orthoKMeansTrainCpp(x, rounds = 1L, k = 3L, iter = 100L, initType = 0L,
verbose = FALSE)
Arguments
x
data matrix
rounds
number of rounds (orthogonal views)
k
number of clusters
iter
numer of iterations in one round
initType
centroid initialization via Random or KMeans++
verbose
verbose output?
Value
a list consisting of
centers these are the resulting centroids of the kmean algorithm (as a std::vector of NumericMatrix)
cluster these are the labels for the resulting clustering (as a std::vector of NumericVector)
obj this is a vector with the final objective value for each round