Use crossprod() and tcrossprod() to replace t(x) %*% y and x %*% t(y)…

… respectively
YangLabHKUST · Jan 30, 2024 · ee7702e · ee7702e
1 parent bf89d38
commit ee7702e
Showing 1 changed file with 15 additions and 9 deletions.
diff --git a/R/mfairELBO.R b/R/mfairELBO.R
@@ -30,14 +30,20 @@ getELBO <- function(Y, object, obs_indices = NULL) {
 
  if (!is.null(obs_indices)) { # Sparse mode
  elbo2 <- -tau *
- sum((Y - projSparse(as.matrix(mu) %*% t(nu), obs_indices))^2 +
- projSparse(as.matrix(mu_sq + a_sq) %*% t(nu_sq + b_sq), obs_indices) -
- projSparse(as.matrix(mu_sq) %*% t(nu_sq), obs_indices))
+ sum((Y - projSparse(
+ tcrossprod(as.matrix(mu), as.matrix(nu)), obs_indices
+ ))^2 +
+ projSparse(
+ tcrossprod(as.matrix(mu_sq + a_sq), as.matrix(nu_sq + b_sq)), obs_indices
+ ) -
+ projSparse(
+ tcrossprod(as.matrix(mu_sq), as.matrix(nu_sq)), obs_indices
+ ))
  } else {
  elbo2 <- -tau * sum(
- (Y - as.matrix(mu) %*% t(nu))^2 +
- as.matrix(mu_sq + a_sq) %*% t(nu_sq + b_sq) -
- as.matrix(mu_sq) %*% t(nu_sq),
+ (Y - tcrossprod(as.matrix(mu), as.matrix(nu)))^2 +
+ tcrossprod(as.matrix(mu_sq + a_sq), as.matrix(nu_sq + b_sq)) -
+ tcrossprod(as.matrix(mu_sq), as.matrix(nu_sq)),
  na.rm = TRUE
  ) / 2
  }
@@ -49,9 +55,9 @@ getELBO <- function(Y, object, obs_indices = NULL) {
  elbo6 <- sum(log(2 * pi * b_sq)) / 2 + M / 2
  } else {
  elbo1 <- -N * M * log(2 * pi / tau) / 2
- elbo2 <- -tau * sum((Y - as.matrix(mu) %*% t(nu))^2 +
- as.matrix(mu_sq + a_sq) %*% t(nu_sq + b_sq) -
- as.matrix(mu_sq) %*% t(nu_sq)) / 2
+ elbo2 <- -tau * sum((Y - tcrossprod(as.matrix(mu), as.matrix(nu)))^2 +
+ tcrossprod(as.matrix(mu_sq + a_sq), as.matrix(nu_sq + b_sq)) -
+ tcrossprod(as.matrix(mu_sq), as.matrix(nu_sq))) / 2
  elbo3 <- -N * log(2 * pi / beta) / 2 -
  beta * (sum(mu_sq) + N * a_sq -
  2 * sum(mu * FX) + sum(FX^2)) / 2