reduce: Reduce latent variable model

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

Description

Aggregate exogeneous variables into a linear predictor

Usage

reduce(x, ...)

## S3 method for class 'lvm'
reduce(x, link = NULL, endo = NULL, clean = TRUE, ...)

Arguments

x	lvm-object
...	additional arguments to be passed to the low level functions
link	the links that should be aggregated into linear predictors
endo	the endogeneous variables for which the related exogeneous variables should be aggregated
clean	should the lvm object be simplified using the clean function

Examples

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

mtest <- m
cancel(mtest) <- y ~ x1 + x2 + x3 + x4 + x5
mtest
coef(mtest)
rm <- reduce(m)
rm
lp(rm, type = "link")
coef(rm)

coef(reduce(m))
reduce(m, rm.exo = TRUE)

reduce(m, link = paste0("y~x",1:5))

## lvm
m <- lvm()
m <- regression(m, x=paste0("x",1:10),y="y1")
m <- regression(m, x=paste0("x",11:30),y="y2")
covariance(m) <- y1~y2

system.time(
rm <- reduce(m)
)
coef(rm)
reduce(m, rm.exo = TRUE)

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