similar to: stopping stepAIC (MASS package)

Dear R-listers, I am using R for Windows Vista 64-bit. I have no experience working with maps, but now I have to plot a map of Denmark in which each Danish county is color-coded according to its incidence rate of a particular disease. My problem is that I don't know where to start. I have read the help files to packages like 'mapplot', and their examples indicate that I am on the

Dear R-listers, I am trying to assign colnames to a data frame within a loop, but I keep getting a "target of assignment expands to non-language object"-error. I need to split up a large dataset into about 20 smaller ones, and I would like to assign colnames within the loop, so I won't have to type the same thing 20 times over. I have concocted this really goofy example which

Hi, I am writing an Sweave document and am using 'xtable' to make frequency tables of diagnoses of people undergoing cholecystectomy. Some of these diagnoses contain Danish characters ("?", "?", and "?"), and these characters are all garbled in the Latex document after I run Sweave. The odd thing is, everything looks absolutely right in the R console, and if

I keep getting the same error in my survival analysis. I have access to a very large database but am just using small subsets to get some results. In this particular subset there is 50 explanatory variables(both factors of many levels and covariates) and 117 data pieces with some of the data being censored. I am using the stepAIC command to find my model. My initial model is built from all

Dear R-listers, I am a relatively inexperienced R-user currently migrating from Stata. I am deeply frustrated by this data manipulation question: I know how I could do it in Stata, but I cannot make it work in R. I have a data frame of hospitalization data where each row represents an admission. I need to know when patients were first discharged, but the problem is that patients were sometimes

Hello, I'm currently dealing with a rather strange problem when using the function "stepAIC" ("MASS" package). The setting is the following: From model learning data sets ("learndata"), I want to be able to build prediction functions (in order to save them in a file for further use). This is done by the function "pred.function" (see below). Therein,

In a generalized linear model with k covariates, there are 2(kth power) - 1 possible models (excluding interactions). Awhile ago a posting to R-help suggested Model Selection and Multimodel Inference, 2nd ed, by Burnham and Anderson as a good source for understanding model selection. They recommend (page 71) computing AIC differences over all candidate models in the set of possible models. After

Dear all, I have tried to modify the code of MASS:::stepAIC.default(), dropterm() and addterm() to use AICc instead of AIC for model selection. The code is appended below. Somehow the calculations are still not correct and I would be grateful if anyone could have a look at what might be wrong with this code... Here is a working example: ## require(nlme) model1=lme(distance ~ age + Sex, data =

Hi, the following R lines work fine in R 2.4.0 alpha (and older R versions), but not in R 2.4.0 beta (details below): library(drc) # to load the dataset 'PestSci' library(nlme) ## Starting values sv <- c(0.328919, 1.956121, 0.097547, 1.642436, 0.208924) ## No error m1 <- nlme(SLOPE ~ c + (d-c)/(1+exp(b*(log(DOSE)-log(e)))), fixed =

Hi, the following R lines work fine in R 2.4.0, but not in R 2.4.1 or any devel versions of R 2.5.0 (see below for details). library(drc) # to load the dataset 'PestSci' library(nlme) ## Setting starting values sv <- c(0.43355869, 2.49963220, 0.05861799, 1.73290589, 0.38153146, 0.24316978) ## No error m1 <- nlme(SLOPE ~ c + (d-c)/(1+exp(b*(log(DOSE)-log(e)))), fixed =

Dear R-Help, I have used R2.6.0 until I recently installed also R2.7.2 (see details below) In R 2.6.0, the following script using the function 'var' (cor(stats)): x.test <- c(NA, NA, NA, NA) var(x.test, na.rm=T) gives the output: NA In R2.7.2 the output of the same script generates an error message and stops R: 'Error in var(x.test, na.rm = T)

Dear list members, is it intentional that: curve(cos, xlim = c(-5, 5)) plot(cos, xlim = c(-5, 5)) produce different plots? Shouldn't the 'xlim' argument in both cases set the 'from' and 'to' argument if they aren't supplied (at least that's what I understood reading the help page for 'curve' and 'plot.function')? Christian >

Dear friends - I am trying to understand fda and working with the 2009 book from Springer. I run the scripts directly from a new installed fda library and include sessionInfo() below. This is from the script fdarm-ch09.R - I seem to have got not fda_2.2.8 but 2.2.7 - is that the problem? Where do I get .8 as I just reinstalled? Best wishes Troels Ring, Nephrology Aalborg, Denmark >

Dear list, Please consider the following call of sort > sort(c("a","f")) [1] "a" "f" > sort(c("f","a")) [1] "a" "f" > > sort(c("aa","ff")) [1] "ff" "aa" > sort(c("ff","aa")) [1] "ff" "aa" The last two results look strange

I have created an R-package with datasets which I want my students to install (the package is not on CRAN). 1) I've put the package on the web in a directory called 'data' and I thought I could do: > install.packages("http://gbi.agrsci.dk/statistics/courses/2011-ISMLS-course/data/LiSciData_0.0-03.tar.gz",repos=NULL,type="source") Warning: invalid package

I would like to do a SS type III analysis on a proportional odds logistic regression model. I use drop1(), but dropterm() shows the same behaviour. It works as expected for regular main effects models, however when the model includes an interaction effect it seems to have problems with matching the parameters to the predictor terms. An example: library("MASS"); options(contrasts =

Hi, I would appreciate of someone could explain how the residual standard error is computed for rlm models (MASS package). Usually, one would expect to get the residual standard error by > sqrt(sum((y-fitted(fm))^2)/(n-2)) where y is the response, fm a linear model with an intercept and slope for x and n the number of observations. This does not seem to work for rlm models and I am wondering

Full_Name: Andrew Robinson Version: 2.6.2 Patched (2008-02-12 r44439) OS: FreeBSD 6.3-RC1 Submission from: (NULL) (211.28.206.186) I believe that the computation for GCV is incorrect in the lm.ridge function in MASS. >From lm.ridge: GCV <- colSums((Y - X %*% coef)^2)/ (n - colSums(matrix(d^2/div, dx)))^2 The denominator does not tally with the formula on p. 141 of Ripley's

I have two specific questions regarding the output of lda function in MASS. #Question1: #========= n: sample size, p: number of variables Some articles in the literature say that LDA is singular for p > n-1. However, my experimentation with lda (default arguments) for two class problems shows collinearity for p > n-2. Does anyone know why this is the case? Does lda (MASS) use a different

The example in the help file for pdf() plots the characters outside the plotting area in this for loop: for(i in c(32:255)) { x <- i y <- i points(x, y, pch=i) } The following loop seems to be working as intended for(i in c(32:255)) { x <- (i-31)%%16 y <- (i-31)%/%16 points(x, y, pch=i) } As an extra