initialise_loading: Initialis Loading
In jpmeagher/vbar: Variational Bayes Ancestral Reconstruction

initialise_loading

R Documentation

Initialis Loading

Description

Initialise loadings for a PLVM. Allows for either a random initialisation, which does not depend on the auxiliary traits, or initialises with the first L principle components, which does.

Usage

initialise_loading(
  D_prime,
  L,
  ard_precision,
  loading_prior_correlation,
  loading = NULL,
  method = "random",
  random_seed = NULL,
  auxiliary_traits = NULL,
  perform_checks = TRUE
)

Arguments

`D_prime`	An integer. The dimension of auxiliary traits traits.
`L`	An integer. The dimension of latent traits.
`ard_precision`	An L-dimensional vector of real numbers. The initial ARD precision hyper-parameters on the loadings.
`loading_prior_correlation`	A D'xD' correlation matrix. The correlation within loadings under the Gaussian prior.
`loading`	A D'xL matrix of real values. The initial loading matrix. When non null checks that the loading matrix is of the correct type and size.
`method`	A string. Either "random" or "pca" indicating the type of initialisation adopted.
`random_seed`	A single value, interpreted as an integer, or NULL.
`auxiliary_traits`	An NxD' matrix of real numbers. The initial matrix of auxiliary traits. Only required when method == "pca"
`perform_checks`	Logical. Check if function inputs are specified correctly.