gaussian_weight: Estimate LVM With Weights

In lavaSearch2: Tools for Model Specification in the Latent Variable Framework

Estimate LVM With Weights

Description

Estimate LVM with weights.

Usage

gaussian_weight.estimate.hook(x, data, estimator, ...)

gaussian_weight_method.lvm

gaussian_weight_logLik.lvm(object, type = "cond", p, data, weights, ...)

gaussian_weight_objective.lvm(x, ...)

gaussian_weight_score.lvm(
  x,
  data,
  p,
  S,
  n,
  mu = NULL,
  weights = NULL,
  debug = FALSE,
  reindex = FALSE,
  mean = TRUE,
  constrain = TRUE,
  indiv = FALSE,
  ...
)

gaussian_weight_gradient.lvm(...)

gaussian_weight_hessian.lvm(x, p, n, weights = NULL, ...)

Arguments

x, object	A latent variable model
data	dataset
estimator	name of the estimator to be used
...	passed to lower level functions.
type	must be "cond"
p	parameter value
weights	weight associated to each iid replicate.
S	empirical variance-covariance matrix between variable
n	number of iid replicates
mu	empirical mean
debug, reindex, mean, constrain, indiv	additional arguments not used

Format

An object of class character of length 1.

Examples

#### linear regression with weights ####

## data
df <- data.frame(Y = c(1,2,2,1,2),
                 X = c(1,1,2,2,2),
                 missing = c(0,0,0,0,1),
                 weights = c(1,1,2,1,NA))

## using lm
e.lm.GS <- lm(Y~X, data = df)
e.lm.test <- lm(Y~X, data = df[df$missing==0,], weights = df[df$missing==0,"weights"])

## using lvm
m <- lvm(Y~X)
e.GS <- estimate(m, df)
## e.lava.test <- estimate(m, df[df$missing==0,], weights = df[df$missing==0,"weights"])
## warnings!!
e.test <- estimate(m, data = df[df$missing==0,],
                   weights = df[df$missing==0,"weights"],
                   estimator = "gaussian_weight")

lavaSearch2 documentation built on April 12, 2023, 12:33 p.m.