Hi, I have been having a tedious issue with lmer models with lots of factors and lots of levels. In order to get the basic information at the beginning of the print out I also have to generate these enormous tables as well. Is there a method command to leave off all of the effects and correlations? Or, do I have to go to string commands?

On 7/4/06, John Christie <jc at or.psychology.dal.ca> wrote:> Hi, > I have been having a tedious issue with lmer models with lots of > factors and lots of levels. In order to get the basic information at > the beginning of the print out I also have to generate these enormous > tables as well. Is there a method command to leave off all of the > effects and correlations? Or, do I have to go to string commands?Depending on exactly which parts of the output you want it may be easiest to obtain them from the result of summary applied to your fitted model. Use str to determine the structure of the object then extract the slots that you want. Warning: This is a somewhat risky practice in that we don't guarantee that the classes and slot names will remain constant. If there is an extractor function that gives the result that you want it is much safer to use that than to reach in to a structure and extract individual slots or components. However, in this case I don't think such an extractor exists. As an example> (fm1 <- lmer(Reaction ~ Days + (Days|Subject), sleepstudy))Linear mixed-effects model fit by REML Formula: Reaction ~ Days + (Days | Subject) Data: sleepstudy AIC BIC logLik MLdeviance REMLdeviance 1753.628 1769.593 -871.8141 1751.986 1743.628 Random effects: Groups Name Variance Std.Dev. Corr Subject (Intercept) 612.090 24.7405 Days 35.072 5.9221 0.066 Residual 654.941 25.5918 number of obs: 180, groups: Subject, 18 Fixed effects: Estimate Std. Error t value (Intercept) 251.4051 6.8246 36.838 Days 10.4673 1.5458 6.771 Correlation of Fixed Effects: (Intr) Days -0.138> so <- summary(fm1) > showClass(class(so))Slots: Name: isG methTitle logLik ngrps sigma coefs Class: logical character logLik integer numeric matrix Name: vcov REmat AICtab flist Zt X Class: dpoMatrix matrix data.frame list dgCMatrix matrix Name: y wts wrkres method useScale family Class: numeric numeric numeric character logical family Name: call cnames nc Gp XtX ZtZ Class: call list integer integer dpoMatrix dsCMatrix Name: ZtX Zty Xty Omega L RZX Class: dgeMatrix numeric numeric list dCHMsuper dgeMatrix Name: RXX rZy rXy devComp deviance fixef Class: dtrMatrix numeric numeric numeric numeric numeric Name: ranef RZXinv bVar gradComp status assign Class: numeric dgeMatrix list list logical integer Name: frame terms Class: data.frame terms Extends: Class "summary.mer", directly Class "lmer", directly Class "mer", by class "summary.mer" Class "mer", by class "lmer"> so at AICtabAIC BIC logLik MLdeviance REMLdeviance 1753.628 1769.593 -871.8141 1751.986 1743.628> print(so at REmat, quote = FALSE)Groups Name Variance Std.Dev. Corr Subject (Intercept) 612.090 24.7405 Days 35.072 5.9221 0.066 Residual 654.941 25.5918

### Apparently Analagous Threads

- identifying convergence or non-convergence of mixed-effects regression model in lme4 from model output
- Matrix package problem: dsyMatrix %*% vector gives error
- question about the "solve" function in library "Matrix"
- expm() within the Matrix package
- R's AIC values differ from published values