search for: rab45+

Thompson's Manual to Accompany Agresti's book refers to a package named "repeated". It's not on CRAN from what I can see. I have seen rpm's for it. Where is the best place to download this package? Rick B.

Where I work a lot of people end up using Excel spreadsheets for storing data. This has limitations and maybe some less than obvious problems. I'd like to recommend a uniform way for storing and archiving data collected in the department. Most of the data could be stored in simple csv type files but it would be nice to have something that stores more information about the variables and units.

> rab45+ at pitt.edu wrote: >> Jerry, >> >> Here are the two R packages. >> >> Thanks. >> >> Rick > All the fortran codes compile OK so the problem is invoking the compiler > you > want. One approach which is a work around is to rename the gfortran > e...

I fit the following model using glmmPQL from MASS: fit.glmmPQL <- glmmPQL(ifelse(class=="Disease",1,0)~age+x1+x2,random=~1|subject,family=binomial) summary(fit.glmmPQL) The response is paired (pairing denoted by subject), although some subjects only have one response. Also, there is a perfect positive correlation between the paired responses. x1 and x2 can and do differ within each

Is there an R package available that would facilitate doing a power/sample size analysis for linear mixed effects models? I have seen the Java applets made available by Russell Length which would seem to be able to handle most any lme, but there is little documentation and it's not clear how the models need to be formulated. Rick B.

I'm estimating models using lme in a for loop. Sometimes the model doesn't converge or there are other problems. This causes the evaluation to stop prematurely. I can't remember the function name that I need to use to allow the loop to continue until the end. Could someone remind me the name of the function? I've tried searching but haven't hit upon the function. Rick B.

The stat.table function in the Epi package won't do standard deviations. It didn't seem that it would be difficult to add an "sd" function to the stat.table function. Following the example for the mean, I set up a similar function for the sd (and included it as an options) but it just won't work. (I tried sending messages to the Epi mailing list after subscribing but my mail

I'm fitting a mixed effects model: fit.1 <- lme(y~x,random=~1|id,data=df) There are two different observations for each id for both x and y. When I use plot(fit.1), there is a strong increasing linear trend in the residuals versus the fitted values (with no outliers). This also happens if I use random=~x|id. Am I specifying something incorrectly? Rick B.

I'm using groupedData from nlme. I set up a groupedData data.frame with outer=~group1. When I try to plot with outer=TRUE, I get "subscript out of bounds." This happens most of the time. When it works, I get spaghetti-type plots for comparing groups. But I don't understand why it doesn't usually work. > longa.mod.1.gd <- groupedData(mod1.logit~time|

When I try to run the example from Variogram with an lme object, I get an error (although summary works): R : Copyright 2005, The R Foundation for Statistical Computing Version 2.2.1 (2005-12-20 r36812) ISBN 3-900051-07-0 ... > fm1 <- lme(weight ~ Time * Diet, BodyWeight, ~ Time | Rat) Error: couldn't find function "lme" > Variogram(fm1, form = ~ Time | Rat, nint =

I'm estimating two models for data with n = 179 with four clusters (21, 70, 36, and 52) named siteid. I'm estimating a logistic regression model with random intercept and another version with random intercept and random slope for one of the independent variables. fit.1 <- lmer(glaucoma~(1|siteid)+x1 +x2,family=binomial,data=set1,method="ML",

I'm running R 2.4.1 (with the latest versions of all packages) on an FC6 32-bit system. When I try to install the rgl package, compilation fails: > install.packages("rgl") --- Please select a CRAN mirror for use in this session --- Loading Tcl/Tk interface ... done trying URL 'http://lib.stat.cmu.edu/R/CRAN/src/contrib/rgl_0.70.tar.gz' Content type

When trying to fit a generalized linear mixed model using glmmPQL: > fit0 <- glmmPQL(ifelse(response=="A",1,0)~gender,data=set1, random=~1|subject,family=binomial) iteration 1 Error in solve.default(pdMatrix(a, fact = TRUE)) : Lapack routine dgesv: system is exactly singular Could this be occuring because the paired responses for each subject are always the same? If

I started my laptop, opened a shell and get this error message: R : Copyright 2005, The R Foundation for Statistical Computing Version 2.2.1 (2005-12-20 r36812) ISBN 3-900051-07-0 lines deleted Error in fun(...) : couldn't find function "assignInNamespace" Error: .onLoad failed in 'loadNamespace' for 'Matrix' Fatal error: unable to restore saved data in .RData

I can't find a way to retrieve z values and p-values from the output from lmer in the lme4 package. How is this done? Rick B.

Is there a way (simple or not) to get the marginal prediction from lme (in nlme) and/or lmer (in lme4)? Rick B.

event.nab.2 is 0/1 and I dichotomized va to get va.2 to see if I could get geeglm to work. glm has no problem with the data but geeglm chokes. Each subject (patient.id) has at most 2 observations and more than 3/4 of the subjects have 2 observations. I have even worse problems trying to use glmmPQL from MASS and worse still trying to use lmer from lme4. But I figured a marginal model would work.

Using the weight argument with a variance function in lme (nlme), you can allow for heteroscedasticity of the within-group error. Is there a way to do this for the other variance components? For example, suppose you had subjects, days nested within subjects, and visits nested within days within subjects (a fully nested two-way design) and you had, say men and women subjects. Could you allow for

I'm running simulation using lme and sometimes the estimated variance-covariance matrix is not positive definite so that the intervals function won't work for the random effect coefficients. I've tried varcomp from the ape package but this does not return all the coefficients. How can I extract the random effect coefficients without using intervals? Rick B.

The publisher requires embedded fonts and the eps files that R produces don't pass its tests. How can I force the fonts (just default fonts) to be embedded. embedFonts seems to only embed unusual fonts, not the default ones that are used in a simple eps generated by using "postscript". I've tried: ps2pdf14 -dPDFSETTINGS=/prepress bland-altman_v2.eps bland-altman_v2.pdf pdftops