Performance

Even without a performance testing, we are still sure the C++ implementations should be faster. A small wine data set from UCI data sets repository is used for benchmarking. rbenchmark package is also used.

suppressMessages(library(rbenchmark))
res <- benchmark(mlKmeans(t(wine),3),
                 kmeans(wine,3),
                 columns=c("test", "replications", "elapsed",
                 "relative", "user.self", "sys.self"), order="relative")

From benchmarking result, we can see that MLPACK version of k-means is 33-time faster than \texttt{kmeans()} in R. However, we should note that R returns more information than the clustering result and there are much more checking functions in R.

test	replications	elapsed	relative	user.self	sys.self
mlKmeans(t(wine), 3)	100	0.028	1.000	0.028	0.000
kmeans(wine, 3)	100	0.947	33.821	0.484	0.424