39 -- XFMail -- __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Upgraded FC4 to FC5 - unable to start device X11 in R
On Fri, 2007-06-15 at 13:50 -0700, Li-Jung Liang wrote: Hi, I upgraded my system from FC4 to FC5. So now I have R version 2.5.0 (2007-04-23). But I ran into a problem with starting device X11 (message below). X11() can't find X11 font Error in X11(display, width, height, pointsize, if (is.null(gamma)) 1 else gamma, : unable to start device X11 Any idea? Thanks, L. Sounds like either you did not install some font related RPMS during your 'upgrade' or your xorg.conf file has perhaps become corrupted relative to defining font paths. Check as 'root': # yum list xorg-x11-fonts* and see what it returns. How did you install R? Using the RPMS from CRAN or from Fedora Extras? Also, note that FC5 will reach EOL (End of Life) on June 29, which means that you will receive no further updates (security, bug fixes, etc) after that date. You should really consider upgrading to FC6. F7 just came out and I am running it, but I would recommend that typical users wait a while before doing so to give it time to stabilize. HTH, Marc Schwartz __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] [OT] 'gv' and fractional points
On 15-Jun-07 20:08:05, Gabor Grothendieck wrote: Check out the engauge digitizer: http://digitizer.sourceforge.net/ Thanks, Gabor. This looks useful, as far as it goes. However, it doesn't deal directly with PS, so one would have to work to pixel resolution of screenshots as displayed by say 'gv'. In the case of an example such as the Fig 4 I mentioned in a previous post, this would involve taking several screenshots to tile over the entirety of the graph, since the pixel accuracy or a screenshot of the entire graph would be rather poor. This would introduce the neccessity of keeing track of points from screenshot to screenshot, and also each screenshot would have its own coordinate system so these would have to be reconciled before the final dataset could be compiled. But thanks -- this could well be worth a try in a different kind of context (e.g. I have some screenshots of old maps which I want to digitise ... ). Best wishes, Ted. E-Mail: (Ted Harding) [EMAIL PROTECTED] Fax-to-email: +44 (0)870 094 0861 Date: 16-Jun-07 Time: 01:13:53 -- XFMail -- __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] plot via xyplot not being saved
Hi everyone, it's been a while I've been trying to save a plot created via lattice:::xyplot if I have a file tst.R with the following code: y - rnorm(100) x - rnorm(100) z - sample(letters[1:4], 100, rep=T) library(lattice) bitmap(tst.png) xyplot(y~x|z) dev.off() and I source it, I get the tst.png file, which is a blank page. If I copy and paste instead, I get the correct plot. Any suggestion? Thank you very much, b sessionInfo() R version 2.5.0 (2007-04-23) x86_64-unknown-linux-gnu locale: LC_CTYPE=en_US.iso885915;LC_NUMERIC=C;LC_TIME=en_US.iso885915;LC_COLLATE =en_US.iso885915;LC_MONETARY=en_US.iso885915;LC_MESSAGES=en_US.iso885915 ;LC_PAPER=en_US.iso885915;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASU REMENT=en_US.iso885915;LC_IDENTIFICATION=C attached base packages: [1] stats graphics grDevices utils datasets methods [7] base other attached packages: lattice 0.15-4 __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] plot via xyplot not being saved
On 6/15/07, Benilton Carvalho [EMAIL PROTECTED] wrote: Hi everyone, it's been a while I've been trying to save a plot created via lattice:::xyplot if I have a file tst.R with the following code: y - rnorm(100) x - rnorm(100) z - sample(letters[1:4], 100, rep=T) library(lattice) bitmap(tst.png) xyplot(y~x|z) dev.off() and I source it, I get the tst.png file, which is a blank page. If I copy and paste instead, I get the correct plot. Any suggestion? Use source(..., echo = TRUE) -Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] plot via xyplot not being saved
So, if those statements are inside a function, I have to make my function to have an 'echo' argument/functionality? eg.: ## begin test.R test - function(n){ y - rnorm(n) x - rnorm(n) z - sample(letters[1:4], n, rep=T) library(lattice) bitmap(tst.png) xyplot(y~x|z) dev.off() } test(100) ## end test.R source(test.R, echo=T) also fails in this case... thanks a lot, b On Jun 15, 2007, at 8:53 PM, [EMAIL PROTECTED] wrote: On 6/15/07, Benilton Carvalho [EMAIL PROTECTED] wrote: Hi everyone, it's been a while I've been trying to save a plot created via lattice:::xyplot if I have a file tst.R with the following code: y - rnorm(100) x - rnorm(100) z - sample(letters[1:4], 100, rep=T) library(lattice) bitmap(tst.png) xyplot(y~x|z) dev.off() and I source it, I get the tst.png file, which is a blank page. If I copy and paste instead, I get the correct plot. Any suggestion? Use source(..., echo = TRUE) -Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] R vs. Splus in Pharma/Devices Industry
On Sat, 2007-06-16 at 00:43 +0100, [EMAIL PROTECTED] wrote: On 15-Jun-07 19:43:55, John Kane wrote: --- Douglas Bates [EMAIL PROTECTED] wrote: On 6/15/07, Nicholas Lewin-Koh [EMAIL PROTECTED] wrote: [...] I think you mean A, not Z. First there was S, then there was R. We're regressing But not to mediocrity! [1] Ted. [1] Practical Regression and Anova using R Julian J. Faraway (July 2002), page 15. http://cran.r-project.org/doc/contrib/Faraway-PRA.pdf Which in turn is of course paraphrasing Galton's Regression Toward Mediocrity in Hereditary Stature”. Journal of the Anthropological Institute 15 : 246-63, 1886. http://galton.org/essays/1880-1889/galton-1886-jaigi-regression-stature.pdf :-) Regards, Marc __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] [Not R question]: Better fit for order probit model
Hi, I have a model which tries to fit a set of data with 10-level ordered responses. Somehow, in my data, the majority of the observations are from level 6-10 and leave only about 1-5% of total observations contributed to level 1-10. As a result, my model tends to perform badly on points that have lower level than 6. I would like to ask if there's any way to circumvent this problem or not. I was thinking of the followings ideas. But I am opened to any suggestions if you could please. 1. Bootstrapping with small size of samples each time. Howevever, in each sample basket, I intentionally sample in such a way that there is a good mix between observations from each level. Then I have to do this many times. But I don't know how to obtain the true standard error of estimated parameters after all bootstrapping has been done. Is it going to be simply the average of all standard errors estimated each time? 2. Weighting points with level 1-6 more. But it's unclear to me how to put this weight back to maximum likelihood when estimating parameters. It's unlike OLS where your objective is to minimize error or, if you'd like, a penalty function. But MLE is obviously not a penalty function. 3. Do step-wise regression. I will segment the data into two regions, first points with response less than 6 and the rest with those above 6. The first step is a binary regression to determine if the point belongs to which of the two groups. Then in the second step, estimate ordered probit model for each group separately. The question here is then, why I am choosing 6 as a cutting point instead of others? Any suggestions would be really appreciated. Thank you. - adschai __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] fSeries - Ox - ver: 240.10068 - Steps to make it work
-Bugs and fixes reported to Diethelm Wuertz. -In the interim. To make the Ox functions part of the fSeries package work please follow the following steps. - 1. Install R-project. 2. Install fSeries. 3. Download: http://www.core.ucl.ac.be/~laurent/[EMAIL PROTECTED]/site/xbdcons/garch42.zip ([EMAIL PROTECTED] package for Ox) 4. Download: http://www.doornik.com/download/b736gw/oxcons.html (Ox console) 5. Install Ox console to directory c:\ox 6. Copy: R-2.5.0\library\fSeries\ox\GarchOxModelling.ox c:\ox\ox\lib\GarchOxModelling.ox 7. Load R-Project 8. Run: library(fSeries) data(dem2gbp) x = dem2gbp[, 1] garchOxFit(~garch(1,1),x) -- Causes the errors: Error in paste(OXPATH, \\bin\\oxl.exe , OXPATH, \\lib\\GarchOxModelling.ox, : object OXPATH not found -- This can be fixed by setting the directory in R-Project to where Ox is installed. ie. OXPATH=c:\ox\ox. 9. garchOxFit(~garch(1,1),x) 10. Copy the [EMAIL PROTECTED] package files to c:\ox\ox\packages\Garch42 11. Change the line in GarchOxModelling.ox: #import packages/Garch40/garch to #import packages/Garch42/garch 13. garchOxFit(~garch(1,1),x) Will give the following error: -- Ox version 4.1 (Windows) (C) J.A. Doornik, 1994-2006 This version may be used for academic research and teaching only Copyright for this package: S. Laurent and J.P. Peters, 2000-2006. [EMAIL PROTECTED] package version 4.2, object created on 7-06-2007 - Starting Values === Parameter Starting Value Cst(M) 0.01 Cst(V) 0.05 ARCH(Alpha1) 0.10 GARCH(Beta1) 0.80 - Bounds == Parameter Lower Bound Upper Bound Cst(M)-100.00 100.00 Cst(V) 0.00 100.00 ARCH(Alpha1) 0.001.00 GARCH(Beta1) 0.001.00 Runtime error: 'DoEstimation' wrong number of arguments Runtime error occurred in main(207), call trace: c:/ox/Ox\lib\GarchOxModelling.ox (207): main Error in file(file, r) : unable to open connection In addition: Warning message: cannot open file 'OxResiduals.csv', reason 'No such file or directory' -- 14. This can be fixed by changing the line in GarchOxModelling.ox: from garchobj.DoEstimation(); to garchobj.DoEstimation(); 15. Running: garchOxFit(~garch(1,1),x) Will give a 'typo' error is at the end of the estimation. Which needs to be fixed by the developer as it is compiled code. However, now the function works and can call Ox from R-Prooject to estimate models. [1] garch(1, 1) Ox version 4.1 (Windows) (C) J.A. Doornik, 1994-2006 This version may be used for academic research and teaching only Copyright for this package: S. Laurent and J.P. Peters, 2000-2006. [EMAIL PROTECTED] package version 4.2, object created on 11-06-2007 - Starting Values === Parameter Starting Value Cst(M) 0.01 Cst(V) 0.05 ARCH(Alpha1) 0.10 GARCH(Beta1) 0.80 - Bounds == Parameter Lower Bound Upper Bound Cst(M)-100.00 100.00 Cst(V) 0.00 100.00 ARCH(Alpha1) 0.001.00 GARCH(Beta1) 0.001.00 Starting values parameters 0.01 0.05 0.1 0.8 Initial function = -1312.86321151 Position after 5 SQPF iterations parameters 0.0054353 0.038781 0.20486 0.60034 function value = -1125.76603281 steplen = 0.03125 Position after 10 SQPF iterations parameters -0.0088801 0.012371 0.17148 0.78507 function value = -1106.89461928 Position after 14 SQPF iterations Status: Strong convergence parameters -0.0061830 0.010761 0.15341 0.80588 function value = -1106.58656395 Ian - Doing output ** SPECIFICATIONS ** Dependent variable : X Mean Equation : ARMA (0, 0) model. No regressor in the mean Variance Equation : GARCH (1, 1) model. No regressor in the variance The distribution is a Gauss distribution. Strong convergence using numerical derivatives Log-likelihood = -1106.59 Please wait : Computing the Std Errors ... Maximum Likelihood Estimation (Std.Errors based
Re: [R] [Not R question]: Better fit for order probit model
At 09:31 PM 6/15/2007, adschai wrote: I have a model which tries to fit a set of data with 10-level ordered responses. Somehow, in my data, the majority of the observations are from level 6-10 and leave only about 1-5% of total observations contributed to level 1-10. As a result, my model tends to perform badly on points that have lower level than 6. I would like to ask if there's any way to circumvent this problem or not. I was thinking of the followings ideas. But I am opened to any suggestions if you could please. 1. Bootstrapping with small size of samples each time. Howevever, in each sample basket, I intentionally sample in such a way that there is a good mix between observations from each level. Then I have to do this many times. But I don't know how to obtain the true standard error of estimated parameters after all bootstrapping has been done. Is it going to be simply the average of all standard errors estimated each time? 2. Weighting points with level 1-6 more. But it's unclear to me how to put this weight back to maximum likelihood when estimating parameters. It's unlike OLS where your objective is to minimize error or, if you'd like, a penalty function. But MLE is obviously not a penalty function. 3. Do step-wise regression. I will segment the data into two regions, first points with response less than 6 and the rest with those above 6. The first step is a binary regression to determine if the point belongs to which of the two groups. Then in the second step, estimate ordered probit model for each group separately. The question here is then, why I am choosing 6 as a cutting point instead of others? Any suggestions would be really appreciated. Thank you. You could do the obvious, and lump categories such as 1-6 or 1-7 together to make a composite category. You don't mention the size of your dataset. If there are 10,000 data, you might live with a 1% category. If you only have 100 data, you have too many categories. Also, next time plan your study and training better so that next time your categories are fully utilized. And don't use so many categories. People have trouble even selecting responses on a 5-level scale. Robert A. LaBudde, PhD, PAS, Dpl. ACAFS e-mail: [EMAIL PROTECTED] Least Cost Formulations, Ltd.URL: http://lcfltd.com/ 824 Timberlake Drive Tel: 757-467-0954 Virginia Beach, VA 23464-3239Fax: 757-467-2947 Vere scire est per causas scire __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Upgraded FC4 to FC5 - unable to start device X11 in R
On Fri, 15-Jun-2007 at 01:50PM -0700, Li-Jung Liang wrote: | Hi, | | I upgraded my system from FC4 to FC5. So now I have R version 2.5.0 | (2007-04-23). | But I ran into a problem with starting device X11 (message below). | | X11() | can't find X11 font | Error in X11(display, width, height, pointsize, if (is.null(gamma)) 1 | else gamma, : | unable to start device X11 | | Any idea? Check what search() gives you. Make sure package:grDevices is on that list. I ran into that problem myself. I never figured out how come it used to be there, but suddenly ceased to be. The problem didn't appear until R-2.5.0. In the USER-VISIBLE CHANGES under CHANGES IN R VERSION 2.0.0, I see this: o Package 'graphics' has been split into 'grDevices' (the graphics devices shared between base and grid graphics) and 'graphics' (base graphics). Each of the 'graphics' and 'grid' packages load 'grDevices' when they are attached. Note that ps.options() has been moved to grDevices and user hooks may need to be updated. My suspicion is that grDevices did not have to be specifically loaded if graphics was one of your default packages loaded at startup. Since few others seem to have had the same experience, I'd hesitate to call it a bug. I've got round the problem by adding grDevices to the list in my ~/.Rprofile file. Others might have a better explanation, of course. HTH -- ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~. ___Patrick Connolly {~._.~} Great minds discuss ideas _( Y )_Middle minds discuss events (:_~*~_:)Small minds discuss people (_)-(_) . Anon ~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] plot via xyplot not being saved
On 6/15/07, Benilton Carvalho [EMAIL PROTECTED] wrote: So, if those statements are inside a function, I have to make my function to have an 'echo' argument/functionality? eg.: ## begin test.R test - function(n){ y - rnorm(n) x - rnorm(n) z - sample(letters[1:4], n, rep=T) library(lattice) bitmap(tst.png) xyplot(y~x|z) dev.off() } test(100) ## end test.R source(test.R, echo=T) also fails in this case... Yes. The following will produce some output (the values of x + y and x - y) if you type it out at the R prompt: x - rnorm(10) y - rnorm(10) x + y x - y If you put that in a file and source it, nothing will get printed, unless you have echo=TRUE. If you define test - function(){ x - rnorm(10) y - rnorm(10) x + y x - y } calling test() at the R prompt will only print x - y and not x + y, and so on. This is all standard R behaviour. If you want something to be printed irrespective of context, use print(), e.g. print(x + y) or print(xyplot(y~x|z)) This is also mentioned in the R FAQ. -Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] plot via xyplot not being saved
Thank you Deepayan, I understand the behavior of not printing out the results inside the functions. What I didn't know was that for xyplot() saving the plot actually meant save the result I see, which does not happen with plot(), in which case my function test() works just fine if I replaced xyplot() by plot(). Thank you very much, b On Jun 16, 2007, at 12:26 AM, [EMAIL PROTECTED] wrote: On 6/15/07, Benilton Carvalho [EMAIL PROTECTED] wrote: So, if those statements are inside a function, I have to make my function to have an 'echo' argument/functionality? eg.: ## begin test.R test - function(n){ y - rnorm(n) x - rnorm(n) z - sample(letters[1:4], n, rep=T) library(lattice) bitmap(tst.png) xyplot(y~x|z) dev.off() } test(100) ## end test.R source(test.R, echo=T) also fails in this case... Yes. The following will produce some output (the values of x + y and x - y) if you type it out at the R prompt: x - rnorm(10) y - rnorm(10) x + y x - y If you put that in a file and source it, nothing will get printed, unless you have echo=TRUE. If you define test - function(){ x - rnorm(10) y - rnorm(10) x + y x - y } calling test() at the R prompt will only print x - y and not x + y, and so on. This is all standard R behaviour. If you want something to be printed irrespective of context, use print(), e.g. print(x + y) or print(xyplot(y~x|z)) This is also mentioned in the R FAQ. -Deepayan __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] selecting cut-off in Logistic regression using ROCR package
Hi, I am using logistic regression to classify a binary psychometric data. using glm() and then predict.glm() i got the predicted odds ratio of the testing data. Next i am going to plot ROC curve for the analysis of my study. Now what i will do: 1. first select a cut-off (say 0.4) and classify the output of predict.glm() into {0,1} segment and then use it to draw ROC curve using ROCR package OR 2. just use the predicted odds ratio in ROCR package to get error rate and use the minimum error rate (as new cut-off) to draw new ROC curve. waiting for reply. with regards and thanks. Tirtha. -- View this message in context: http://www.nabble.com/selecting-cut-off-in-Logistic-regression-using-ROCR-package-tf3931603.html#a11151210 Sent from the R help mailing list archive at Nabble.com. __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.