similar to: solved mystery of difference between glmmPQL and lme

I have a longitudinal data analysis project. There are 10 observations on each of 15 units, and I'm estimating this with randomly varying intercepts along with an AR1 correction for the error terms within units. There is no correlation across units. Blundering around in R for a long time, I found that for linear/gaussian models, I can use either the MASS method glmmPQL (thanks to

On a RedHat 7.3 system with R-1.6.1, I did this > x11(width=3.5,height=4,colortype="gray") Then plotted (with matplot) a nice looking no-color graph on the screen, then I did this: > dev.copy2eps(file="test.eps",height=4,width=3.5) I was surprised that the output in the eps file included the colored lines from the plot, even though the screen device was set to

I handed out some results from glm() and the students ask "how many observations were dropped due to missing values"? How would I know? In other stat programs, the results will typically include N and the number dropped because of missings. Without going back to R and fiddling about to find the total number of rows in the dataframe, there is no way to tell. Somewhat

Hello, everybody: My friends in Ukraine are starting a research lab at a national university and they asked what programs to use. I said "R" of course, and they then asked me 'what support does it have for Cyrillic'? i've done some snooping in the R website and all the references i find to foreign languages concern c and fortran, not Ukrainian or Russian. Since i'm an

Dear R users who might like to use the package RNetCDF on Fedora Linux: Fedora (versions 4 and 5) users might have noticed that the default install of the netcdf and netcdf-devel packages from the Fedora Extra archive is inconsistent with the R package RNetCDF. The attempt to install RNetCDF results in a failure in the configure stage because the header & library info for netcdf cannot

Dear everybody: We used the fine foreign library to bring in an SPSS dataset that was about 9 megabytes and I can squeeze it into a much smaller R object using compression with save(ndat, file="NatAnnES2000.rda", compress=T). I can use load() to get the "ndat" dataframe back, that's all good as far as I can see. If I put that file in the data subdirectory, then the

OK, my journey to make lab machines automagically install & update all desirable R packages is nearing an end! The only question I have now is this: How can I control where the system prints the *.Rout file that is created automatically when the R batch program runs. In "man R" I don't find any information about it. When the cron job runs "R_installAll.sh" (see

Dear Everybody: I have Fedora Core 2 and R-1.9.0 does not build "out of the box". After applying the daily patch file R-release.diff, I find it does build and I've made RPMS and SRPM. In case you want to save yourself a recompile, you can find Fedora Core RPMs in here: http://lark.cc.ku.edu/~pauljohn/software/R These are based on the standard R distribution SPEC file, only

Last Friday, I noticed that it is difficult to work with regression models in which there are factors. It is easier to do the old fashioned thing of coding up "dummy" variables with 0-1 values. The predict function's newdata argument is not suited to insertion of hypothetical values for the factor, whereas it has no trouble with numeric variables. For example, if one uses a

I have a linux system with Fedora Core 2 and R-2.0. I was comparing plots made with plot() and dotplot() and discovered a problem. Although the dots are positioned correctly, the numerical labels in the dotplot y axis are not correct. I put copies here: http://lark.cc.ku.edu/~pauljohn/R/plotTrouble1.jpg That is the "correct" one from plot, with the higest value on y showing at 18.

Hello, everybody: On a Fedora Core 3 Linux system, I built R-2.1 using an updated version of the spec file that was used to make the RPMs for version 2.0.1 on the CRAN system. The build was fine, and packages updates perfectly. Thanks! Then I got curious about the package gnomeGUI. While trying to build that, I see errors ============= * Installing *Frontend* package 'gnomeGUI' ...

Greetings: I'm running R-1.9.1 on Fedora Core 2 Linux. I tested a proportional odds logistic regression with MASS's polr and Design's lrm. Parameter estimates between the 2 are consistent, but the standard errors are quite different, and the conclusions from the t and Wald tests are dramatically different. I cranked the "abstol" argument up quite a bit in the polr

I've been reading the R mail archives and I've found a lot of messages with this same kind of problem, but I can't understand the answers. Can one of you try to explain this to me? Here's my example. Given a regression model and a variable, I want to use unstack() on the vector of residuals and make some magic with the result. But unstack hates me. PCSE <- function

Concerning this article, Christopher Zorn, "Generalized Estimating Equation Models for Correlated Data: A Review with Applications." 2001. American Journal of Political Science 45(April):470-90. The author very kindly provides data for replication on his web page: http://www.emory.edu/POLS/zorn/Data/GEE.zip. I've been comparing the Professor Zorn's results obtained with

I've used the HDF5 library to bring some data into R. THe verbose output looks like this: > hdf5load("hdfGraphWed_Mar_16_13_33_37_2005.hdf",load=T,verbosity=1,tidy=T) Processing object: cprSeats ...... which is a Group Processing object: Seats 0 ...... its a dataset......Finished dataset Processing object: Seats 1 ...... its a dataset......Finished dataset Processing object:

Good day: I'm administering 6 linux systems (FC4) in a student lab and worry that users may want packages that are not installed. I get tired of adding them one by one. Then I happened upon this page http://support.stat.ucla.edu/view.php?supportid=30 about installing all R packages from CRAN. That did not run as it was, but after some fiddling I arrived at the following script, which

Dear R users: I assigned students to make some graphs and I'm having trouble answering some questions that they have. We are all working on R 2.1 on Fedora Core Linux 4 systems. 1. In the plot, the axis is not labeled by "numbers", but rather scientific notation like "-2e+08" or such. We realize that means -200,000,000. We want to beautify the plot. We would rather

Greetings, R-help! On 2 Fedora Core 2 Linux systems, i've completely erased the previous R and all packages and then installed R-2.0 and installed fresh packages. In using Rcmdr, I see some trouble and I wonder if other people see this and if it is due to the tcl/tk, or R, or Rcmdr. (If readers have not yet tried Rcmdr, I recommend it not just because of the GUI it provides, but also

HDF5 is a high quality data storage format from NCSA (http://hdf.ncsa.uiuc.edu/index.html) Marcus Daniels of the Swarm Development Group has made available an R library called hdf5 which allows loading and saving of data in hdf5 format. The source for his library is here: ftp://ftp.swarm.org/pub/swarm/src/testing/hdf5_1.4.7.tar.gz I prefer to use RPMs to maintain software across many linux

Greetings, everybody. Can I ask some glm questions? 1. How do you find out -2*lnL(saturated model)? In the output from glm, I find: Null deviance: which I think is -2[lnL(null) - lnL(saturated)] Residual deviance: -2[lnL(fitted) - lnL(saturated)] The Null model is the one that includes the constant only (plus offset if specified). Right? I can use the Null and Residual deviance to