residplot: Plot the Residuals of a LVM Object

In bozenne/butils: Various useful functions

Plot the Residuals of a LVM Object

Description

Plot the residuals for each outcome of a lvm object.

Usage

residplot(object, ...)

## S3 method for class 'lvmfit'
residplot(
  object,
  res.variables = endogenous(object),
  obs.variables = res.variables,
  sd.kernel = 0.5,
  kernel = "dnorm",
  plot.weights = FALSE,
  ncol = NULL,
  smooth.mean = TRUE,
  smooth.sd = TRUE,
  plot = TRUE,
  ...
)

Arguments

object	a lvm model.
...	additional arguments.
res.variables	the endogenous variable for which the residuals should be displayed.
obs.variables	same as res.variables or a variable present in the model
sd.kernel	the standard deviation of the kernel used to smooth the variance.
kernel	the type of kernel used to smooth the variance.
plot.weights	should the weights used to compute the variance be displayed?
ncol	the number of columns in the graphical display.
smooth.mean	should the mean be displayed across the obs.variables values?
smooth.sd	should the variance be displayed across the obs.variables values?
plot	should the fit be displayed in a graphical window.

Value

a list containing:

• plot the display of the fit

• data the dataset used to make the display

Examples

m <- lvm(y1~eta+x1,y2~eta,y3~eta+x3)
latent(m) <- ~ eta
dd <- lava::sim(m,100) ## Simulate 100 observations from model
e <- estimate(m, dd) ## Estimate parameters

res <- residplot(e)
residplot(e, obs.variables = "x1")

m <- lvm(y~x)
distribution(m,~y) <- function(n,mean,x) rnorm(n,mean,exp(x)^.5)
d <- lava::sim(m,1e3)

e <- estimate(m,data = d)

residplot(e)
# residplot(e, plot.weights = TRUE)
residplot(e, res.variables = "y", obs.variables = "x")

bozenne/butils documentation built on Oct. 14, 2023, 6:19 a.m.