similar to: Data packages in R V 1.3.1 in Win2K

Platform: WIN2000 Version of R: 1.6.2 I'm interested in plotting fitted values in a trellis xyplot. I believe the following should work; however, I only get the points (not the fitted lines). library(lattice) trellis.device(bg="white") xyplot(MULTDV~TIME|SUBNUM,data=TEMP, panel=function(x,y){ panel.xyplot(x,y) lines(x,fitted(lm(y~poly(x,1),na.action=na.omit)))

I apologize if this is a FAQ -- I kind of recall seeing something along these lines before, but I couldn't find the message when I searched the archives. Problem: 1. I have hundreds of small files in a subdirectory ("c:\\temp") and I would like to combine the files into a single data frame. 2. Individually, it is easy to read each file

Looking to the wonderful statistical advice that this group can offer. In behavioral science applications of stats, we are often introduced to coefficients for orthogonal polynomials that are nice integers. For instance, Kirk's experimental design book presents the following coefficients for p=4: Linear -3 -1 1 3 Quadratic 1 -1 -1 1 Cubic -1 3 -3 1 In R orthogonal

As usual, the R newsgroup set me straight (thanks to Douglas Bates, Robert Balshaw and Albyn Jones). There is really no difference between using orthogonal polynomials of the form: Linear -3 -1 1 3 Quadratic 1 -1 -1 1 Cubic -1 3 -3 1 Versus > poly(c(1:4),3) 1 2 3 [1,] -0.6708204 0.5 -0.2236068 [2,] -0.2236068 -0.5 0.6708204 [3,] 0.2236068

In lm and glm, the weights command should only be used when the variances of the observations are inversely proportional to the weights. Recently, however, a question came up regarding how one could estimate lm and glm models with weights where the weights refer to number of observations (counts). Because lm and glm do not handle this case, SE values will be wrong if one uses the weight command

Behavioral science data is often collected from nested structures (students in schools, in districts, etc.). This can produce nonindependence among responses from individuals in the same groups. Consequently, researchers are advised to model the nested nature of the data to avoid biases in SE estimates. Failing to account for nonindependence can lead to SE estimates that are too large or too

I'm reluctant to draw the S-PLUS and R comparison (these are different programs after all), but could someone tell me why the following matrix substitution works in S-PLUS, but not R. I'm curious because matrix substitution is a really slick way to "cleaning up" columns of data in data frames. For example, in the following I change values of 1 to values of 10, but only for

It turns out that there is a color named "transparent" that does the trick with the trellis device. Thanks to Brian Ripley for pointing it out. trellis.device(bg="transparent") My original post is below, but it boils down to "how can I make a trellis graph that doesn't contain any background color?" This can be useful when you paste graphics into documents

Newsgroup members, I appreciate the help on this topic. David Duffy provided a solution (below) that was quite helpful, and came close to what I needed. It did a great job creating two vectors of dichotomous variables with a known correlation (what I referred to as a phi-coefficient). My situation is a bit more complicated and I'm not sure it is easily solved. The problem is that I must

I would appreciate some thoughts on using the bootstrap functions in the library "bootstrap" to estimate confidence intervals of ICC values calculated in lme. In lme, the ICC is calculated as tau/(tau+sigma-squared). So, for instance the ICC in the following example is 0.116: > tmod<-lme(CINISMO~1,random=~1|IDGRUP,data=TDAT) > VarCorr(tmod) IDGRUP = pdLogChol(1)

If you found my subject heading to be confusing then I'm sure you'll enjoy the example I've included below. I find the apply type functions to be wonderful for avoiding loops but when I use them with existing functions, I end up using loops anyway to extract the vectors I want. I would appreciate it if someone could show me how to avoid these loops. Thanks. EXAMPLE:

Newsgroup members, Does anyone have a clever way to simulate a correlation matrix such that each column contains dichotomous variables (0,1) and where each column has different prevalence rates. For instance, I would like to simulate the following correlation matrix: > CORMAT[1:4,1:4] PUREPT PTCUT2 PHQCUT2T ALCCUTT2 PUREPT 1.0000000 0.5141552 0.1913139 0.1917923 PTCUT2

I'm sure there is a simple answer to this question, but it stumps me so far. I've been using R to create graphic images and paste them into other programs (Word, PowerPoint). Recently, I've been using lattice (wonderful package), but I can't figure out how to create graphs in lattice that don't have a background color -- that is, my graphs carry over the grey background if I

Hi List, I have the following problem: I am using the multilevel package and make.univ function for available in the package and then xyplot from lattice and I want to know how could I be able to use the ?coefficient? for the straight line that passes the data ? Example from help: library(multilevel) data(univbct) #a data set already in univariate or stacked form for job satisfaction

It is frustrating to see the labels I want in the dimensions of a list but not be able to extract those labels into titles for plots generated from component objects. If someone could set me straight, I would appreciate it. For your amusement, I have provided an example of the Byzantine code I am currently using to avoid loops: # Simulate ANOVA type test data sex<-c(rep(1,8),rep(0,8))

It is my understanding that Tcl/Tk does not come with the base installation of R under OS X ( http://www.sciviews.org/_rgui/projects/TclTk.html). Is there a simple way to explain how a user with limited tolerance for computer complexity can obtain and install Tcl/Tk for OS X? Thanks. Chuck Background: Roughly a third of my target audience uses OS X. I do not and I do not have access to a

I installed the MinGW-5.0.0 compilers today (gcc-3.4.2 is now contained in the 'Current' distribution. When the current compilers failed to compile the referenced Matrix update and not knowing how to check the exact version numbers for individual files, I applied the Id.exe and f771.exe fixes from the 'Building R for Windows' page (http://www.murdoch-sutherland.com/Rtools/#ldbug).

1) In general, I would appreciate help in adding functions to the Rcmdr menu system. I've been able to modify the menus themselves and source test code but I can't get R functions to execute from the menu. My latest proof of concept code follows: Three lines added to "Rcmdr-menus.txt": menu junkMenu topMenu "" ""

Hello Laurent, Ray and Greg, I tried the gmake settings. Yet, I still got an Error code 1 message when I run the test. running regression tests don't know how to make ../library/base/R/base (bu42). *** Error code 1 (bu21) Thanks you if you have any idea. Manuel -----Message d'origine----- De : Laurent Gautier [ mailto:laurent at genome.cbs.dtu.dk <mailto:laurent at

Hi All, I found an old article on running WINE alongside Windows 2000 yesterday. It advocated using the command line of "wine --winver nt2k [program name]", in the example "wine --winver nt2k mspaint.exe". I tried this out with Paintshop Pro which WINE HQ says does run with WINE, but had even less success than before. At least without using that extra part I got an error