kill.lp: Remove variables from linear predictors

In bozenne/lava.reduce: Latent Variable Models with linear predictor

Description

Remove one or several variable from the linear predictors

Usage

lavaReduce.remove.hook(x, var, expar = TRUE, restaure = FALSE, ...)

lavaReduce.cancel.hook(x, value, expar = TRUE, restaure = FALSE, ...)

Arguments

x	lvm-object
var	the names of the variables that should be removed
expar	should the external parameters be also removed
restaure	should the link be kept while removing the linear predictor
...	argument passed to clean
value	the names of the variables that should be removed

Examples

m <- lvm()
m <- regression(m, x=paste0("x",1:10),y="y")
kill(m) <- "x7"

rm <- reduce(m, rm.exo = FALSE)

kill(rm) <- ~x6
kill(rm, value = "x1")
kill(rm, value = c("x2","x3"))
kill(rm, value = ~x4+x5)
kill(rm, value = ~LPy)


m <- lvm.reduced()
regression(m) <- y1 ~ x1+x2+x3
regression(m) <- y2 ~ z
covariance(m) <- y1 ~ y2
cancel(m) <- y1~x1+x2+x3
m
cancel(m) <- "y1 ~~ y2"
coef(m)

# see test/testthat/test-cancel.R

bozenne/lava.reduce documentation built on May 13, 2019, 1:41 a.m.