update_W: Update function for expansion points (~predicted latent...

In koekvall/lvmmr: Mixed-type multivariate response regression

Description

Update function for expansion points (~predicted latent variables)

Usage

update_W(
  Y,
  X,
  W,
  Beta,
  Sigma,
  psi,
  type,
  pen = 1e-04,
  tol = 1e-08,
  maxit = 100,
  quiet = T
)

Arguments

Y	Matrix (n x r) of responses.
X	Matrix (nr x p) of predictors.
W	Matrix (n x r) of expansion points (~predicted latent variables).
Beta	Vector (p by 1) of regression coefficients.
Sigma	Covariance matrix(r x r) of the latent vector.
psi	Vector (r x 1) of conditional variance parameters.
type	Vector (r x 1) indicating response types.
pen	Regularization parameter lower bouding the eigenvalue of a covariance matrix and multiplying an L2 penalty (see paper).
tol	Tolerance parameter for trust region algorithm
maxit	Maximum number of iterations for trust region algorithm.
quiet	Logical indicating whether to print information while running.

Value

Matrix of updated expansion points

koekvall/lvmmr documentation built on Dec. 13, 2021, 2:35 p.m.