R/RcppExports.R

In EandrewJones/mmBPFA: Multi-Modal Beta-Process Factor Analysis

# Generated by using Rcpp::compileAttributes() -> do not edit by hand
# Generator token: 10BE3573-1514-4C36-9D1C-5A225CD40393

#' Sampling step for alpha (item-intercepts)
#'
#' @keywords internal
#'
#' @param x A multivariate Gaussian copula matrix of dimenion N x P.
#' @param lambda A standard multivariate normal matrix of factor loadings of dimension P x K.
#' @param omega A standard multivariate normal matrix of factor scores of dimenion N x K.
#'
#' @return alpha Numeric vector of item-intercepts of length P.
#'
sample_alpha <- function(x, lambda, omega) {
    .Call(`_mmBPFA_sample_alpha`, x, lambda, omega)
}

#' Sampling step for gamma_k (factor precisions)
#'
#' @keywords internal
#'
#' @param zeros A binary, sparsity inducing matrix of dimenion P x K.
#' @param lambda A standard multivariate normal matrix of factor loadings of dimension P x K.
#' @param c A tunable hyperparameter for the shape parameter of the gamma draw. Default is 0.
#' @param d A tunable hyperparameter for the inverse-scale (rate) parameter of the gamma draw. Default is 100
#'
#' @return gamma_k Numeric vector of factor precisions of length K.
#'
sample_gamma_k <- function(zeros, lambda, c = 0, d = 100) {
    .Call(`_mmBPFA_sample_gamma_k`, zeros, lambda, c, d)
}

#' Sampling step for d (hyperparameter for sample_gamma_k)
#'
#' @keywords internal
#'
#' @param zeros A binary, sparsity inducing matrix of dimenion P x K.
#' @param gamma_k A gamma-distributed vector of factor precisions of length K.
#' @param c A tunable hyperparameter for the shape parameter of the gamma draw. Default is 1.
#' @param c0 A tunable hyperparameter for the shape parameter of the gamma draw.
#' @param d0 A tunable hyperparameter for the inverse-scale (rate) parameter of the gamma draw.
#'
#' @return scalar double
#'
sample_d <- function(zeros, gamma_k, c, c0, d0) {
    .Call(`_mmBPFA_sample_d`, zeros, gamma_k, c, c0, d0)
}

#' Sampling step for Indian Buffet Process Alpha Parameter
#'
#' @keywords internal
#'
#' @param ibp_b A scalar double, Indian Buffet Process beta parameter.
#' @param zeros A binary, sparsity inducing matrix of dimenion P x K.
#' @param e A scalar double, tunable hyperparameter for the shape parameter of the gamma draw. Default is 1.
#' @param f A scalar double, tunable hyperparameter for the rate parameter of the gamma draw. Default is 1.
#'
#' @return ibp_a Scalar double, Indian Buffet Process alpha parameter.
#'
sample_IBP_a <- function(ibp_b, zeros, e = 1, f = 1) {
    .Call(`_mmBPFA_sample_IBP_a`, ibp_b, zeros, e, f)
}

#' Metropolis-Hastings sampling step for Indian Buffet Process Beta Parameter
#'
#' @keywords internal
#'
#' @param ibp_a A scalar double, Indian Buffet Process alpha parameter.
#' @param ibp_b A scalar double, prior Indian Buffet Process Beta parameter .
#' @param zeros A binary, sparsity inducing matrix of dimenion P x K.
#'
#' @return ibp_b Scalar double, Indian Buffet Process beta parameter.
#'
sample_IBP_b <- function(ibp_a, ibp_b, zeros) {
    .Call(`_mmBPFA_sample_IBP_b`, ibp_a, ibp_b, zeros)
}

#' Sampling step for lambda and binary, sparsity-inducing zero matrix
#'
#' @keywords internal
#'
#' @param x A multivariate Gaussian copula matrix of dimenion N x P.
#' @param lambda A standard multivariate normal matrix of factor loadings of dimension P x K.
#' @param zeros A binary, sparsity inducing matrix of dimenion P x K.
#' @param omega A standard multivariate normal matrix of factor scores of dimenion N x K.
#' @param alpha A numeric vector of item-intercepts of length P.
#' @param gamma_k A numeric vector of factor precisions of length K.
#' @param dc A integer vector of counts for the number of times each dimensions has been sampled.
#' @param ibp_a A scalar double, Indian Buffet Process beta parameter.
#' @param ibp_b A scalar double, Indian Buffet Process beta parameter.
#' @param tau A tunable hyperparameter for the poisson draws of new dishes to be sampled.
#' @param sparse A boolean indicating whether to include sparsity-inducing prior.
#' @param infinite A boolen indicating whether to sample new potential dishes at each step.
#'
#' @return zeros A binary, sparsity inducing matrix of dimenion P x K.
#' @return lambda A standard multivariate normal matrix of factor loadings of dimension P x K.
#' @return dc A integer vector of counts for the number of times each dimensions has been sampled.
#'
sample_lambda_and_zeros <- function(x, lambda, zeros, omega, alpha, gamma_k, dc, ibp_a, ibp_b, tau, sparse, infinite) {
    .Call(`_mmBPFA_sample_lambda_and_zeros`, x, lambda, zeros, omega, alpha, gamma_k, dc, ibp_a, ibp_b, tau, sparse, infinite)
}

#' Sampling step for omega matrix (Factor Scores) of dimension N x K.
#'
#' @keywords internal
#'
#' @param lambda A standard multivariate normal matrix of factor loadings of dimension P x K.
#' @param x A multivariate Gaussian copula matrix of dimenion N x P.
#' @param alpha A numeric vector of item-intercepts of length P.
#'
#' @return omega A multivariate Gaussian matrix of factor scores of dimension N x K.
#'
sample_omega <- function(lambda, x, alpha) {
    .Call(`_mmBPFA_sample_omega`, lambda, x, alpha)
}