plot.glmmlasso: Diagnostic Plots for a lmmlasso object

In glmmixedlasso: Generalized Linear Mixed Models with Lasso

Description

Plots four diagnostic plots for checking the model assumptions and supporting model selection for a glmmlasso object

Usage

## S3 method for class 'glmmlasso'
plot(x, ...)

Arguments

x	a lmmlasso object
...	not used.

Details

plot.glmmlasso shows four diagnostic plots which support checking the model assumption, model fit and may give hints for another model. 1) The first plot depicts the Tukey-Anscombe plot on the predictor scale. Points with the same color belong to the same group. 2) Plot depending on the GLM family. For family="poisson", the fitted values against the observed values is shown. For family="binomial", the Tukey-Anscombe plot on the response scale is depicted. 3) QQ-Plot of the predicted random effects. Be careful with the interpretation since the random effects have not been standardized. The color shows which points belong to the same random-effects covariate. 4) A histogram of the fixed-effects coefficients. For the interpretation of the Tukey-Anscombe plot in GLMs, see Faraway (2006).

References

Julian J. Faraway (2006) Extending the Linear Model with R, Chapman and Hall/CRC.

Examples

data(xerop)
fit <- glmmlasso(y=xerop$y,group=xerop$group,X=xerop$X,Z=xerop$Z,
        family="binomial",covStruct="Identity",lambda=30)
plot(fit)

glmmixedlasso documentation built on May 2, 2019, 5:26 p.m.