Predict function for latent variable models
Predictions of conditinoal mean and variance and calculation of jacobian with respect to parameter vector.
predictlvm(object, formula, p = coef(object), data = model.frame(object), ...)
object |
Model object |
formula |
Formula specifying which variables to predict and which to condition on |
p |
Parameter vector |
data |
Data.frame |
... |
Additional arguments to lower level functions |
predict.lvm
m <- lvm(c(x1,x2,x3)~u1,u1~z,
c(y1,y2,y3)~u2,u2~u1+z)
latent(m) <- ~u1+u2
d <- simulate(m,10,"u2,u2"=2,"u1,u1"=0.5,seed=123)
e <- estimate(m,d)
## Conditional mean given covariates
predictlvm(e,c(x1,x2)~1)$mean
## Conditional variance of u1,y1 given x1,x2
predictlvm(e,c(u1,y1)~x1+x2)$varPlease choose more modern alternatives, such as Google Chrome or Mozilla Firefox.