Plots the outcome probabilities for a randomLCA object, for random effects objects this can be either marginal or conditional or both. For a 2 level random effects model conditional2 will condition on the subject random effect and integrate over the period random effects. Note that plot is based on the xyplot function.
Usage
## S3 method for class 'randomLCA'
plot(x, ... , graphtype=ifelse(x$random, "marginal","conditional"),
conditionalp=0.5,classhorizontal=TRUE)
Arguments
x
randomLCA object
graphtype
Type of graph
conditionalp
For a conditional graph the percentile corresponding to the random effect at which the outcome probability is to be calculated
## Not run:
# standard latent class with 2 classes
uterinecarcinoma.lca2 <- randomLCA(uterinecarcinoma[,1:7],freq=uterinecarcinoma$freq)
plot(uterinecarcinoma.lca2)
uterinecarcinoma.lcarandom2 <- randomLCA(uterinecarcinoma[,1:7],
freq=uterinecarcinoma$freq,random=TRUE,probit=TRUE,quadpoints=61)
# default for random effects models is marginal
plot(uterinecarcinoma.lcarandom2)
# default for random effects models conditional is p=0.5 i.e. median
plot(uterinecarcinoma.lcarandom2,graphtype="conditional")
# look at variability by plotting conditional probabilities at 0.05,0.5 and 0.95
plot(uterinecarcinoma.lcarandom2,graphtype="conditional",conditionalp=c(0.05,0.5,0.95))
## End(Not run)