knitr::opts_chunk$set( collapse = TRUE, comment = "#>", fig.path = "README-" )
The rust
package implements the multivariate generalized ratio-of-uniforms method of simulating random variates from a d-dimensional continuous distribution. The user specifies (the log of) a positive target function f
that is proportional to the density function of the distribution.
We use the function ru
to simulate a sample of size 1000 from a two-dimensional standard normal distribution with strong positive correlation between the components. Of course, this particular example is purely illustrative: there are better ways to simulate from a multivariate normal distribution.
rho <- 0.9 covmat <- matrix(c(1, rho, rho, 1), 2, 2) log_dmvnorm <- function(x, mean = rep(0, d), sigma = diag(d)) { x <- matrix(x, ncol = length(x)) d <- ncol(x) - 0.5 * (x - mean) %*% solve(sigma) %*% t(x - mean) } x <- ru(logf = log_dmvnorm, sigma = covmat, d = 2, n = 1000, init = c(0, 0))
From version 1.2.0 onwards the faster function ru_rcpp
can be used. See the vignette "Rusting Faster: Simulation using Rcpp" for details.
# Create an external pointer to a C++ function to evaluate the log-density. ptr_bvn <- create_xptr("logdnorm2") # Pass the external pointer to `ru_rcpp`. x <- ru_rcpp(logf = ptr_bvn, rho = rho, d = 2, n = 1000, init = c(0, 0))
To get the current released version from CRAN:
install.packages("rust")
See vignette("rust-a-vignette", package = "rust")
for an overview of the package,
vignette("rust-b-when-to-use-vignette", package = "rust")
for guidance on when rust
can be used and vignette("rust-c-using-rcpp-vignette", package = "rust")
for information on how to take advantage of the Rcpp package.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.