I've created a package with the new version of the MCLE code:
# Source the old code: library(sn) library(mnormt) library(rbenchmark) dir = "~/GitHub/Robust-Skew-t/" source(paste0(dir, "sn-funct.R")) source(paste0(dir, "modified_mst.pdev.grad.R")) source(paste0(dir, "robust_ST_for_mandy.R"))
Now, we can perform a quick speed comparison:
mst = list(dev = MCLE:::devMST, grad = MCLE:::gradDevMST, paramList2Vec = MCLE:::paramList2VecMST, paramVec2List = MCLE:::paramVec2ListMST) initialNew = list(xi = c(0, 0), Omega = diag(c(1, 1)), alpha = c(0, 0), nu = 100) initialOld = dplist2optpar(initialNew) benchmark( { data = matrix(rnorm(10), ncol = 2) robustST(data, start = initialOld) }, { data = matrix(rnorm(10), ncol = 2) MCLE::MCLE(data = data, dist = mst, initial = initialNew) } )
And now, let's try it with a larger sample size:
benchmark( { data = matrix(rnorm(500), ncol = 2) robustST(data, start = initialOld) }, { data = matrix(rnorm(500), ncol = 2) MCLE::MCLE(data = data, dist = mst, initial = initialNew) }, replications = 10)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.