doMIsaul (version 1.0.1)

initiate_centers: Initiate centers for clustering algorithm

Description

Initiate centers for clustering algorithm

Usage

initiate_centers(data, N = 1000, t = 1, k, algorithms = NULL, seeds.N = NULL)

Arguments

data

Dataset that clustering will be applied on

N

Integer. Number clustering initialization (set of centers) to generate

t

Numeric between 0 and 1. weight coefficient between only random centers (t=1) and only centers from clustering (t=0).

k

Vector of size N containing the number of centers for each initialization.

algorithms

list of algorithm(s) (size N * (1-t) to generate centers if t!=1, given as characters. Possible values are "km" for 'K-means', "kmed" for 'K-medians', "hclust.mean", "hclust.med" for hierarchical clustering with mean or median position of the center.

seeds.N

(optional) vector of size N containing seeds for each initialization.

Value

list of size N containing coordinates of centers for clustering initialization.

Examples

# NOT RUN {
Cent.init <- initiate_centers(data = iris[, 1:4], N = 10,
                              k = sample(c(2:7), 10, replace = TRUE))
# }